Remote printer monitoring and fault prediction system based on Internet of Things

Abstract

The invention discloses a remote printer monitoring and fault prediction system based on the Internet of Things, and relates to the technical field of the Internet of Things and artificial intelligence, and the system comprises a data collection module which is used for collecting equipment state data of a remote printer in real time and carrying out the preprocessing of the equipment state data; the data processing module is used for integrating the preprocessed equipment state data into a data matrix and a label vector; the model construction module is used for outputting a final feature set by using a random forest model and combining a Gumbel-Sigmoid activation function; and the fault prediction module is used for constructing a fault prediction model by using the bidirectional LSTM to output the predicted fault probability of the current time point and dividing a fault interval data pool. By introducing the combination of a random forest algorithm and a Gumbel-Sigmoid activation function, a feature selection process can be automatically optimized, a feature set which is crucial to fault prediction can be accurately screened out, and a bidirectional LSTM model is adopted to capture time sequence information, so that the fault prediction precision and the real-time response capability are remarkably improved.

Description

Technical Field

[0001] The present invention relates to the technical field of Internet of Things and artificial intelligence, in particular to a remote printing press monitoring and fault prediction system based on Internet of Things. Background Art

[0002] With the rapid development of Internet of Things technology, remote monitoring and intelligent management of industrial equipment have become an important part of modern manufacturing. Especially in large-scale distributed environments, remote monitoring technology realizes real-time control of equipment operation status through data collection, processing and analysis, thereby improving system operation efficiency and reducing operation and maintenance costs. As a typical representative of industrial production equipment, the reliability and efficiency of remote printing machines are crucial to printing quality and production continuity. In recent years, the combination of Internet of Things and big data analysis technology has enabled the status monitoring and fault prediction of remote printing machines to gradually develop from passive maintenance to predictive maintenance. By deploying sensors to obtain the equipment operation status and combining machine learning algorithms to process and analyze data, enterprises can predict potential failures in advance and avoid production interruptions.

[0003] At present, most technical solutions for remote printing machine monitoring rely on the collection of basic sensor data and a single rule analysis method. These solutions usually transmit the real-time collected equipment status data directly to the central processing system, which judges the equipment operation status through preset thresholds or fixed rules. This traditional method is prone to misjudgment in the face of complex operation scenarios and it is difficult to accurately capture the potential trend of equipment failure. Traditional methods lack the ability to deeply mine high-dimensional feature data, and it is difficult to achieve global failure risk prediction by relying on a single indicator. Summary of the invention

[0004] In view of the problems existing in the above-mentioned existing remote printing machine monitoring and fault prediction system based on the Internet of Things, the present invention is proposed.

[0005] Therefore, the problem to be solved by the present invention is that the traditional monitoring system lacks flexibility in data processing, and can usually only perform simple preprocessing on the raw data, and fails to effectively integrate multi-dimensional equipment status information, which limits the model's understanding of complex data correlation characteristics.

[0006] In order to solve the above technical problems, the present invention provides the following technical solutions: a remote printing machine monitoring and fault prediction system based on the Internet of Things, which includes:

[0007] Data acquisition module, used to collect the equipment status data of the remote printing machine in real time and perform pre-processing;

[0008] A data processing module, used to integrate the pre-processed device status data into a data matrix and a label vector;

[0009] Model building module, which uses the random forest model and combines the Gumbel-Sigmoid activation function to output the final feature set;

[0010] Fault prediction module, which uses bidirectional LSTM to build a fault prediction model to output the predicted fault probability at the current time point and divide the fault interval data pool;

[0011] The classification processing module is used to screen important features and perform classification processing for data pools in different risk intervals through dynamic causal relationship analysis;

[0012] The data storage module is used to store and manage all data in a relational database.

[0013] As a preferred solution of the remote printer monitoring and fault prediction system based on the Internet of Things of the present invention, the real-time collection of the device status data of the remote printer includes:

[0014] Determine the remote printing press that needs to be monitored and predicted, install a collection device in the remote printing press, set the sampling frequency of the collection device, and collect the equipment status data of the printing press in real time;

[0015] The device status data includes the paper status, vibration signal, ink remaining percentage, device temperature, network delay and number of tasks to be printed of the remote printer at each time point.

[0016] As a preferred solution of the remote printing machine monitoring and fault prediction system based on the Internet of Things of the present invention, the preprocessing includes:

[0017] Check the collected equipment status data and process missing values, detect and eliminate outliers, remove duplicate records, and standardize the equipment status data.

[0018] As a preferred solution of the remote printing machine monitoring and fault prediction system based on the Internet of Things of the present invention, wherein: the integration of the pre-processed device status data into a data matrix and a label vector includes:

[0019] The preprocessed device status data is sorted by timestamp and integrated into time series data. The vibration signal, ink remaining percentage, device temperature, network delay and number of tasks to be printed in the time series data are extracted as features and integrated into the feature vector corresponding to each time point. The feature vectors of all time points are divided into sliding windows with a window size of L. , the step length is S;

[0020] Count the number of occurrences of each paper state at each time point in the sliding window, and calculate the proportion of normal states in the sliding window as the paper state feature;

[0021] Calculate the mean of each feature dimension of the sliding window and the paper status feature as time series statistical features and combine them into a time series feature vector;

[0022] The dynamic detection range of each feature dimension is defined using the mean and standard deviation. All features in each sliding window are analyzed in relation to the dynamic detection range of the corresponding feature dimension. If any feature in the sliding window is larger than the dynamic detection range, the classification label of the sliding window is marked as a fault window. Otherwise, the classification label of the sliding window is marked as a normal window.

[0023] Construct a label vector Y, where each element corresponds to the classification label of a sliding window, the fault window is marked as 1, and the normal window is marked as 0;

[0024] The time series feature vectors of all sliding windows are combined into a global feature matrix, and the global feature matrix and label vector are combined into a data matrix. Each row represents a sliding window, and each column represents the time series feature vector and classification label of the sliding window.

