Bucket wheel machine automatic monitoring and alarming system and method based on internet of things

Abstract

The embodiment of the disclosure relates to the technical field of intelligent monitoring of bucket wheel machines, and provides a bucket wheel machine automatic monitoring and alarm system and method based on Internet of Things, wherein time series of bucket wheel machine operation parameters are acquired, time series analysis and mutual fusion of the bucket wheel machine operation parameters are carried out by using data acquisition, processing and analysis algorithms based on Internet of Things and artificial intelligence, so as to automatically judge the operation state of the bucket wheel machine according to the representation features of the various operation parameters of the bucket wheel machine in time series, and timely send early warning prompts according to the operation state of the bucket wheel machine, thereby effectively improving the accuracy of fault detection of the bucket wheel machine, and improving the reliability and working efficiency of the equipment.

Description

TECHNICAL FIELD

[0001] The present disclosure relates to the technical field of intelligent monitoring of bucket wheel machines, and in particular to a bucket wheel machine automatic monitoring and alarm system and method based on the Internet of Things. BACKGROUND

[0002] A bucket wheel machine is a high-efficiency loading and unloading machine used in large dry bulk cargo yards, capable of continuous conveying of stacking and taking material. Bucket wheel machines are widely used in port, wharf, metallurgy, cement and other bulk material storage yards for stacking and taking operations, and are an indispensable important equipment in modern industry.

[0003] The bucket wheel machine is one of the main equipment in coal yards, and its stable operation is crucial to ensure the continuous coal supply of industrial systems such as thermal power plants. The importance of bucket wheel machine fault detection lies in the ability to promptly detect and handle problems that may cause production interruptions, thereby avoiding greater economic losses and safety risks.

[0004] However, traditional bucket wheel machine fault detection methods usually involve periodic inspections or regular maintenance of the bucket wheel machine, which cannot achieve real-time monitoring and early warning. Once the bucket wheel machine fails, it may delay the processing time and affect the normal operation of the equipment. Moreover, when faced with complex faults, the traditional bucket wheel machine fault detection method often fails to accurately diagnose the fault, which requires a lot of time and effort to troubleshoot, resulting in low detection efficiency and accuracy. SUMMARY

[0005] The present disclosure aims to at least solve one of the problems existing in the prior art, and provides a bucket wheel machine automatic monitoring and alarm system and method based on the Internet of Things.

[0006] In one aspect of the present disclosure, a bucket wheel machine automatic monitoring and alarm system based on the Internet of Things is provided, which comprises:

[0007] A bucket wheel machine operating parameter acquisition module for acquiring a time series of bucket wheel machine operating parameters, the bucket wheel machine operating parameters including motor temperature values, motor speed values and hydraulic system pressure values of the bucket wheel machine;

[0008] A data regularization module for data regularization of the time series of bucket wheel machine operating parameters according to time dimension and sample dimension to obtain corresponding motor temperature time series input vectors, motor speed time series input vectors and hydraulic system pressure time series input vectors;

[0009] The operation parameter time sequence mode feature extraction module is configured to perform bucket wheel machine operation parameter time sequence mode feature extraction on the motor temperature time sequence input vector, the motor speed time sequence input vector, and the hydraulic system pressure time sequence input vector respectively to obtain a motor temperature time sequence mode feature vector, a motor speed time sequence mode feature vector, and a hydraulic system pressure time sequence mode feature vector.

[0010] The motor operation state feature fusion module is configured to perform motor operation state capture fusion on the motor temperature time sequence mode feature vector and the motor speed time sequence mode feature vector to obtain a motor temperature-speed time sequence mode fusion representing motor operation state feature vector.

[0011] The operation state representation module is configured to perform dependency scale calculation on the motor temperature-speed time sequence mode fusion representing motor operation state feature vector and the hydraulic system pressure time sequence mode feature vector to obtain a bucket wheel machine operation state representation matrix as a bucket wheel machine operation state representation feature.

