Transient temperature sensor data processing device and working method thereof

Abstract

The invention discloses a transient temperature sensor data processing device and a working method thereof, and belongs to the technical field of temperature sensor data processing. Comprising a transient temperature sensor, the transient temperature sensor is connected with a data processing unit, and a device outer shell is arranged outside the data processing unit; and the data processing unit internally comprises a data acquisition module, a signal processing module, a data analysis module, a data storage module, a display and output module and an early warning and alarming module. Due to the existence of the data analysis module, the automation and intelligence level of the system is remarkably improved, so that the system not only can passively collect and store data, but also can actively analyze and utilize the data. And the early warning and alarm module carries out corresponding early warning and alarm operation according to a data analysis result, and the efficiency in the data transmission process is improved through fault diagnosis and prediction analysis.

Description

Technical Field

[0001] The invention belongs to the technical field of transient temperature sensors, and in particular relates to a transient temperature sensor data processing device and a working method thereof. Background Art

[0002] A transient temperature sensor is a sensor that can respond quickly to temperature changes and can measure temperature changes in a short period of time. It is widely used in aerospace, automotive industry, medical equipment and other fields. The working principle of a transient temperature sensor is usually based on the following principles: Thermocouple principle: Use the thermoelectric potential difference generated at the contact point of two different metals to measure temperature changes. Thermistor principle: Use the property that the resistance value of a thermistor changes with temperature to measure temperature changes. Infrared radiation principle: Use the property that the intensity of infrared radiation radiated by an object is proportional to the temperature to measure temperature changes. In existing technologies, transient temperature sensors are increasingly used. They are used in various environmental monitoring, industrial control, scientific research and other fields.

[0003] Traditional transient temperature sensor data processing devices usually only include basic data acquisition and simple signal processing functions, which cannot meet the growing demand for in-depth and intelligent data analysis. They lack a comprehensive data analysis module and cannot achieve all-round analysis of temperature data such as preprocessing, feature extraction, statistical analysis, pattern recognition, predictive analysis, and fault diagnosis. This limits the user's ability to obtain valuable information from the data and affects the user's judgment. Summary of the invention

[0004] The purpose of the present invention is to overcome the problem that the existing transient temperature sensor data processing device can only perform simple signal processing and lacks specific methods for specific analysis, predictive analysis and fault diagnosis of data during data transmission, resulting in low data processing efficiency, and to provide a transient temperature sensor data processing device and its working method.

[0005] In order to achieve the above object, the present invention adopts the following technical solution:

[0006] A transient temperature sensor data processing device, characterized in that it comprises a transient temperature sensor, the transient temperature sensor is connected to a data processing unit, and a device outer shell is arranged outside the data processing unit;

[0007] The data processing unit comprises:

[0008] Data acquisition module: connected to the interface of the transient temperature sensor, used for real-time acquisition of the signal of the transient temperature sensor;

[0009] Signal processing module: used to convert the signal collected by the transient temperature sensor into a digital signal;

[0010] Data analysis module: used to analyze digital signals and obtain data analysis results;

[0011] Data storage module: used to store data analysis results;

[0012] Display and output module: used to display data analysis results;

[0013] Early warning and alarm module: used to monitor abnormal conditions during signal transmission and generate alarms if abnormal conditions are found;

[0014] Wherein, the data analysis module includes:

[0015] Data preprocessing submodule: used to convert digital signals into a form with zero mean and unit variance to obtain a data set;

[0016] Feature extraction submodule: used to extract features from the data set to obtain a data set with reduced dimension;

[0017] Statistical analysis submodule: used to calculate the mean and standard deviation of the reduced-dimensional data set to obtain statistical analysis results;

[0018] Pattern recognition submodule: used to separate data points of different categories in the data set according to the statistical analysis results to obtain pattern recognition results;

[0019] Prediction analysis submodule: used to perform prediction analysis on future data trends based on statistical analysis results to obtain prediction analysis results;

[0020] Fault diagnosis submodule: used to perform fault detection on the statistical analysis results and obtain fault diagnosis results;

[0021] Result output submodule: input statistical analysis results, pattern recognition results, prediction analysis results and fault diagnosis results to generate data analysis results.