[0025] As a preferred solution of the remote printing machine monitoring and fault prediction system based on the Internet of Things of the present invention, wherein: the final feature set outputted by using the random forest model in combination with the Gumbel-Sigmoid activation function includes:

[0026] Divide the data matrix into training set and validation set, and initialize the parameters of the random forest model;

[0027] The parameters include the number of decision trees, the maximum tree depth, the number of features at each split, and the minimum number of sample splits;

[0028] Use the training set to fit the model, calculate the average Gini index drop value of all features in the training set in all trees, and calculate the importance value of each feature based on the average Gini index drop value ;

[0029] The importance value of each feature Normalize, set the importance threshold, traverse the training set and remove features whose importance values ​​are lower than the importance threshold, sort the features whose importance values ​​are greater than or equal to the importance threshold from high to low according to the importance values, and generate a candidate feature set, which includes the candidate features and the classification labels of the time points corresponding to the candidate features;

[0030] Retrain a new random forest model using the candidate feature set, evaluate the model's performance on the validation set, calculate the AUC value, and define the AUC threshold based on the actual scenario. If the AUC value of the validation set is greater than the AUC threshold, the classification ability of the current candidate feature set is considered to be up to standard and the candidate feature set is output. Otherwise, lower the importance threshold and regenerate the candidate feature set.

[0031] Initialize the initial selection probability, temperature parameter, maximum number of iterations and number of unselected records;

[0032] According to the importance value of the random forest feature, define the initial global weight of each candidate feature in the output candidate feature set and the initial local weights , the expression is:

[0033] , ,

[0034] in, represents the importance value of the i-th candidate feature in the candidate feature set, Represents the maximum importance value among all candidate features, represents the Pearson correlation coefficient between the i-th candidate feature in the candidate feature set and the classification label, Represents the maximum value of the Pearson correlation coefficient between all candidate features and the corresponding classification labels;

[0035] For each candidate feature, generate a random variable, and based on the random variable, calculate the Gumbel noise of each candidate feature ;

[0036] Based on the initial global weight , initial local weight and Gumbel noise , using the Gumbel-Sigmoid activation function to generate the selection probability of each candidate feature in each iteration , the formula is:

[0037] ,

[0038] in, represents the global weight of the i-th candidate feature in the h-1th iteration, represents the local weight of the i-th candidate feature in the h-1th iteration, represents the temperature parameter in the h-1th iteration, controlling the distribution width of the probability;

[0039] According to the selection probability of each candidate feature in each round of iteration , set the probability threshold , judge the selection result of candidate features, if , then select the candidate feature if , then the candidate feature is not selected;

[0040] In each iteration, the global and local weights of the candidate features are dynamically adjusted, and the temperature parameter is gradually reduced through the temperature attenuation factor. ;

[0041] The dynamic adjustment of the global weight and local weight of the candidate features refers to using the updated candidate feature set to train the random forest model in each iteration, and calculating the difference between the AUC value of the candidate feature set in the previous round and the AUC value of the updated candidate feature set. , adjust the global weight of the selected candidate features by difference;

[0042] like , indicating that dynamic feature selection improves performance, then the global weight of the selected candidate features is increased, and the formula is:

[0043] ,

[0044] in, represents the global learning rate;

[0045] like , indicating that the dynamic selection feature does not improve the performance, then the global weight of the selected candidate feature is reduced. The formula is:

[0046] ,

[0047] in, represents the global learning rate;

[0048] The local weights of the unselected candidate features are adjusted by introducing a decay factor, which is decayed according to the number of times they are not selected. The formula is:

[0049]

[0050] in, represents the decay rate control parameter, Indicates the cumulative number of unselected times from the 1st to the hth round of iterations;

[0051] After dynamically adjusting the global and local weights of the candidate features, the next round of dynamic selection is carried out. When the maximum number of iterations is reached, the iteration is terminated, the final selected features are output, and they are sorted in chronological order and integrated into the final feature set.

[0052] As a preferred solution of the remote printing machine monitoring and fault prediction system based on the Internet of Things of the present invention, wherein: the use of bidirectional LSTM to construct a fault prediction model to output the predicted fault probability at the current time point includes:

[0053] The final feature set is divided into a training set and a test set;

[0054] A fault prediction model is constructed using bidirectional LSTM. The feature input at each time point is fed into the forward and backward LSTM sub-networks respectively. The hidden states of the forward and backward sub-networks are concatenated to form a complete hidden state.

[0055] Design the model structure of the fault prediction model and use the binary cross entropy loss function to measure the error between the predicted fault probability output by the model and the classification label;

[0056] The Adam optimizer is used for weight update, the initial learning rate is set to 0.001, and the learning rate is dynamically adjusted according to the validation set loss;

[0057] Initialize the model parameters and input the training set for training. At each time point, the model calculates a probability of stopping prediction, maps through a fully connected layer and uses the Sigmoid activation function to set a stopping threshold. If the probability of stopping prediction is greater than the stopping threshold, it means that the current prediction result meets the standard and the model stops calculating in advance. Otherwise, continue training.

[0058] By inputting real-time equipment status data into the fault prediction model, the predicted fault probability at the corresponding time point is output .

[0059] As a preferred solution of the remote printing machine monitoring and fault prediction system based on the Internet of Things of the present invention, the fault interval data pool includes:

[0060] According to the predicted failure probability, set Divide the failure probability interval;

[0061] like , it means that the current remote printer is in normal state with no signs of failure, and is judged to be in a low-risk range;

[0062] like , it means that the current remote printer has a potential fault and is judged to be in the medium risk range;

[0063] like , it means that the current remote printer is highly abnormal and is judged to be in a high-risk range;

[0064] The risk interval corresponding to each predicted failure probability is used as a data pool, which includes the vibration signal, ink remaining percentage, device temperature, network delay and number of tasks to be printed of the device status data at the time point corresponding to each predicted failure probability.

[0065] As a preferred solution of the remote printing machine monitoring and fault prediction system based on the Internet of Things of the present invention, wherein: the data pool for different risk intervals, through dynamic causal relationship analysis to screen important features, includes:

[0066] Taking each data in each data pool as a feature, analyzing the causal weight of each feature in each data pool on the probability of failure, and verifying the influence of the feature on the status of the remote printing machine by dynamically adjusting the feature value;

[0067] In real-time analysis, the changes in predicted failure probability are observed by adjusting the feature values ​​and updating the causal weights;

[0068] The causal weight thresholds for the low-risk interval and the high-risk interval are set respectively. If the causal weight of a feature is greater than the causal weight threshold, the feature is extracted; otherwise, no operation is performed.

[0069] As a preferred solution of the remote printing machine monitoring and fault prediction system based on the Internet of Things of the present invention, the classification processing includes:

[0070] Classification is performed based on the features extracted from each interval:

[0071] Low-risk zone: regularly monitor the fluctuation of this feature and issue reminders;

[0072] High-risk interval, based on which fault optimization suggestions are generated.