[0012] The recognition result generation module is configured to obtain a corresponding recognition result based on the bucket wheel machine operation state representation feature, and determine whether to issue a fault warning prompt based on the recognition result.

[0013] Optionally, the operation parameter time sequence mode feature extraction module is specifically configured to input the motor temperature time sequence input vector, the motor speed time sequence input vector, and the hydraulic system pressure time sequence input vector into a bucket wheel machine operation parameter time sequence mode feature extractor based on a causal convolution network to obtain the motor temperature time sequence mode feature vector, the motor speed time sequence mode feature vector, and the hydraulic system pressure time sequence mode feature vector.

[0014] Optionally, the motor operation state feature fusion module is specifically configured to input the motor temperature time sequence mode feature vector and the motor speed time sequence mode feature vector into a motor operation state capturer based on a motor operation parameter time sequence fusion module to obtain the motor temperature-speed time sequence mode fusion representing motor operation state feature vector.

[0015] Optionally, the motor operation state feature fusion module includes:

[0016] The representation feature construction unit is configured to construct a corresponding motor temperature time sequence mode representation feature vector and a motor speed time sequence mode representation feature vector based on the motor temperature time sequence mode feature vector and the motor speed time sequence mode feature vector respectively.

[0017] The motor parameter feature interaction unit is configured to add the motor temperature time sequence mode feature vector and the motor speed time sequence mode feature vector position by position to obtain a motor temperature-speed time sequence mode fusion feature vector of a motor operating state.

[0018] Optionally, the feature construction unit is specifically configured to:

[0019] The motor temperature time sequence mode feature vector is point convolution coded to obtain a convolution motor temperature time sequence mode feature vector, and the convolution motor temperature time sequence mode feature vector is multiplied by a first hyperparameter position by position to obtain the motor temperature time sequence mode feature vector;

[0020] The motor speed time sequence mode feature vector is point convolution coded to obtain a convolution motor speed time sequence mode feature vector, and the convolution motor speed time sequence mode feature vector is multiplied by a second hyperparameter position by position to obtain the motor speed time sequence mode feature vector.

[0021] Optionally, the operating state feature construction module comprises:

[0022] The motor pressure time sequence correlation unit is configured to calculate a product between a transposed vector of the motor temperature-speed time sequence mode fusion feature vector of a motor operating state and a hydraulic system pressure time sequence mode feature vector to obtain a motor state pressure time sequence correlation input matrix;

[0023] The correlation value calculation unit is configured to calculate a sum of each element on a diagonal line in the motor state pressure time sequence correlation input matrix to obtain a motor state pressure time sequence correlation value;

[0024] The operating state feature construction unit is configured to divide feature values of each position of the motor state pressure time sequence correlation input matrix by the motor state pressure time sequence correlation value respectively to obtain a bucket wheel machine operating state feature matrix composed of a plurality of motor state pressure time sequence correlation trace sets.

[0025] Optionally, the identification result generation module comprises:

[0026] The fault judgment unit is configured to input the bucket wheel machine operating state feature matrix into a bucket wheel machine fault identifier based on a classifier to obtain a corresponding identification result, and the identification result is used to indicate whether the bucket wheel machine has a fault;

[0027] The early warning prompt generation unit is configured to issue the fault early warning prompt in response to the identification result indicating that the bucket wheel machine has a fault.

[0028] Optionally, the bucket wheel machine automatic monitoring and alarm system further comprises a training module.

[0029] The training module is configured to:

[0030] The operation parameter time sequence mode feature extraction module is configured to train the causal convolution network-based bucket wheel machine operation parameter time sequence mode feature extractor by inputting the motor temperature time sequence input vector, the motor speed time sequence input vector and the hydraulic system pressure time sequence input vector into the causal convolution network-based bucket wheel machine operation parameter time sequence mode feature extractor, respectively.

[0031] The motor operation state feature fusion module is configured to train the motor operation state capturer based on the motor operation parameter time sequence fusion module by inputting the motor temperature time sequence mode feature vector and the motor speed time sequence mode feature vector into the motor operation state capturer based on the motor operation parameter time sequence fusion module.