[0022] A method for processing a transient temperature sensor comprises the following steps:

[0023] Real-time acquisition of transient temperature sensor signals;

[0024] Performing filtering and linearization processing on the collected signal of the transient temperature sensor to obtain a processed digital signal;

[0025] The processed data signal is converted into a form with zero mean and unit variance to obtain a data set;

[0026] The PCA algorithm is used to extract features from the data set to obtain a data set with reduced dimensions;

[0027] Calculate the mean and standard deviation of the reduced-dimensional data set to describe the central tendency and dispersion of the data points in the data set and obtain statistical analysis results;

[0028] Through the SVM algorithm, the data points of different categories in the data set are separated according to the statistical analysis results, the normal operation mode or abnormal mode in the system is identified, and the pattern recognition results are obtained;

[0029] According to the statistical analysis results, the future data trend is predicted and analyzed through the autoregressive moving average model to obtain the prediction analysis results;

[0030] Perform fault detection on the statistical analysis results through the FTA algorithm to obtain the fault diagnosis results;

[0031] Input statistical analysis results, pattern recognition results, predictive analysis results and fault diagnosis results to generate data analysis results;

[0032] The data analysis results are stored in the data storage module on the one hand, and presented to the user in a visual form through the display and output module on the other hand;

[0033] Monitor abnormal conditions during signal transmission and immediately issue an alarm if any abnormality is found.

[0034] A further improvement of the present invention is that the process of converting the processed data signal into a form having zero mean and unit variance is as follows:

[0035] For each digital signal feature, calculate its mean Mean and standard deviation Std;

[0036] For each eigenvalue x, the transformation is performed using the following formula:

[0037] Among them, x std is the transformed value;

[0038] After the above transformation, the mean of each digital signal feature is 0, the standard deviation is 1, and the distribution of the data set will be similar to the standard normal distribution.

[0039] A further improvement of the present invention is that the working method of extracting features from a data set by a PCA algorithm to obtain a data set with reduced dimensionality is:

[0040] Data standardization: ensure that the variance of each digital signal feature is 1 and the mean is 0;

[0041] Calculate the covariance matrix;

[0042] Compute the eigenvalues ​​and eigenvectors of the covariance matrix;

[0043] Select the main eigenvectors to form a new feature space;

[0044] Project the original data onto the new feature space.

[0045] The formula for calculating the mean and standard deviation of the reduced-dimensional dataset is as follows:

[0046] The formula for calculating the mean is: in, is the mean, n is the number of data points, x i is the i-th data point;

[0047] The formula for calculating the standard deviation is:

[0048] Where σ is the standard deviation, n is the number of data points, and x i is the ith data point, is the mean.

[0049] A further improvement of the present invention is that, by using the SVM algorithm, data points of different categories in the data set are separated according to the statistical analysis results, and the normal operation mode or abnormal mode in the system is identified. The operation steps for obtaining the pattern recognition result are as follows:

[0050] Use kernel functions to electrically map the data in the dataset into a high-dimensional space;

[0051] Construct an optimization problem to obtain the hyperplane with the farthest distance between data points;

[0052] Solve the optimization problem and obtain the support vector;

[0053] Use support vectors to build decision functions;

[0054] A further improvement of the present invention is that, in the case of linear separability of the SVM algorithm, the formula for the hyperplane with the farthest data point distance is:

[0055] The constraints are: i (w·x i +b)≥1;

[0056] Where: i = 1,…,n; w is the weight vector; b is the bias term; x i is the i-th eigenvector; y i is the category label of the i-th feature vector; n is the total number of feature vectors;

[0057] The nonlinear SVM algorithm formula using kernel function is:

[0058]

[0059] The constraints are: 0≤α i ≤C;

[0060] where i=1,…,n; α i is the Lagrange multiplier; C is the penalty parameter; K(x i ,x j ) is the kernel function used to calculate the eigenvector x i and x j The inner product between is in the high-dimensional feature space.

[0061] A further improvement of the present invention is that the specific operation steps of predicting and analyzing future data trends through the autoregressive moving average model are as follows:

[0062] Determine the order of the model, that is, the order p of the autoregressive term and the order q of the moving average term;

[0063] Estimate model parameters, including autoregressive coefficients φ1, φ2, …, Φp

[0064] Use the estimated model to make predictions;

[0065] The algorithmic form of the autoregressive moving average model is expressed as:

[0066] X t =φ1X t-1 +φ2X t-2 +...+φ p X t-p +θ1∈ t-1 +θ2∈ t-2 +...+θ q ∈ t-q +∈ t ;

[0067] Where: Xt is the value of the time series at time t; φ1, φ2, …, φp are the autoregressive coefficients; θ1, θ2, …, θq are the moving average coefficients; ∈t is the white noise error term, representing the forecast error at time t; ∈t-1, ∈t-2, …, ∈tq are the past forecast errors; p is the order of the autoregressive term, indicating the number of past values ​​considered by the model; q is the order of the moving average term, indicating the number of past forecast errors considered by the model.