[0073] As a preferred solution of the remote printing machine monitoring and fault prediction system based on the Internet of Things of the present invention, wherein: the storing and managing of all data in a relational database includes:

[0074] Choose a relational database to store and manage data and its analysis results, design the database table structure to store different types of data, set up regular backup tasks, back up all data in the database, manage the permissions of database users and encrypt and store data.

[0075] The beneficial effects of the present invention are as follows: by introducing the combination of random forest algorithm and Gumbel-Sigmoid activation function, the feature selection process can be automatically optimized, and the feature set that is crucial to fault prediction can be accurately screened out. The bidirectional LSTM model is used to capture timing information, so that the fault prediction accuracy and real-time response capability are significantly improved. By dividing the risk intervals and implementing feature importance analysis and causal relationship analysis for data pools in different risk intervals, refined management and precise response are achieved. BRIEF DESCRIPTION OF THE DRAWINGS

[0076] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other accompanying drawings can be obtained based on these accompanying drawings without paying creative work.

[0077] Figure 1 This is a schematic diagram of the structure of the remote printing machine monitoring and fault prediction system based on the Internet of Things.

[0078] Figure 2 Schematic diagram of the process of remote printing press monitoring and fault prediction. DETAILED DESCRIPTION

[0079] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the specific embodiments of the present invention are described in detail below in conjunction with the accompanying drawings.

[0080] In the following description, many specific details are set forth to facilitate a full understanding of the present invention, but the present invention may also be implemented in other ways different from those described herein, and those skilled in the art may make similar generalizations without violating the connotation of the present invention. Therefore, the present invention is not limited to the specific embodiments disclosed below.

[0081] Secondly, the term "one embodiment" or "embodiment" as used herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation of the present invention. The term "in one embodiment" that appears in different places in this specification does not necessarily refer to the same embodiment, nor is it a separate or selective embodiment that is mutually exclusive with other embodiments.

[0082] Example 1

[0083] Reference Figure 1 and Figure 2 , which is the first embodiment of the present invention, and provides a remote printing machine monitoring and fault prediction system based on the Internet of Things. The remote printing machine monitoring and fault prediction system based on the Internet of Things includes:

[0084] S1, data acquisition module, used for real-time acquisition of equipment status data of remote printing machines and preprocessing;

[0085] Specifically, real-time collection of remote printer equipment status data includes:

[0086] Determine the remote printing press that needs to be monitored and predicted, install a collection device in the remote printing press, set the sampling frequency of the collection device, and collect the equipment status data of the printing press in real time;

[0087] The device status data includes the paper status, vibration signal, ink remaining percentage, device temperature, network delay and number of tasks to be printed of the remote printer at each time point;

[0088] The paper status includes normal status, paper jam status, paper out status and multiple paper status;

[0089] The normal state means that the paper position and quantity are in accordance with expectations, and this state is marked as 0;

[0090] The paper jam state means that the sensor detects that the paper is stagnant or in an abnormal position, and the state is marked as 1;

[0091] The paper out state means that the paper inlet sensor does not detect paper, and this state is marked as 2;

[0092] The multiple paper state indicates that the paper outlet sensor detects that the paper thickness or reflection signal is abnormal, and this state is marked as 3.

[0093] By collecting the device status data of the remote printing press in real time, including paper status, vibration signals, ink remaining percentage, device temperature, network delay and the number of tasks to be printed, the device operation status and potential risks can be fully reflected. The paper status classification is marked with digital codes, which provides standardized support for quickly identifying and processing physical anomalies in printing tasks. This real-time monitoring method helps to accurately capture key data in device operation, improve problem diagnosis efficiency, reduce downtime caused by equipment failure, and thus improve production continuity and operational reliability.

[0094] Furthermore, preprocessing includes,

[0095] Check the collected equipment status data and process missing values, detect and eliminate outliers, remove duplicate records, and standardize the equipment status data.

[0096] By preprocessing the collected equipment status data, including checking and processing missing values, detecting and eliminating outliers, removing duplicate records, and standardizing, the quality and consistency of the data can be significantly improved. Processing missing values ​​and outliers can avoid deviations caused by invalid data during model training and prediction. Removing duplicate records can reduce the impact of redundant data on computing efficiency, while standardization helps to eliminate dimensional differences between different features, making the data more suitable for input into the fault prediction model, thereby improving the robustness of the model and prediction accuracy.

[0097] S2, a data processing module, used to integrate the pre-processed device status data into a data matrix and a label vector;

[0098] Specifically, the preprocessed device status data is integrated into a data matrix and a label vector including:

[0099] The preprocessed device status data is sorted by timestamp and integrated into time series data. The vibration signal, ink remaining percentage, device temperature, network delay and number of tasks to be printed in the time series data are extracted as features and integrated into the feature vector corresponding to each time point. The feature vectors of all time points are divided into sliding windows with a window size of L. , the step size is S, the values ​​of L and S are determined by cross-validation, and the sliding window The expression is:

[0100] ,

[0101] in, represents the feature vector at time point t;

[0102] Count the number of occurrences of each paper state at each time point in the sliding window, and calculate the proportion of normal states in the sliding window as the paper state feature;

[0103] Calculate the mean of each feature dimension of the sliding window and the paper status feature as time series statistical features and combine them into a time series feature vector;

[0104] Use mean and standard deviation to define the dynamic range of detection for each feature dimension ,in, represents the mean of the feature, represents the standard deviation of the feature, Indicates the sensitivity coefficient that controls the dynamic detection range;

[0105] Analyze all features in each sliding window and the dynamic detection range of the corresponding feature dimension. If any feature in the sliding window is larger than the dynamic detection range, the classification label of the sliding window is marked as a fault window. Otherwise, the classification label of the sliding window is marked as a normal window.

[0106] Construct a label vector Y, where each element corresponds to the classification label of a sliding window, the fault window is marked as 1, and the normal window is marked as 0;

[0107] The time series feature vectors of all sliding windows are combined into a global feature matrix, and the global feature matrix and label vector are combined into a data matrix. Each row represents a sliding window, and each column represents the time series feature vector and classification label of the sliding window.