[0032] When the identification result generation module comprises a fault judgment unit, the classifier-based bucket wheel machine fault identifier is trained.

[0033] Optionally, the training module comprises:

[0034] The training data acquisition unit is configured to acquire training data and a true value of whether the bucket wheel machine has a fault, wherein the training data comprises a time sequence of bucket wheel machine operation training parameters, and the bucket wheel machine operation training parameters comprise training motor temperature values, training motor speed values and training hydraulic system pressure values of the bucket wheel machine.

[0035] The training data normalization unit is configured to perform data normalization on the time sequence of the bucket wheel machine operation training parameters in terms of time dimension and sample dimension, to obtain a training motor temperature time sequence input vector, a training motor speed time sequence input vector and a training hydraulic system pressure time sequence input vector.

[0036] The training parameter time sequence mode feature extraction unit is configured to input the training motor temperature time sequence input vector, the training motor speed time sequence input vector and the training hydraulic system pressure time sequence input vector into the causal convolution network-based bucket wheel machine operation parameter time sequence mode feature extractor, respectively, to obtain a corresponding training motor temperature time sequence mode feature vector, a training motor speed time sequence mode feature vector and a training hydraulic system pressure time sequence mode feature vector.

[0037] The training parameter time sequence fusion unit is configured to input the training motor temperature time sequence mode feature vector and the training motor speed time sequence mode feature vector into the motor operation state capturer based on the motor operation parameter time sequence fusion module, to obtain a training motor temperature-speed time sequence mode fusion representation motor operation state feature vector.

[0038] The training feature correlation unit is configured to perform dependency scale calculation on the training motor temperature-speed time sequence mode fusion representation motor operation state feature vector and the training hydraulic system pressure time sequence mode feature vector to obtain a training bucket wheel machine operation state representation matrix.

[0039] The classification loss unit is configured to input the training bucket wheel machine operation state representation matrix into the classifier-based bucket wheel machine fault identifier to obtain a predicted value of whether the bucket wheel machine has a fault, and determine a classification loss function value according to the predicted value of whether the bucket wheel machine has a fault and a true value of whether the bucket wheel machine has a fault.

[0040] The training unit is configured to train the causal convolution network-based bucket wheel machine operation parameter time sequence mode feature extractor, the motor operation state capturer based on the motor operation parameter time sequence fusion module, and the classifier-based bucket wheel machine fault identifier based on the classification loss function value and through back propagation of gradient descent.

[0041] Another aspect of the present disclosure provides a bucket wheel machine automatic monitoring and alarming method based on the Internet of Things, which comprises:

[0042] Obtaining a time sequence of bucket wheel machine operation parameters, wherein the bucket wheel machine operation parameters include a motor temperature value, a motor speed value, and a hydraulic system pressure value of the bucket wheel machine.

[0043] Performing data regularization on the time sequence of the bucket wheel machine operation parameters according to time dimensions and sample dimensions to obtain corresponding motor temperature time sequence input vectors, motor speed time sequence input vectors, and hydraulic system pressure time sequence input vectors.

[0044] Performing bucket wheel machine operation parameter time sequence mode feature extraction on the motor temperature time sequence input vectors, the motor speed time sequence input vectors, and the hydraulic system pressure time sequence input vectors to obtain corresponding motor temperature time sequence mode feature vectors, motor speed time sequence mode feature vectors, and hydraulic system pressure time sequence mode feature vectors.

[0045] Performing motor operation state capture fusion on the motor temperature time sequence mode feature vectors and the motor speed time sequence mode feature vectors to obtain a motor temperature-speed time sequence mode fusion representation motor operation state feature vector.

[0046] Performing dependency scale calculation on the motor temperature-speed time sequence mode fusion representation motor operation state feature vector and the hydraulic system pressure time sequence mode feature vector to obtain a bucket wheel machine operation state representation matrix as a bucket wheel machine operation state representation feature.