[0068] A further improvement of the present invention is that the specific operation steps of performing fault detection on the statistical analysis results by using the FTA algorithm to obtain the fault diagnosis results are as follows:

[0069] Identify the top event: the failure or accident analyzed;

[0070] Determine the bottom event: the most basic cause that can lead to a failure or accident;

[0071] Construct a fault tree: connect the top event with the bottom event through logic gates;

[0072] Analyze fault trees and calculate the probability of various failure modes;

[0073] A further improvement of the present invention is that the fault diagnosis submodule can be expressed by the following formula through the logic gate of the fault tree in the FTA algorithm:

[0074] OR gate calculation formula:

[0075] P(A)=P(B1)+P(B2)+…+P(B n )-P(B1∩B2)-...-P(B n-1 ∩B n )+…;

[0076] AND gate calculation formula: P(A)=P(B1)×P(B2)×...×P(B n );

[0077] Among them, P(A) is the probability of the top event occurring, P(Bi) is the probability of the i-th bottom event occurring; ∩ represents the logical AND operation, indicating that two events occur at the same time; × represents the multiplication operation, indicating that the probabilities of the events occurring are multiplied.

[0078] Compared with the prior art, the present invention has the following beneficial effects:

[0079] The present invention provides a transient temperature sensor processing device and a working method thereof. Specifically, a data acquisition module acquires signals of a transient temperature sensor in real time, a signal processing module processes the acquired signals, and the processed digital signals flow into a data analysis module. The data analysis module extracts features of the digital signals, performs statistical analysis on the feature data set, identifies different temperature patterns according to the statistical analysis results, and performs predictive analysis. At the same time, a diagnosis is performed according to the statistical analysis results to obtain a fault diagnosis result. The statistical analysis results, pattern recognition results, predictive analysis results, and fault diagnosis results are input into a result output submodule to generate a data analysis result. The data analysis results are stored in a data storage module on the one hand, and presented to a user in a visual form through a display and output module on the other hand. At the same time, the early warning and alarm module monitors the signal transmission situation, and immediately issues an alarm if an abnormal situation is found. These modules are interconnected to form a coherent data processing and communication network, which improves the efficiency of the data transmission process.

[0080] Furthermore, the data analysis results are presented to users in a visual form through the display and output module, so that users can obtain data information in a timely manner. At the same time, problems in the data transmission process can be discovered in a timely manner based on the alarm information. Users can take corresponding strategies in a timely manner based on the alarm information, thereby improving the efficiency of problem handling.

[0081] Furthermore, the existence of the data analysis module significantly improves the automation and intelligence level of the system, enabling the system to not only passively collect and store data, but also actively analyze and utilize data. This intelligent data analysis capability enables the system to automatically identify abnormal conditions, predict potential problems and issue warnings without human intervention, thereby achieving real-time monitoring and rapid response to temperature changes. This not only improves the reliability and safety of the system, but also reduces operation and maintenance costs. The value of the data analysis module is particularly prominent, especially in high-risk or unattended environments. In addition, through in-depth analysis of historical data, the data analysis module can also provide data support for system improvement and optimization, and promote the continuous development and upgrading of the system. Therefore, the data analysis module is a key component for the entire system to achieve efficient and intelligent operation. BRIEF DESCRIPTION OF THE DRAWINGS

[0082] Figure 1 It is a schematic diagram of the overall structure of the device of the present invention;

[0083] Figure 2 is a system block diagram of the present invention;

[0084] Figure 3 This is a system block diagram of the data analysis module in the present invention.

[0085] Explanation of the reference numerals in the figure: 1. transient temperature sensor; 2. data processing unit; 3. device outer shell. DETAILED DESCRIPTION

[0086] In order to further understand the content of the present invention, the present invention is described in detail below in conjunction with the accompanying drawings and specific embodiments. It should be understood that the embodiments are only for explaining the present invention and are not intended to limit it.