[0108] The above method integrates and extracts features of equipment status data through sliding window technology, effectively captures the dynamic characteristics of remote printing machine operation, and uses statistical features within the window and dynamic detection range to constrain feature dimensions. It can accurately identify abnormal fluctuations in equipment and improve the sensitivity and robustness of fault detection. It introduces paper status features into time series statistical features and combines the classification labels of fault windows and normal windows to further enhance the comprehensive characterization of equipment operation status. The data matrix finally generated not only retains the dynamic change information of multi-dimensional features, but also provides complete input for model training and fault prediction in an efficient structural form, which greatly improves the accuracy and efficiency of fault identification and prediction.

[0109] S3, model building module, used to output the final feature set using the random forest model combined with the Gumbel-Sigmoid activation function;

[0110] Specifically, the random forest model is used in combination with the Gumbel-Sigmoid activation function to output the final feature set, including:

[0111] Divide the data matrix into training set and validation set, and initialize the parameters of the random forest model;

[0112] The parameters include the number of decision trees, the maximum tree depth, the number of features at each split, and the minimum number of sample splits;

[0113] Use the training set to fit the model, calculate the average Gini index drop value of all features in the training set in all trees, and calculate the importance value of each feature based on the average Gini index drop value , the formula is:

[0114]

[0115] in, represents the number of decision trees in the random forest, Represents the decrease in the Gini index of the jth feature in the rth decision tree in the training set;

[0116] The Gini index is an indicator for measuring the purity of a node. The smaller the Gini index, the purer the node (i.e., the samples in the current node are more likely to belong to the same category).

[0117] The importance value of each feature Normalize, set the importance threshold, traverse the training set and remove features whose importance values ​​are lower than the importance threshold, sort the features whose importance values ​​are greater than or equal to the importance threshold from high to low according to the importance values, and generate a candidate feature set, which includes the candidate features and the classification labels of the time points corresponding to the candidate features;

[0118] Retrain a new random forest model using the candidate feature set, evaluate the model's performance on the validation set, calculate the AUC value, and define the AUC threshold based on the actual scenario. If the AUC value of the validation set is greater than the AUC threshold, the classification ability of the current candidate feature set is considered to be up to standard and the candidate feature set is output. Otherwise, lower the importance threshold and regenerate the candidate feature set.

[0119] The AUC is the area under the ROC curve, which ranges from [0, 1]. It indicates the ability of the model to correctly classify positive and negative samples. The closer the AUC is to 1, the stronger the classification ability of the model. When AUC = 0.5, the performance of the model is no different from random guessing.

[0120] Initialize the initial selection probability, temperature parameter, maximum number of iterations and number of unselected records;

[0121] According to the importance value of the random forest feature, define the initial global weight of each candidate feature in the output candidate feature set and the initial local weights , the expression is:

[0122] , ,

[0123] in, represents the importance value of the i-th candidate feature in the candidate feature set, Represents the maximum importance value among all candidate features, represents the Pearson correlation coefficient between the i-th candidate feature in the candidate feature set and the classification label, Represents the maximum value of the Pearson correlation coefficient between all candidate features and the corresponding classification labels;

[0124] The combination of global weights (based on feature importance) and local weights (based on the relevance of features to labels) allows for a comprehensive assessment of the value of features:

[0125] The global weight reflects the overall contribution of the feature in the entire dataset;

[0126] The local weight further quantifies the direct relationship between features and labels through the Pearson correlation coefficient;

[0127] This combined weight mechanism is more robust than relying solely on a single indicator and can more accurately identify key features;

[0128] For each candidate feature, generate a random variable, and based on the random variable, calculate the Gumbel noise of each candidate feature , the formula is:

[0129] ,

[0130] in, A random variable representing the i-th candidate feature;

[0131] The temperature parameter refers to a key parameter in the Gumbel-Sigmoid activation function, which controls the smoothness of the feature selection probability;

[0132] Based on the initial global weight , initial local weight and Gumbel noise , using the Gumbel-Sigmoid activation function to generate the selection probability of each candidate feature in each iteration , the formula is:

[0133] ,

[0134] in, represents the global weight of the i-th candidate feature in the h-1th iteration, represents the local weight of the i-th candidate feature in the h-1th iteration, represents the temperature parameter in the h-1th iteration, controlling the distribution width of the probability;

[0135] After the Gumbel-Sigmoid dynamic selection mechanism is introduced, each round of iteration can be dynamically adjusted according to the current model performance and feature selection probability, avoiding the overly rigid limitations of traditional feature selection methods based on fixed thresholds.

[0136] The Gumbel-Sigmoid activation function can simulate probability distribution, smooth the randomness of feature selection, and ensure the stability of feature selection results;

[0137] According to the selection probability of each candidate feature in each round of iteration , set the probability threshold , judge the selection result of candidate features, if , then select the candidate feature if , then the candidate feature is not selected;

[0138] In each iteration, the global and local weights of the candidate features are dynamically adjusted, and the temperature parameter is gradually reduced through the temperature attenuation factor. ;

[0139] The dynamic adjustment of the global weight and local weight of the candidate features refers to using the updated candidate feature set to train the random forest model in each iteration, and calculating the difference between the AUC value of the candidate feature set in the previous round and the AUC value of the updated candidate feature set. , adjust the global weight of the selected candidate features by difference;

[0140] like , indicating that dynamic feature selection improves performance, then the global weight of the selected candidate features is increased, and the formula is:

[0141] ,

[0142] in, represents the global learning rate;

[0143] like , indicating that the dynamic selection feature does not improve the performance, then the global weight of the selected candidate feature is reduced. The formula is:

[0144] ,

[0145] in, represents the global learning rate;

[0146] The local weights of the unselected candidate features are adjusted by introducing a decay factor, which is decayed according to the number of times they are not selected. The formula is:

[0147]

[0148] in, represents the decay rate control parameter, Indicates the cumulative number of unselected times from the 1st to the hth round of iterations;

[0149] After dynamically adjusting the global and local weights of the candidate features, the next round of dynamic selection is carried out. When the maximum number of iterations is reached, the iteration is terminated, the final selected features are output, and they are sorted in chronological order and integrated into the final feature set.

[0150] By combining the random forest model and the Gumbel-Sigmoid activation function, the present invention realizes the dynamic selection and optimization of candidate features, significantly improving the flexibility and accuracy of feature selection. The random forest model quantifies the importance of features by calculating the average decline value of the Gini index, and globally screens out candidate feature sets that contribute more to fault prediction. In the dynamic selection process, the Gumbel-Sigmoid activation function further optimizes the feature selection results and balances randomness and stability by combining global weights and local weights, dynamic adjustment mechanism and temperature attenuation factor. This method avoids the limitations of traditional fixed threshold feature selection methods, can adaptively adjust feature selection strategies, and optimize the importance ranking of features in real time according to model performance. The final generated feature set is more robust and interpretable, providing high-quality input for subsequent fault prediction models, and effectively improving the prediction ability and applicability of the model.