[0047] Based on the bucket wheel machine operation state representation features, a corresponding recognition result is obtained, and based on the recognition result, it is determined whether to issue a fault early warning prompt.

[0048] The present disclosure, relative to the prior art, automatically judges the operation state of the bucket wheel machine according to the time sequence representation features of various operation parameters of the bucket wheel machine by acquiring the time sequence of the operation parameters of the bucket wheel machine, using data collection, processing and analysis algorithms based on the Internet of Things and artificial intelligence to perform time sequence analysis and mutual fusion on the operation parameters of the bucket wheel machine, and can issue a prompt in a timely manner according to the operation state of the bucket wheel machine, effectively improving the accuracy of fault detection of the bucket wheel machine, and improving the reliability and work efficiency of the equipment. BRIEF DESCRIPTION OF DRAWINGS

[0049] One or more embodiments are illustrated by way of example in the figures that are part of this disclosure and which are illustrative, but not restrictive, of the embodiments, wherein elements having the same reference number designates like elements, unless otherwise explicitly stated, the figures in the drawings do not constitute a proportional limitation.

[0050] Figure 1 A structural schematic diagram of an automatic monitoring and alarm system for a bucket wheel machine based on the Internet of Things is provided for an embodiment of the present disclosure;

[0051] Figure 2 A data flow direction schematic diagram of an automatic monitoring and alarm system for a bucket wheel machine based on the Internet of Things is provided for another embodiment of the present disclosure;

[0052] Figure 3 A structural block diagram of a motor operation state feature fusion module in an automatic monitoring and alarm system for a bucket wheel machine based on the Internet of Things is provided for another embodiment of the present disclosure;

[0053] Figure 4 A structural block diagram of an operation state representation module in an automatic monitoring and alarm system for a bucket wheel machine based on the Internet of Things is provided for another embodiment of the present disclosure;

[0054] Figure 5 A structural block diagram of a training module in an automatic monitoring and alarm system for a bucket wheel machine based on the Internet of Things is provided for another embodiment of the present disclosure;

[0055] Figure 6 A flowchart of an automatic monitoring and alarm method for a bucket wheel machine based on the Internet of Things is provided for an embodiment of the present disclosure. DETAILED DESCRIPTION

[0056] In order to make the purposes, technical solutions and advantages of the embodiments of the present disclosure clearer, the embodiments of the present disclosure will be described in detail below with reference to the drawings. However, those skilled in the art can understand that, in the embodiments of the present disclosure, many technical details are presented in order to make the readers better understand the present disclosure. However, the technical solutions claimed by the present disclosure can be implemented even without these technical details and based on various changes and modifications of the following embodiments. The division of the following embodiments is for the convenience of description, and should not constitute any limitation on the specific embodiments of the present disclosure, and the embodiments can be combined with each other and referred to each other without contradiction.

[0057] Unless the context clearly indicates otherwise, the words "one", "an", "a", and / or "the" do not mean a singular object, but can include a plurality of objects. Generally, the terms "comprising" and "including" only indicate the inclusion of the steps and elements explicitly identified, and these steps and elements do not constitute an exclusive list, and the method or device can also include other steps or elements.

[0058] Although the present disclosure makes various references to certain modules in the system according to the embodiments of the present disclosure, however, any number of different modules can be used and run on the user terminal and / or server. The modules are only illustrative, and different aspects of the system and method can use different modules.

[0059] The present disclosure uses flowcharts to illustrate the operations performed by the method according to the embodiments of the present disclosure. It should be understood that the preceding or following operations are not necessarily performed in sequence. On the contrary, various steps can be processed in reverse order or simultaneously as needed. Meanwhile, other operations can also be added to these processes, or one or more steps of operations can be removed from these processes.