[0087] Embodiment 1:

[0088] like Figure 1 As shown, a transient temperature sensor data processing device comprises a transient temperature sensor 1 , wherein the transient temperature sensor 1 is connected to a data processing unit 2 , and a device outer shell 3 is arranged outside the data processing unit 2 .

[0089] Further, such as Figure 2 As shown, the data processing unit 2 includes a data acquisition module, a signal processing module, a data storage module, an early warning and alarm module, a data analysis module and a display and output module.

[0090] The data acquisition module is the front end of the system, which directly interfaces with the temperature sensor. For example, if we are using a thermocouple temperature sensor, the data acquisition module will include a cold junction compensation circuit to ensure the accuracy of the temperature reading. The A / D converter in this module may be a 16-bit analog-to-digital converter, which can convert analog signals to digital signals at a rate of 1000 times per second, thus providing high-precision, high-frequency digital temperature data to the signal processing module.

[0091] The signal processing module optimizes the digital signal from the data acquisition module. For example, the module may contain a digital low-pass filter to eliminate signal components with frequencies higher than 1kHz caused by environmental noise. In addition, the signal processing module also performs linearization processing to correct nonlinear errors in the sensor output. Through these processes, the signal processing module ensures the quality and reliability of the data, providing a solid foundation for subsequent data analysis.

[0092] The data analysis module can extract valuable information from the pre-processed temperature data and perform in-depth analysis and interpretation.

[0093] The data storage module is responsible for long-term storage of temperature data. For example, the module may use a SQL database to store data, and each data point contains a timestamp, temperature value, and sensor identifier. The database may be deployed on a local server or use a cloud service for storage. By backing up data regularly, the data storage module ensures that data will not be lost even in the event of hardware failure.

[0094] The early warning and alarm module takes action when an abnormal situation is detected. For example, if the temperature exceeds the preset safety range, the module will notify the operator by sending an email, text message or sound alarm. The early warning and alarm module can also integrate with the control system to automatically start the cooling system or shut down the equipment to prevent further damage.

[0095] The display and output module provides a user interface to display temperature data and system status. For example, the module may include an LED display that displays the current temperature reading in real time, and a graphical user interface (GUI) that displays historical data trends in charts and tables. Users can interact with the display interface through a touch screen or keyboard to adjust display settings or query data for a specific time period.

[0096] Further, such as Figure 3 As shown, the data analysis module includes a data preprocessing submodule, a feature extraction submodule, a statistical analysis submodule, a pattern recognition submodule, a prediction analysis submodule, a fault diagnosis submodule and a result output submodule.

[0097] Data preprocessing submodule: used to convert digital signals into a form with zero mean and unit variance to obtain a data set;

[0098] Feature extraction submodule: used to extract features from the data set to obtain a data set with reduced dimension;

[0099] Statistical analysis submodule: used to calculate the mean and standard deviation of the reduced-dimensional data set to obtain statistical analysis results;

[0100] Pattern recognition submodule: used to separate data points of different categories in the data set according to the statistical analysis results to obtain pattern recognition results;

[0101] Prediction analysis submodule: used to perform prediction analysis on future data trends based on statistical analysis results to obtain prediction analysis results;

[0102] Fault diagnosis submodule: used to perform fault detection on the statistical analysis results and obtain fault diagnosis results;

[0103] Result output submodule: input statistical analysis results, pattern recognition results, prediction analysis results and fault diagnosis results to generate data analysis results.

[0104] In this embodiment, the data acquisition module collects the signal of the transient temperature sensor in real time, the signal processing module processes the collected signal, and the processed digital signal flows into the data analysis module, the data analysis module extracts the features of the digital signal, performs statistical analysis on the feature data set, identifies different temperature patterns according to the statistical analysis results and performs predictive analysis, and diagnoses according to the statistical analysis results to obtain fault diagnosis results, and inputs the statistical analysis results, pattern recognition results, predictive analysis results and fault diagnosis results into the result output submodule to generate data analysis results; the data analysis results are stored in the data storage module on the one hand, and presented to the user in a visual form through the display and output module on the other hand. At the same time, the early warning and alarm module monitors the signal transmission situation, and immediately alarms if an abnormal situation is found. These modules are interconnected to form a coherent data processing and communication network, which improves the efficiency of the data transmission process.