[0151] S4, fault prediction module, used to use bidirectional LSTM to build a fault prediction model to output the predicted fault probability at the current time point and divide the fault interval data pool;

[0152] Specifically, the bidirectional LSTM is used to build a fault prediction model to output the predicted fault probability at the current time point, including:

[0153] The final feature set is divided into a training set and a test set;

[0154] A fault prediction model is built using bidirectional LSTM. The feature input at each time point is fed into the forward and backward LSTM sub-networks respectively. The forward network processes the data at each time point from the beginning to the end of the time series, and the backward network processes the data in reverse from the last time point of the time series. At each time point, the forward and backward hidden states are concatenated to form a complete hidden state. This structure can capture the bidirectional dependency relationship in the time series.

[0155] The number of hidden units in the LSTM layer is designed to be 60. A fully connected layer is connected after the output of the LSTM layer to perform feature transformation on the hidden state. The output dimension of the fully connected layer is 32, and the activation function is ReLU to ensure that the output value is non-negative and enhance the nonlinear expression ability of the model. The fully connected layer is connected to the output layer of a single node, and the Sigmoid activation function is used to map the output of the fully connected layer to the [0,1] interval to represent the predicted fault probability;

[0156] The binary cross entropy loss function is used to measure the error between the predicted failure probability and the classification label output by the model;

[0157] The Adam optimizer is used for weight update, the initial learning rate is set to 0.001, and the learning rate is dynamically adjusted according to the validation set loss;

[0158] Initialize the model parameters and input the training set for training. At each time point, the model calculates a probability of stopping prediction, maps through a fully connected layer and uses the Sigmoid activation function to set a stopping threshold. If the probability of stopping prediction is greater than the stopping threshold, it means that the current prediction result meets the standard and the model stops calculating in advance. Otherwise, continue training.

[0159] By inputting real-time equipment status data into the fault prediction model, the predicted fault probability at the corresponding time point is output .

[0160] The bidirectional LSTM fault prediction model utilizes the bidirectional time series processing capability, combines the hidden states of the forward and backward LSTM networks, captures the bidirectional dependencies in the data, and thus comprehensively mines the time series characteristics of the equipment operation status. By designing the number of hidden units, fully connected layers, and activation functions, the model has strong nonlinear expression capabilities and accurate fault probability prediction capabilities. By using the binary cross entropy loss function and the Adam optimizer with dynamic learning rate adjustment, the model can converge quickly and adapt to different data distributions, optimizing the prediction accuracy. By introducing the stop prediction mechanism, the model can dynamically determine whether to terminate the calculation in advance based on the current prediction results, significantly improving the computing efficiency. This design not only improves the accuracy and real-time performance of the prediction, but also effectively saves computing resources, providing efficient and reliable technical support for intelligent monitoring and fault warning.

[0161] Furthermore, the fault interval data pool is divided into:

[0162] According to the predicted failure probability, set Divide the failure probability interval;

[0163] like , it means that the current remote printer is in normal state with no signs of failure, and is judged to be in a low-risk range;

[0164] like , it means that the current remote printer has a potential fault and is judged to be in the medium risk range;

[0165] like , it means that the current remote printer is highly abnormal and is judged to be in a high-risk range;

[0166] The risk interval corresponding to each predicted failure probability is used as a data pool, which includes the vibration signal, ink remaining percentage, device temperature, network delay and number of tasks to be printed of the device status data at the time point corresponding to each predicted failure probability.

[0167] By dividing the predicted failure probability into three intervals of low risk, medium risk and high risk, and classifying and storing the equipment status data at relevant time points in different data pools according to the risk intervals corresponding to the failure probability, refined management of the equipment status can be achieved. This division method not only helps to accurately distinguish between normal operation, potential failures and highly abnormal states, but also provides targeted data support for subsequent feature importance analysis and dynamic causal relationship analysis, thereby significantly improving the efficiency and accuracy of fault prediction, and providing a scientific basis for risk intervention and optimization strategies.

[0168] S5, classification processing module, is used to screen important features and perform classification processing for data pools in different risk intervals through dynamic causal relationship analysis;

[0169] Specifically, for data pools in different risk ranges, dynamic causal relationship analysis is used to screen important features, including:

[0170] Each data in each data pool is used as a feature, and the causal weight of each feature in each data pool on the failure probability is analyzed. The influence of the feature on the status of the remote printing machine is verified by dynamically adjusting the feature value. The formula is:

[0171]

[0172] in, represents the causal weight of the feature on the prediction result, It indicates the change in predicted failure probability after the extracted feature value is adjusted. Indicates the magnitude of eigenvalue adjustment;

[0173] In real-time analysis, the changes in predicted failure probability are observed by adjusting the feature values ​​and updating the causal weights;

[0174] The causal weight thresholds for the low-risk interval and the high-risk interval are set respectively. If the causal weight of a feature is greater than the causal weight threshold, the feature is extracted; otherwise, no operation is performed.

[0175] Through dynamic causal analysis, the actual impact of features in different risk intervals on the status of remote printing machines can be accurately evaluated. By gradually adjusting the feature values, the changes in the predicted failure probability can be observed in real time, and the causal weights can be dynamically updated to quantify the contribution of each feature to the failure risk. This method effectively solves the problems of static feature importance and limitations in traditional analysis, and realizes dynamic correlation verification between features and fault prediction results. It can provide a scientific basis for precise optimization and fault prevention of equipment, and help improve the stability and reliability of remote printing machines.

[0176] Furthermore, the classification process includes,

[0177] Classification is performed based on the features extracted from each interval:

[0178] Low-risk zone: regularly monitor the fluctuation of this feature and issue reminders;

[0179] High-risk interval, based on which fault optimization suggestions are generated.

[0180] By classifying remote printing machines through interval causal analysis, accurate status management and fault prevention can be achieved. Feature analysis of low-risk intervals helps to identify key factors for maintaining stable equipment operation, while in-depth analysis of high-risk intervals focuses on key fault characteristics and generates targeted optimization solutions, thereby effectively reducing the probability of failure and extending equipment life. This classification and processing method significantly improves the reliability and intelligence level of equipment operation, while optimizing operation and maintenance efficiency.