[0060] In the prior art, the Internet of Things can collect information of objects or processes that need to be monitored, connected and interacted in real time through various information sensors, radio frequency identification technology, global positioning system and other technologies, and realize ubiquitous connection between things and things, and between things and people. In the automatic monitoring and alarm system of the bucket wheel machine, the Internet of Things technology plays a crucial role. For example, by installing sensors and monitoring devices, the bucket wheel machine can realize functions such as automatic addressing, automatic coal piling and automatic stacking, thereby improving work efficiency and safety. For another example, by using the Internet of Things technology, the management personnel can view the running state and operation data of the bucket wheel machine in real time at the remote monitoring center. Therefore, the application of the Internet of Things in the monitoring of the bucket wheel machine not only improves the accuracy and efficiency of the operation of the bucket wheel machine, but also provides strong support for modern industrial automation.

[0061] Based on this, in view of the problems existing in the traditional bucket wheel machine fault detection method, the disclosure provides a bucket wheel machine automatic monitoring alarm system and method based on Internet of Things, and the technical concept is: obtaining the time sequence of the bucket wheel machine operation parameters, wherein the bucket wheel machine operation parameters include motor temperature value, motor speed value and hydraulic system pressure value, and using the data collection, processing and analysis algorithm based on Internet of Things and artificial intelligence to perform time sequence analysis and mutual fusion on the bucket wheel machine operation parameters, so as to automatically judge whether the bucket wheel machine has a fault according to the characteristic features of the bucket wheel machine operation parameters in time sequence, and determine whether to issue a corresponding fault warning prompt based on whether the bucket wheel machine has a fault. Through this way, the bucket wheel machine automatic monitoring alarm system can automatically judge the operation state of the bucket wheel machine and timely issue a warning prompt, thereby improving the accuracy of the bucket wheel machine fault detection, and improving the reliability and working efficiency of the equipment.

[0062] One embodiment of the disclosure relates to a bucket wheel machine automatic monitoring alarm system 100 based on Internet of Things, the structure of which is as shown in Figure 1 , and the data flow direction is as shown in Figure 2 .

[0063] In combination with Figure 1 and Figure 2 , the bucket wheel machine automatic monitoring alarm system 100 based on Internet of Things provided by the embodiment includes a bucket wheel machine operation parameter acquisition module 110, a data regularization module 120, an operation parameter time sequence mode feature extraction module 130, a motor operation state feature fusion module 140, an operation state representation module 150 and an identification result generation module 160.

[0064] The bucket wheel machine operation parameter acquisition module 110 is used to obtain the time sequence of the bucket wheel machine operation parameters, and the bucket wheel machine operation parameters include the motor temperature value, the motor speed value and the hydraulic system pressure value of the bucket wheel machine.

[0065] Specifically, the bucket wheel machine operating parameter acquisition module 110 can obtain the time series of the bucket wheel machine operating parameters by acquiring the time series of the bucket wheel machine operating parameters collected by the sensor group deployed on the bucket wheel machine. Considering that the motor temperature value, the motor speed value and the hydraulic system pressure value are three key operating parameters of the bucket wheel machine, in order to help monitor the operating state of the bucket wheel machine and ensure that the equipment operates according to the predetermined performance and safety standards, the bucket wheel machine operating parameters in the embodiment include the motor temperature value, the motor speed value and the hydraulic system pressure value of the bucket wheel machine. Among them, the motor temperature value refers to the temperature generated by the motor in the bucket wheel machine during operation, which is an important indicator for evaluating whether the motor is overheated. Overheating of the motor may cause damage to the motor or reduce its efficiency. The motor speed value refers to the rotational speed of the motor in the bucket wheel machine, which is usually measured in revolutions per minute (RPM) and is related to the operating speed and yield of the bucket wheel machine. The hydraulic system pressure value refers to the pressure level in the hydraulic system of the bucket wheel machine. The hydraulic system of the bucket wheel machine is responsible for providing power and performing various mechanical actions, and its pressure value is an important indicator of the normal operation and load capacity of the hydraulic system of the bucket wheel machine.

[0066] The data normalization module 120 is configured to normalize the time series of the bucket wheel machine operating parameters according to the time dimension and the sample dimension to obtain the corresponding motor temperature time series input vector, motor speed time series input vector and hydraulic system pressure time series input vector.