[0105] Embodiment 2:

[0106] A method for processing a transient temperature sensor comprises the following steps:

[0107] Collect the signal of transient temperature sensor 1 in real time;

[0108] Performing filtering and linearization processing on the collected signal of the transient temperature sensor 1 to obtain a processed digital signal;

[0109] The processed data signal is converted into a form with zero mean and unit variance to obtain a data set;

[0110] The PCA algorithm is used to extract features from the data set to obtain a data set with reduced dimensions;

[0111] Calculate the mean and standard deviation of the reduced-dimensional data set to describe the central tendency and dispersion of the data points in the data set and obtain statistical analysis results;

[0112] Through the SVM algorithm, the data points of different categories in the data set are separated according to the statistical analysis results, the normal operation mode or abnormal mode in the system is identified, and the pattern recognition results are obtained;

[0113] According to the statistical analysis results, the future data trend is predicted and analyzed through the autoregressive moving average model to obtain the prediction analysis results;

[0114] Perform fault detection on the statistical analysis results through the FTA algorithm to obtain the fault diagnosis results;

[0115] Input statistical analysis results, pattern recognition results, predictive analysis results and fault diagnosis results to generate data analysis results;

[0116] The data analysis results are stored in the data storage module on the one hand, and presented to the user in a visual form through the display and output module on the other hand;

[0117] Monitor abnormal conditions during signal transmission and immediately issue an alarm if any abnormality is found.

[0118] Furthermore, the data preprocessing submodule converts the data set into a form with zero mean and unit variance; the specific steps are as follows:

[0119] S1. Calculate the mean and standard deviation: For each digital signal feature, calculate its mean (Mean) and standard deviation (Std);

[0120] S2. Apply the transformation formula: For each eigenvalue x, use the following formula to transform:

[0121]

[0122] Among them, x std is the transformed value;

[0123] S3. Transform the dataset: Apply the above transformation to each digital signal feature; after Z-score normalization, the mean of each feature is 0 and the standard deviation is 1, and the distribution of the dataset will be similar to the standard normal distribution.

[0124] Furthermore, the feature extraction submodule extracts features from the data set through the PCA algorithm to obtain a data set with reduced dimensions. The specific operation method includes:

[0125] S1: Data standardization: Ensure that the variance of each feature is 1 and the mean is 0.

[0126] S2: Calculate the covariance matrix.

[0127] S3: Calculate the eigenvalues ​​and eigenvectors of the covariance matrix.

[0128] S4: Select the main eigenvectors to form a new feature space.

[0129] S5: Project the original data onto the new feature space.

[0130] Furthermore, the statistical analysis submodule describes the central tendency and dispersion of the data by calculating the mean and standard deviation of the new data set;

[0131] The calculation formula of mean is: in, is the mean, n is the number of data points, x i is the i-th data point;

[0132] The formula for calculating the standard deviation is:

[0133] Where σ is the standard deviation, n is the number of data points, and x i is the ith data point, is the mean.

[0134] Furthermore, the pattern recognition submodule is used for supervised learning of classification and regression analysis through the support vector machine (SVM) algorithm. In the classification problem, the goal of the SVM algorithm is to find an optimal hyperplane that separates data points of different categories and maximizes the margin, that is, the two types of data points are as far away from the hyperplane as possible.

[0135] The operation steps are as follows:

[0136] S1. Select a kernel function to map data into a high-dimensional space.

[0137] S2. Construct an optimization problem to find the hyperplane with the maximum margin.

[0138] S3. Solve the optimization problem and find the support vector.

[0139] S4. Use support vectors to construct decision functions;

[0140] The maximum margin hyperplane formula for the linearly separable case of the support vector machine (SVM) algorithm is:

[0141] The constraints are: i (w·x i +b)≥1;

[0142] Where: i = 1,…,n; w is the weight vector; b is the bias term; x i is the i-th eigenvector; y i is the class label (+1 or -1) of the i-th feature vector; n is the total number of feature vectors;

[0143] The algorithm formula of nonlinear support vector machine using kernel function is:

[0144]

[0145] The constraints are: 0≤α i ≤C;

[0146] where i=1,…,n; α i is the Lagrange multiplier. C is the penalty parameter; K(x i ,x j ) is the kernel function used to calculate the eigenvector x i and x j The inner product between is in the high-dimensional feature space.