[0181] S6, data storage module, used to store all data in a relational database and manage it;

[0182] All data is stored in a relational database and managed, including:

[0183] Select a relational database to store and manage data and its analysis results, design the database table structure to store different types of data, set up regular backup tasks, back up all data in the database, manage the permissions of database users and encrypt and store data.

[0184] Adopt relational database to store and manage data and its analysis results. By designing a reasonable database table structure, it can efficiently organize and store different types of data, set regular backup tasks to ensure data security and recoverability, implement permission management for database users, restrict unauthorized access, and ensure data security. At the same time, by encrypting and storing sensitive information, it can effectively prevent data leakage and tampering, thereby improving the security, reliability and availability of data management, and providing solid support for subsequent analysis and decision-making.

[0185] Example 2

[0186] The second embodiment of the present invention provides experimental steps of a remote printing press monitoring and fault prediction system based on the Internet of Things, including:

[0187] The experimental steps include:

[0188] Data collection module, data processing module, model building module and classification processing module;

[0189] First, in the data acquisition module, by setting the sampling frequency, the real-time device status data of the remote printer is collected, as shown in Table 1, the collected device status data is checked and missing values ​​are processed, abnormal values ​​are detected and eliminated, duplicate records are removed, and the device status data is standardized;

[0190] Set the parameters of the sliding window, the window size is 5, and the window step is 1;

[0191] Table 1 Equipment status data for each time window

[0192] Timestamp Vibration signal Ink remaining (%) Equipment temperature Network latency Number of tasks to be printed 2023-11-20 10:00:00 0.15 80 45 50 2 2023-11-20 10:00:01 0.16 79 45.2 48 2 2023-11-20 10:00:02 0.30 78 46 60 3 2023-11-20 10:00:03 0.12 77 47 55 1 2023-11-20 10:00:04 0.40 76 48.5 70 2

[0193] Second, in the data processing module, the statistical features of vibration signals, ink remaining, device temperature, network delay and number of tasks to be printed are calculated, and the number of occurrences of each paper state in the sliding window is counted, as shown in Table 2. The proportion of the normal state is calculated as the paper state feature, and the paper state feature is obtained as follows: , summarize the statistical features and paper status features;

[0194] Table 2 Collected paper status data

[0195] Timestamp Paper Status 2023-11-20 10:00:00 0 2023-11-20 10:00:01 0 2023-11-20 10:00:02 3 2023-11-20 10:00:03 2 2023-11-20 10:00:04 0

[0196] Analyze the features of each sliding window. For each feature dimension in each sliding window, check the feature value according to the dynamic detection range, and conclude that the sliding window is a fault window with a classification label of 1. Combine the time series feature vectors of all sliding windows into a global feature matrix, and combine the global feature matrix and the label vector into a data matrix.

[0197] 3. In the model building module, the random forest model is used in combination with the Gumbel-Sigmoid activation function to iterate the features of the obtained data matrix;

[0198] The data matrix is ​​divided into training sets, and the model is fitted using the training sets, as shown in Table 3;

[0199] Table 3 Training set

[0200] Sliding window sequence number Feature 1 Feature 2 Feature 3: ... <![CDATA[O1]]> 0.90 40.0 60.0 ...

[0201] Calculate the average Gini index drop value of all features in the training set in all trees, calculate the importance value MDG of each feature based on the average Gini index drop value, normalize the importance values ​​of all features, and calculate the Pearson correlation coefficient between the feature and the classification label , as shown in Table 4:

[0202] Table 4 Pearson correlation coefficients between MDG and features and classification labels

[0203] #timg# (normalization) C #timg# #timg# Feature 1 0.08 0.60 0.25 0.9 Feature 2 0.06 0.50 0.25 0.9 Feature 3 0.15 0.8 0.25 0.9

[0204] Calculate the initial global weight and local weight, the formula is:

[0205] , ,

[0206] Feature 1:

[0207] , 0.67,

[0208] Feature 2:

[0209] , 0.56,

[0210] Feature 3:

[0211] , 0.89,

[0212] Set the temperature parameter to 0.1, Gumbel noise of feature 1, feature 2 and feature 3 They are -0.05, 0.01 and 0.01 respectively. The feature selection probability is calculated based on the global weight and local weight. The formula is:

[0213] ,

[0214] Feature 1:

[0215] ,

[0216] Feature 2:

[0217] ,

[0218] Feature 3:

[0219] ,

[0220] According to the selection probability of each candidate feature in each round of iteration , set the probability threshold , judge the selection result of candidate features, if , then select the candidate feature if , then the candidate feature is not selected;

[0221] The dynamic adjustment of the global weight and local weight of the candidate features refers to using the updated candidate feature set to train the random forest model in each iteration, and calculating the difference between the AUC value of the candidate feature set in the previous round and the AUC value of the updated candidate feature set. , adjust the global weight of the selected candidate features by difference;

[0222] Set the characteristics of feature 1, feature 2 and feature 3 The learning rates are 0.01, 0.02, and -0.03 respectively. is 0.01;

[0223] like , indicating that dynamic feature selection improves performance, then the global weight of the selected candidate features is increased, and the formula is:

[0224] ,

[0225] in, represents the global learning rate;

[0226] like , indicating that the dynamic selection feature does not improve the performance, then the global weight of the selected candidate feature is reduced. The formula is:

[0227] ,

[0228] Feature 1:

[0229] 0.3201,

[0230] Feature 2:

[0231] ,

[0232] Feature 3:

[0233] ,

[0234] By introducing a decay factor to adjust the local weights of unselected candidate features, the decay is performed according to the number of unselected times. After dynamically adjusting the global and local weights of the candidate features, the next round of dynamic selection is carried out. When the maximum number of iterations is reached, the iteration is terminated, the final selected features are output, and they are sorted in chronological order and integrated into the final feature set.