[0067] Specifically, considering that the bucket wheel machine operating parameters all have characteristic information in the time dimension, and the motor temperature value, the motor speed value and the hydraulic system pressure value are different types of physical parameters, i.e. different sample dimensions, which have different characteristic information in time series, in order to preserve the order of the bucket wheel machine operating parameters in time to capture the dynamic characteristics of each parameter data in the time dimension, the data normalization module 120 normalizes the time series of the bucket wheel machine operating parameters according to the time dimension and the sample dimension to obtain the corresponding motor temperature time series input vector, motor speed time series input vector and hydraulic system pressure time series input vector. This can independently analyze the time series characteristics of each bucket wheel machine operating parameter in order to more accurately capture and extract the time series fluctuations and periodic changes in each operating parameter.

[0068] The operating parameter time series pattern feature extraction module 130 is configured to extract the bucket wheel machine operating parameter time series pattern features from the motor temperature time series input vector, the motor speed time series input vector and the hydraulic system pressure time series input vector respectively to obtain the corresponding motor temperature time series pattern feature vector, motor speed time series pattern feature vector and hydraulic system pressure time series pattern feature vector.

[0069] The motor temperature time series mode feature extraction module 130 is configured to input the motor temperature time series input vector, the motor speed time series input vector and the hydraulic system pressure time series input vector into the bucket wheel machine running parameter time series mode feature extractor based on the causal convolution network, to obtain the motor temperature time series mode feature vector, the motor speed time series mode feature vector and the hydraulic system pressure time series mode feature vector.

[0070] Specifically, considering that the motor temperature value, the motor speed value and the hydraulic system pressure value are collected and obtained in time sequence, that is, the data value at the current time point is affected by the previous time point and has nothing to do with the future time, therefore, in order to capture the important causal mode features of time from the motor temperature value, the motor speed value and the hydraulic system pressure value, the motor temperature time series input vector, the motor speed time series input vector and the hydraulic system pressure time series input vector are input into the bucket wheel machine running parameter time series mode feature extractor based on the causal convolution network, to obtain the motor temperature time series mode feature vector, the motor speed time series mode feature vector and the hydraulic system pressure time series mode feature vector.

[0071] The causal convolution network is a convolution network considering the time causal relationship, which can maintain the causality of time series data, that is, in time series data, past data will affect future data, therefore, the causal convolution can better capture this causal relationship. Specifically, the bucket wheel machine running parameter time series mode feature extractor based on the causal convolution network can better capture and mine important time series causal mode features in the motor temperature time series input vector, the motor speed time series input vector and the hydraulic system pressure time series input vector, thereby helping to understand the regularity and time series features of the bucket wheel machine running parameters.

[0072] The motor running state feature fusion module 140 is configured to perform motor running state capture fusion on the motor temperature time series mode feature vector and the motor speed time series mode feature vector, to obtain the motor temperature-speed time series mode fusion representing motor running state feature vector.

[0073] The motor running state feature fusion module 140 is configured to input the motor temperature time series mode feature vector and the motor speed time series mode feature vector into the motor running state capturer based on the motor running parameter time series fusion module, to obtain the motor temperature-speed time series mode fusion representing motor running state feature vector.

[0074] Specifically, the motor temperature time series pattern feature vector and the motor speed time series pattern feature vector respectively represent the feature information of the motor temperature and the motor speed on the time scale, and there is mutual correlation and interaction information about the motor operating state between the motor temperature and the motor speed, therefore, in order to more comprehensively and comprehensively represent the time series feature information of the motor operating state, facilitate more accurate fault judgment subsequently, the motor operating state feature fusion module 140 is specifically configured to input the motor temperature time series pattern feature vector and the motor speed time series pattern feature vector into the motor operating state capturer based on the motor operating parameter time series fusion module for fusion, to comprehensively consider the pattern features of the motor temperature and the motor speed on the time series, and to generate a more comprehensive motor temperature-speed time series pattern fusion representing motor operating state feature vector based on the feature correlation between the motor temperature and the motor speed on the time series pattern, so that different aspect information about the motor operating state can be fused, so as to capture the comprehensive operating state time series feature of the motor, improve the grasp of the motor operating state, and thus realize more accurate subsequent fault prediction.