[0147] Furthermore, the prediction analysis submodule performs prediction analysis through the autoregressive moving average model. The specific operation steps are as follows:

[0148] S1. Determine the order of the model, that is, the order p of the autoregressive term and the order q of the moving average term.

[0149] S2. Estimate model parameters, including autoregressive coefficients φ1, φ2,…, φp.

[0150] S3. Make predictions using the estimated model.

[0151] The algorithmic form of the autoregressive moving average model is expressed as:

[0152] X t =φ1X t-1 +φ2X t-2 +...+φ p X t-p +θ1∈ t-1 +θ2∈ t-2 +...+θ q ∈ t-q +∈ t ;

[0153] Where: Xt is the value of the time series at time t; φ1, φ2, …, φp are the autoregressive coefficients; θ1, θ2, …, θq are the moving average coefficients; ∈t is the white noise error term, representing the forecast error at time t; ∈t-1, ∈t-2, …, ∈tq are the past forecast errors; p is the order of the autoregressive term, indicating the number of past values ​​considered by the model; q is the order of the moving average term, indicating the number of past forecast errors considered by the model.

[0154] Furthermore, the fault diagnosis submodule uses the fault tree analysis (FTA) algorithm to trace back from the top event (i.e., the undesirable event) to the basic cause of the fault. FTA constructs a logical tree diagram to identify and analyze the cause of the system failure. The specific operation steps are as follows:

[0155] S1. Determine the top event, that is, the fault or accident that needs to be analyzed.

[0156] S2. Determine the basic events, that is, the most basic fault causes that may lead to the top events.

[0157] S3. Build a fault tree and connect the top event with the bottom event through logic gates (AND gate, OR gate, etc.).

[0158] S4. Analyze the fault tree and calculate the probability of various failure modes. ;

[0159] The fault diagnosis submodule can be expressed by the following formula through the logic gate of the fault tree in the fault tree analysis (FTA) algorithm:

[0160] OR gate calculation formula:

[0161] P(A)=P(B1)+P(B2)+…+P(B n )-P(B1∩B2)-...-P(B n-1 ∩B n )+…;

[0162] AND gate calculation formula: P(A)=P(B1)×P(B2)×...×P(B n )

[0163] Where P(A) is the probability of the top event occurring, and P(Bi) is the probability of the i-th bottom event occurring. ∩ represents the logical AND operation, indicating that two events occur at the same time. × represents the multiplication operation, indicating that the probabilities of the events occurring are multiplied.

[0164] Furthermore, the result output submodule converts complex analysis data into easy-to-understand reports, charts, graphs and visualization interfaces so that users can quickly grasp key information. It generates detailed reports, including key performance indicators (KPIs), statistical data summaries, abnormal event lists, fault diagnosis reports, forecast trend charts, etc., and outputs this information in a variety of ways, such as printing, emailing, online viewing or API interface for other system integration. In addition, the result output submodule also provides customized services, allowing users to select specific data views, report formats and content according to their needs to ensure accurate communication and effective use of analysis results.

[0165] In this embodiment, the data analysis module can extract valuable information from the original temperature data through the collaborative work of its multiple submodules, and conduct in-depth analysis and interpretation. The data preprocessing submodule ensures the quality and consistency of the data, and the feature extraction submodule reduces the dimension of the data and improves the efficiency of the analysis. The statistical analysis submodule provides the basic statistical characteristics of the data to help users understand the distribution of the data. The pattern recognition submodule can identify specific patterns in the data, which provides possibilities for fault detection and system optimization. The predictive analysis submodule provides users with forward-looking decision support by predicting future data trends. The fault diagnosis submodule can promptly discover potential problems of the system and give diagnostic results, thereby reducing the risks and losses caused by system failures. Finally, the result output submodule presents the complex data analysis results to the user in an intuitive way, allowing the user to respond and make decisions quickly. The integration of these functions has greatly improved the intelligence level and user experience of the system.