[0235] 4. In the classification processing module, the real-time equipment status data is input into the fault prediction model to output the predicted fault probability at the corresponding time point. ,set up The failure probability at each time point is mapped to the normal interval, low risk interval, and high risk interval, as shown in Table 4;

[0236] Table 4 Real-time equipment status data and risk interval division

[0237] Time Vibration signal Ink level percentage temperature Network latency Number of tasks Predicting the probability of failure Risk range <![CDATA[t1]]> 0.12 80% 45℃ 20 2 0.10 Normal range <![CDATA[t2]]> 0.15 75% 46 25 3 0.20 Normal range <![CDATA[t3]]> 0.35 60% 50 40 4 0.50 Low risk range <![CDATA[t4]]> 0.70 30% 60 70 5 0.85 High risk area <![CDATA[t5]]> 1.00 20% 65 100 6 0.95 High risk area

[0238] 4. Experimental Results Analysis

[0239] 1. Through the failure probability output by the fault prediction model, the experiment realizes risk division. The experimental data shows that the present invention can accurately divide the risk intervals by the failure probability. The division logic of each interval is clear, and the characteristic distribution is highly consistent with the risk level.

[0240] 2. From the variation range of features in different risk intervals, it can be seen that the distribution of each feature in the risk interval is significantly different, which verifies the rationality of feature selection. The present invention can effectively capture key features related to faults through dynamic feature extraction and sensitive feature analysis;

[0241] 3. Compared with traditional feature selection technology, this invention realizes dynamic feature optimization by introducing Gumbel-Sigmoid activation function, which significantly improves the prediction accuracy, recall rate and precision of the model;

[0242] If the functions are implemented in the form of software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, or the part of the technical solution, can be embodied in the form of a software product. The computer software product is stored in a storage medium, including several instructions to enable a computer device (which can be a personal computer, server, or network device, etc.) to perform all or part of the steps of the methods described in various embodiments of the present invention. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), disk or optical disk, etc., which can store program codes.

[0243] The logic and / or steps represented in the flowchart or otherwise described herein, for example, can be considered as an ordered list of executable instructions for implementing logical functions, and can be embodied in any computer-readable medium for use by an instruction execution system, device or apparatus (such as a computer-based system, a system including a processor, or other system that can fetch instructions from an instruction execution system, device or apparatus and execute instructions), or in conjunction with such instruction execution systems, devices or apparatuses. For the purposes of this specification, "computer-readable medium" can be any device that can contain, store, communicate, propagate or transmit a program for use by an instruction execution system, device or apparatus, or in conjunction with such instruction execution systems, devices or apparatuses.

[0244] More specific examples of computer-readable media (a non-exhaustive list) include the following: an electrical connection with one or more wires (electronic device), a portable computer disk case (magnetic device), a random access memory (RAM), a read-only memory (ROM), an erasable and programmable read-only memory (EPROM or flash memory), an optical fiber device, and a portable compact disk read-only memory (CDROM). In addition, the computer-readable medium may even be a paper or other suitable medium on which the program is printed, since the program may be obtained electronically, for example, by optically scanning the paper or other medium, followed by editing, deciphering or, if necessary, processing in another suitable manner, and then stored in a computer memory.

[0245] It should be understood that the various parts of the present invention can be implemented by hardware, software, firmware or a combination thereof. In the above embodiments, multiple steps or methods can be implemented by software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if implemented by hardware, as in another embodiment, it can be implemented by any one of the following technologies known in the art or their combination: a discrete logic circuit with a logic gate circuit for implementing a logic function for a data signal, a dedicated integrated circuit with a suitable combination of logic gate circuits, a programmable gate array (PGA), a field programmable gate array (FPGA), etc.

[0246] 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 preferred embodiments, those skilled in the art should understand that the technical solutions of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical solutions of the present invention, which should all be included in the scope of the claims of the present invention.

Claims

1. A remote printing machine monitoring and fault prediction system based on the Internet of Things, characterized by: include, Data collection module, used to collect the equipment status data of the remote printing machine in real time and perform pre-processing; A data processing module, used to integrate the pre-processed device status data into a data matrix and a label vector; Model building module, which uses the random forest model and combines the Gumbel-Sigmoid activation function to output the final feature set; Fault prediction module, which uses bidirectional LSTM to build a fault prediction model to output the predicted fault probability at the current time point and divide the fault interval data pool; The classification processing module is used to screen important features and perform classification processing for data pools in different risk intervals through dynamic causal relationship analysis; The data storage module is used to store and manage all data in a relational database.

2. The remote printing press monitoring and fault prediction system based on the Internet of Things as claimed in claim 1, characterized in that: The real-time collection of the device status data of the remote printer includes: Determine the remote printing press that needs to be monitored and predicted, install a collection device in the remote printing press, set the sampling frequency of the collection device, and collect the equipment status data of the printing press in real time; The device status data includes the paper status, vibration signal, ink remaining percentage, device temperature, network delay and number of tasks to be printed of the remote printer at each time point.

3. The remote printing machine monitoring and fault prediction system based on the Internet of Things as claimed in claim 2, characterized in that: The pre-processing comprises: Check the collected equipment status data and process missing values, detect and eliminate outliers, remove duplicate records, and standardize the equipment status data.

4. The remote printing machine monitoring and fault prediction system based on the Internet of Things as claimed in claim 3, characterized in that: The integration of the pre-processed device status data into a data matrix and a label vector comprises: The preprocessed device status data is sorted by timestamp and integrated into time series data. The vibration signal, ink remaining percentage, device temperature, network delay and number of tasks to be printed in the time series data are extracted as features and integrated into the feature vector corresponding to each time point. The feature vectors of all time points are divided into sliding windows with a window size of L. , the step length is S; Count the number of occurrences of each paper state at each time point in the sliding window, and calculate the proportion of normal states in the sliding window as the paper state feature; Calculate the mean of each feature dimension of the sliding window and the paper status feature as time series statistical features and combine them into a time series feature vector; The dynamic detection range of each feature dimension is defined using the mean and standard deviation. All features in each sliding window are analyzed in relation to the dynamic detection range of the corresponding feature dimension. If any feature in the sliding window is larger than the dynamic detection range, the classification label of the sliding window is marked as a fault window. Otherwise, the classification label of the sliding window is marked as a normal window. Construct a label vector Y, where each element corresponds to the classification label of a sliding window, the fault window is marked as 1, and the normal window is marked as 0; The time series feature vectors of all sliding windows are combined into a global feature matrix, and the global feature matrix and label vector are combined into a data matrix. Each row represents a sliding window, and each column represents the time series feature vector and classification label of the sliding window.