[0075] As shown in the figure, Figure 3 The motor operating state feature fusion module 140 includes a feature representation construction unit 141 and a motor parameter feature interaction unit 142.

[0076] The feature representation construction unit 141 is configured to construct a motor temperature time series pattern feature vector and a motor speed time series pattern feature vector based on the motor temperature time series pattern feature vector and the motor speed time series pattern feature vector.

[0077] The motor parameter feature interaction unit 142 is configured to add the motor temperature time series pattern feature vector and the motor speed time series pattern feature vector by position to obtain a motor temperature-speed time series pattern fusion representing motor operating state feature vector.

[0078] The feature representation construction unit 141 is specifically configured to: point convolution coding is performed on the motor temperature time series pattern feature vector to obtain a convolution motor temperature time series pattern feature vector, the convolution motor temperature time series pattern feature vector is multiplied by position with the first hyperparameter to obtain the motor temperature time series pattern feature vector; point convolution coding is performed on the motor speed time series pattern feature vector to obtain a convolution motor speed time series pattern feature vector, and the convolution motor speed time series pattern feature vector is multiplied by position with the second hyperparameter to obtain the motor speed time series pattern feature vector.

[0079] In other words, the motor operating state feature fusion module 140 can be specifically configured to use the motor operating state capturer based on the motor operating parameter time series fusion module to fuse the motor temperature time series pattern feature vector and the motor speed time series pattern feature vector using the following fusion formula to obtain the motor temperature-speed time series pattern fusion motor operating state feature vector, wherein the fusion formula is:

[0080] V c = a * Conv 1×1 (V1) + b * Conv 1×1 (V2)

[0081] wherein V1 and V2 represent the motor temperature time series pattern feature vector and the motor speed time series pattern feature vector respectively, Conv 1×1 (·) represents point convolution encoding of the vector in the parentheses, a and b represent the first and second hyperparameters respectively, and V c represents the motor temperature-speed time series pattern fusion motor operating state feature vector.

[0082] The operating state representation module 150 is configured to perform dependent scale calculation on the motor temperature-speed time series pattern fusion motor operating state feature vector and the hydraulic system pressure time series pattern feature vector to obtain a bucket wheel machine operating state representation matrix as the bucket wheel machine operating state representation feature.

[0083] Specifically, considering that the motor temperature-speed time series pattern fusion motor operating state feature vector contains comprehensive time series feature information of motor temperature and motor speed, it can more comprehensively describe the operating state of the motor, while the hydraulic system pressure is also an important parameter of the bucket wheel machine operation, and the hydraulic system pressure time series pattern feature vector can provide important time series information about the hydraulic system state. At the same time, there is some dependent correlation between the motor operating state feature and the hydraulic system pressure time series pattern feature, such as in some cases, the change of motor speed may cause the fluctuation of hydraulic system pressure, thereby reflecting the change of the overall working state of the bucket wheel machine. Therefore, in order to better understand and describe the operating state of the bucket wheel machine, and thus obtain a more comprehensive and accurate bucket wheel machine operating state representation, the operating state representation module 150 is configured to perform dependent scale calculation on the motor temperature-speed time series pattern fusion motor operating state feature vector and the hydraulic system pressure time series pattern feature vector to obtain a bucket wheel machine operating state representation matrix. Through dependent scale calculation, the time series correlation and mutual dependence between different operating parameters of the bucket wheel machine can be quantified, which helps to more comprehensively understand the overall operating state of the bucket wheel machine to form a more comprehensive and global bucket wheel machine operating state representation matrix, providing stronger support for subsequent fault diagnosis analysis and decision-making, and helping to discover potential problems or abnormal situations.