[0166] The feature extraction submodule within the data analysis module extracts the most important features from a large amount of raw data through algorithms such as PCA, thereby reducing the dimension of the data. The advantage of this is that it reduces the computational complexity, speeds up the analysis, and retains the most representative information in the data, which helps to improve the efficiency and accuracy of subsequent analysis. The statistical analysis submodule provides basic descriptive information of the data by performing statistical calculations such as mean and variance on the data. These statistical results help users understand the distribution, trend, and degree of dispersion of the data, provide a quantitative basis for decision-making, and enable users to make more reasonable judgments based on data rather than intuition. The pattern recognition submodule can identify specific patterns or classifications in the data through algorithms such as SVM. This helps the system automatically identify normal operating modes and abnormal modes, thereby achieving automated fault detection and classification, and improving the intelligence level and self-diagnosis capabilities of the system. The predictive analysis submodule predicts future data trends. This forward-looking analysis enables the system to predict potential problems or changes, providing users with time for early warning and planning, thereby reducing the impact of emergencies. The fault diagnosis submodule can trace the root cause of the fault and provide a detailed diagnostic report through algorithms such as FTA. This helps to quickly locate and repair problems, reduces system downtime, and improves system reliability and maintainability.

[0167] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the above embodiments, ordinary technicians in the relevant field should understand that the specific implementation methods of the present invention can still be modified or replaced by equivalents. Any modification or equivalent replacement that does not depart from the spirit and scope of the present invention should be covered within the scope of protection of the claims of the present invention.

Claims

1. A transient temperature sensor data processing device, characterized in that: The device comprises a transient temperature sensor (1), wherein the transient temperature sensor (1) is connected to a data processing unit (2), and a device outer shell (3) is arranged outside the data processing unit (2); The data processing unit (2) comprises: A data acquisition module: connected to the interface of the transient temperature sensor (1) and used for real-time acquisition of the signal of the transient temperature sensor (1); Signal processing module: used for converting the signal collected by the transient temperature sensor (1) into a digital signal; Data analysis module: used to analyze digital signals and obtain data analysis results; Data storage module: used to store data analysis results; Display and output module: used to display data analysis results; Early warning and alarm module: used to monitor abnormal conditions during signal transmission and generate alarms if abnormal conditions are found; Wherein, the data analysis module includes: Data preprocessing submodule: used to convert digital signals into a form with zero mean and unit variance to obtain a data set; Feature extraction submodule: used to extract features from the data set to obtain a data set with reduced dimension; Statistical analysis submodule: used to calculate the mean and standard deviation of the reduced-dimensional data set to obtain statistical analysis results; Pattern recognition submodule: used to separate data points of different categories in the data set according to the statistical analysis results to obtain pattern recognition results; Prediction analysis submodule: used to predict and analyze future data trends based on statistical analysis results to obtain prediction analysis results; Fault diagnosis submodule: used to perform fault detection on the statistical analysis results and obtain fault diagnosis results; Result output submodule: input statistical analysis results, pattern recognition results, prediction analysis results and fault diagnosis results to generate data analysis results.

2. A method for processing a transient temperature sensor, characterized in that: The steps include: Collecting the signal of the transient temperature sensor (1) in real time; Performing filtering and linearization processing on the collected signal of the transient temperature sensor (1) to obtain a processed digital signal; The processed data signal is converted into a form with zero mean and unit variance to obtain a data set; The PCA algorithm is used to extract features from the data set to obtain a data set with reduced dimensions; Calculate the mean and standard deviation of the reduced-dimensional data set to describe the central tendency and dispersion of the data points in the data set and obtain statistical analysis results; Through the SVM algorithm, the data points of different categories in the data set are separated according to the statistical analysis results, the normal operation mode or abnormal mode in the system is identified, and the pattern recognition results are obtained; According to the statistical analysis results, the future data trend is predicted and analyzed through the autoregressive moving average model to obtain the prediction analysis results; Perform fault detection on the statistical analysis results through the FTA algorithm to obtain the fault diagnosis results; Input statistical analysis results, pattern recognition results, predictive analysis results and fault diagnosis results to generate data analysis results; The data analysis results are stored in the data storage module on the one hand, and presented to the user in a visual form through the display and output module on the other hand; Monitor abnormal conditions during signal transmission and immediately issue an alarm if any abnormality is found.

3. The working method of a transient temperature sensor processing device according to claim 2, characterized in that: The process of converting the processed data signal into a form with zero mean and unit variance is as follows: For each digital signal feature, calculate its mean Mean and standard deviation Std; For each eigenvalue x, the transformation is performed using the following formula: Among them, x std is the transformed value; After the above transformation, the mean of each digital signal feature is 0, the standard deviation is 1, and the distribution of the data set will be similar to the standard normal distribution.