5. The remote printing press monitoring and fault prediction system based on the Internet of Things as claimed in claim 4, characterized in that: The random forest model is used in combination with the Gumbel-Sigmoid activation function to output the final feature set, including: Divide the data matrix into training set and validation set, and initialize the parameters of the random forest model; The parameters include the number of decision trees, the maximum tree depth, the number of features at each split, and the minimum number of sample splits; Use the training set to fit the model, calculate the average Gini index drop value of all features in the training set in all trees, and calculate the importance value of each feature based on the average Gini index drop value ; The importance value of each feature Normalize, set the importance threshold, traverse the training set and remove features whose importance values ​​are lower than the importance threshold, sort the features whose importance values ​​are greater than or equal to the importance threshold from high to low according to the importance values, and generate a candidate feature set, which includes the candidate features and the classification labels of the time points corresponding to the candidate features; Retrain a new random forest model using the candidate feature set, evaluate the model's performance on the validation set, calculate the AUC value, and define the AUC threshold based on the actual scenario. If the AUC value of the validation set is greater than the AUC threshold, the classification ability of the current candidate feature set is considered to be up to standard and the candidate feature set is output. Otherwise, lower the importance threshold and regenerate the candidate feature set. Initialize the initial selection probability, temperature parameter, maximum number of iterations and number of unselected records; According to the importance value of the feature of the random forest, define the initial global weight of each candidate feature in the output candidate feature set and the initial local weights , the expression is: ， ， in, represents the importance value of the i-th candidate feature in the candidate feature set, Represents the maximum importance value among all candidate features, represents the Pearson correlation coefficient between the i-th candidate feature in the candidate feature set and the classification label, Represents the maximum value of the Pearson correlation coefficient between all candidate features and the corresponding classification labels; For each candidate feature, generate a random variable, and based on the random variable, calculate the Gumbel noise of each candidate feature ; Based on the initial global weight , initial local weight and Gumbel noise , using the Gumbel-Sigmoid activation function to generate the selection probability of each candidate feature in each iteration , the formula is: ， in, represents the global weight of the i-th candidate feature in the h-1th iteration, represents the local weight of the i-th candidate feature in the h-1th iteration, represents the temperature parameter in the h-1th iteration, controlling the distribution width of the probability; According to the selection probability of each candidate feature in each round of iteration , set the probability threshold , judge the selection result of candidate features, if , then select the candidate feature if , then the candidate feature is not selected; In each iteration, the global and local weights of the candidate features are dynamically adjusted, and the temperature parameter is gradually reduced through the temperature attenuation factor. ; The dynamic adjustment of the global weight and local weight of the candidate features refers to using the updated candidate feature set to train the random forest model in each iteration, and calculating the difference between the AUC value of the candidate feature set in the previous round and the AUC value of the updated candidate feature set. , adjust the global weight of the selected candidate features by difference; like , indicating that dynamic feature selection improves performance, then the global weight of the selected candidate features is increased, and the formula is: ， in, represents the global learning rate; like , indicating that the dynamic selection feature does not improve the performance, then the global weight of the selected candidate feature is reduced. The formula is: ， in, represents the global learning rate; The local weights of the unselected candidate features are adjusted by introducing a decay factor, which is decayed according to the number of times they are not selected. The formula is: ； in, represents the decay rate control parameter, Indicates the cumulative number of unselected times from the 1st to the hth round of iterations; After dynamically adjusting the global and local weights of the candidate features, the next round of dynamic selection is carried out. When the maximum number of iterations is reached, the iteration is terminated, the final selected features are output, and they are sorted in chronological order and integrated into the final feature set.

6. The remote printing press monitoring and fault prediction system based on the Internet of Things as claimed in claim 5, characterized in that: The use of bidirectional LSTM to construct a fault prediction model to output the predicted fault probability at the current time point includes: The final feature set is divided into a training set and a test set; A fault prediction model is constructed using bidirectional LSTM. The feature input at each time point is fed into the forward and backward LSTM sub-networks respectively. The hidden states of the forward and backward sub-networks are concatenated to form a complete hidden state. Design the model structure of the fault prediction model and use the binary cross entropy loss function to measure the error between the predicted fault probability output by the model and the classification label; The Adam optimizer is used for weight update, the initial learning rate is set to 0.001, and the learning rate is dynamically adjusted according to the validation set loss; Initialize the model parameters and input the training set for training. At each time point, the model calculates a probability of stopping prediction, maps through a fully connected layer and uses the Sigmoid activation function to set a stopping threshold. If the probability of stopping prediction is greater than the stopping threshold, it means that the current prediction result meets the standard and the model stops calculating in advance. Otherwise, continue training. By inputting real-time equipment status data into the fault prediction model, the predicted fault probability at the corresponding time point is output .

7. The remote printing press monitoring and fault prediction system based on the Internet of Things as claimed in claim 6, characterized in that: The partitioning of the fault interval data pool includes: According to the predicted failure probability, set Divide the failure probability interval; like , it means that the current remote printer is in normal state, with no signs of failure, and is judged to be in the normal range; like , it means that the current remote printer has a potential fault and is judged to be in the low risk range; like , it means that the current remote printer is highly abnormal and is judged to be in a high-risk range; The medium-risk interval and the high-risk interval are respectively used as data pools, which include the vibration signal, ink remaining percentage, device temperature, network delay and number of tasks to be printed of the equipment status data at the time point corresponding to each predicted failure probability.

8. The remote printing machine monitoring and fault prediction system based on the Internet of Things as claimed in claim 7, characterized in that: The data pool for different risk intervals screens important features through dynamic causal analysis, including: Taking each data in each data pool as a feature, analyzing the causal weight of each feature in each data pool on the probability of failure, and verifying the influence of the feature on the status of the remote printing machine by dynamically adjusting the feature value; In real-time analysis, the changes in predicted failure probability are observed by adjusting the feature values ​​and updating the causal weights; The causal weight thresholds for the low-risk interval and the high-risk interval are set respectively. If the causal weight of a feature is greater than the causal weight threshold, the feature is extracted; otherwise, no operation is performed.

9. The remote printing machine monitoring and fault prediction system based on the Internet of Things as claimed in claim 8, characterized in that: The classification process comprises: Classification is performed based on the features extracted from each interval: Low-risk zone: regularly monitor the fluctuation of this feature and issue reminders; High-risk interval, based on which fault optimization suggestions are generated.

10. The remote printing press monitoring and fault prediction system based on the Internet of Things as claimed in claim 9, characterized in that: The storing and managing of all data in a relational database includes: Select a relational database to store and manage data and its analysis results, design the database table structure to store different types of data, set up regular backup tasks, back up all data in the database, manage the permissions of database users and encrypt and store data.

Images