4. The working method of a transient temperature sensor processing device according to claim 2, characterized in that: The working method of extracting features from a data set through the PCA algorithm to obtain a data set with reduced dimensionality is as follows: Data standardization: ensure that the variance of each digital signal feature is 1 and the mean is 0; Calculate the covariance matrix; Compute the eigenvalues ​​and eigenvectors of the covariance matrix; Select the main eigenvectors to form a new feature space; Project the original data onto the new feature space.

5. The working method of a transient temperature sensor processing device according to claim 2, characterized in that: The formula for calculating the mean and standard deviation of the reduced-dimensional dataset is as follows: The formula for calculating the mean is: in, is the mean, n is the number of data points, x i is the i-th data point; The formula for calculating the standard deviation is: Where σ is the standard deviation, n is the number of data points, and x i is the ith data point, is the mean.

6. The working method of a transient temperature sensor processing device according to claim 2, characterized in that: Through the SVM algorithm, the data points of different categories in the data set are separated according to the statistical analysis results, and the normal operation mode or abnormal mode in the system is identified. The operation steps to obtain the pattern recognition results are as follows: Use kernel functions to electrically map the data in the dataset into a high-dimensional space; Construct an optimization problem to obtain the hyperplane with the farthest distance between data points; Solve the optimization problem and obtain the support vector; Use support vectors to build decision functions.

7. The working method of a transient temperature sensor processing device according to claim 6, characterized in that: The formula for the hyperplane with the farthest distance between data points obtained in the case of linear separability of the SVM algorithm is: The constraints are: i (w·x i +b)≥1; Where: i = 1,…,n; w is the weight vector; b is the bias term; x i is the i-th eigenvector; y i is the category label of the i-th feature vector; n is the total number of feature vectors; The nonlinear SVM algorithm formula using kernel function is: The constraints are: 0≤α i ≤C; where i=1,…,n;α i is the Lagrange multiplier; C is the penalty parameter; K(x i ,x j ) is the kernel function used to calculate the eigenvector x i and x j The inner product between is in the high-dimensional feature space.

8. The working method of a transient temperature sensor processing device according to claim 2, characterized in that: The specific operation steps of predicting and analyzing future data trends through the autoregressive moving average model are as follows: Determine the order of the model, that is, the order p of the autoregressive term and the order q of the moving average term; Estimate model parameters, including autoregressive coefficients φ1, φ2, …, φp Use the estimated model to make predictions; The algorithmic form of the autoregressive moving average model is expressed as: X t =φ1X t-1 +φ2X t-2 +...+φ p X t-p +θ1∈ t-1 +θ2∈ t-2 +...+θ q ∈ t-q +∈ t ; Where: Xt is the value of the time series at time t; φ1, φ2, …, φp are autoregressive coefficients; θ1, θ2, …, θq are moving average coefficients; ∈t is the white noise error term, representing the prediction error at time t; ∈t-1, ∈t-2, …, ∈tq are past prediction errors; p is the order of the autoregressive term, indicating the number of past values ​​considered by the model; q is the order of the moving average term, indicating the number of past prediction errors considered by the model.

9. The working method of a transient temperature sensor processing device according to claim 2, characterized in that: The specific operation steps of performing fault detection on the statistical analysis results by using the FTA algorithm to obtain the fault diagnosis results are as follows: Identify the top event: the failure or accident analyzed; Determine the bottom event: the most basic cause that can lead to a failure or accident; Construct a fault tree: connect the top event with the bottom event through logic gates; Analyze fault trees and calculate the probability of various failure modes.

10. The working method of the transient temperature sensor processing device according to claim 9, characterized in that: The fault diagnosis submodule can be expressed by the following formula through the logic gate of the fault tree in the FTA algorithm: OR gate calculation formula: P(A)=P(B1)+P(B2)+...+P(B n )-P(B1∩B2)-...-P(B n-1 ∩B n )+...; AND gate calculation formula: P(A)=P(B1)×P(B2)×...×P(B n ); Among them, P(A) is the probability of the top event occurring, P(Bi) is the probability of the i-th bottom event occurring; ∩ represents the logical AND operation, indicating that two events occur at the same time; × represents the multiplication operation, indicating that the probabilities of the events occurring are multiplied.

Images