[0084] As shown in the figure, the running state representation module 150 comprises a motor pressure time sequence correlation unit 151, a correlation value calculation unit 152, and a running state representation unit 153. Figure 4

[0085] The motor pressure time sequence correlation unit 151 is configured to calculate the product between the motor temperature-rotation speed time sequence mode fusion representation motor running state feature vector and the transposed vector of the hydraulic system pressure time sequence mode feature vector, to obtain a motor state pressure time sequence correlation input matrix.

[0086] The correlation value calculation unit 152 is configured to calculate the sum of each element on the diagonal of the motor state pressure time sequence correlation input matrix, to obtain a motor state pressure time sequence correlation value.

[0087] The running state representation unit 153 is configured to divide the feature value of each position of the motor state pressure time sequence correlation input matrix by the motor state pressure time sequence correlation value, to obtain a bucket wheel machine running state representation matrix composed of a plurality of motor state pressure time sequence correlation trace vectors.

[0088] In other words, the running state representation module 150 can be specifically configured to: perform dependency scale calculation on the motor temperature-rotation speed time sequence mode fusion representation motor running state feature vector and the hydraulic system pressure time sequence mode feature vector by using the following correlation formula, to obtain the bucket wheel machine running state representation matrix, wherein the correlation formula is:

[0089]

[0090] wherein V c represents the motor temperature-rotation speed time sequence mode fusion representation motor running state feature vector, V s represents the hydraulic system pressure time sequence mode feature vector, tr(·) represents the sum of each element on the diagonal of the matrix in the parentheses, M V represents the bucket wheel machine running state representation matrix.

[0091] The recognition result generation module 160 is configured to obtain a corresponding recognition result based on the bucket wheel machine running state representation feature, and determine whether to issue a fault warning prompt based on the recognition result.

[0092] As shown in the figure, the recognition result generation module 160 comprises a fault judgment unit and a warning prompt generation unit.

[0093] The fault judgment unit is configured to input the bucket wheel machine running state representation matrix into a bucket wheel machine fault recognizer based on a classifier, to obtain a corresponding recognition result, which is used to indicate whether the bucket wheel machine has a fault.

[0094] ​The early warning generation unit is used to issue a fault warning in response to the identification result indicating that there is a fault in the bucket wheel excavator.

[0095] Specifically, the identification result generation module 160 uses the characteristic vector representing the motor's operating state through a fusion of the motor temperature-speed time-series pattern and the characteristic vector representing the hydraulic system's pressure time-series pattern to perform dependency calculations on the bucket wheel excavator's operating state characteristics. Based on the classification results, it automatically determines whether the bucket wheel excavator has a fault and issues a corresponding fault warning when a fault is detected. In this way, the bucket wheel excavator automatic monitoring and alarm system 100 can automatically determine the bucket wheel excavator's operating state and issue timely warnings when a fault occurs, improving not only the accuracy of fault detection but also the reliability and efficiency of the equipment.

[0096] It should be noted that before applying the deep neural network model for inference, the deep neural network model needs to be trained first so that it can perform specific functions. Therefore, the IoT-based bucket wheel excavator automatic monitoring and alarm system 100 also includes a training module 200. The training module 200 is used to: train the bucket wheel excavator operating parameter time-series pattern feature extractor when the operating parameter time-series feature extraction module specifically inputs the motor temperature time-series input vector, motor speed time-series input vector, and hydraulic system pressure time-series input vector into the bucket wheel excavator operating parameter time-series pattern feature extractor based on a causal convolutional network; train the motor operating state feature fusion module when the motor temperature time-series pattern feature vector and motor speed time-series pattern feature vector are input into the motor operating state capture device based on the motor operating parameter time-series fusion module; and train the bucket wheel excavator fault identifyr based on a classifier when the identification result generation module includes a fault judgment unit. That is, the training module 200 is used to train the bucket wheel excavator operating parameter time-series pattern feature extractor based on causal convolutional network, the motor operating state capturer based on motor operating parameter time-series fusion module, and the bucket wheel excavator fault identifyr based on classifier in the above embodiments.

[0097] For example, such as Figure 5 As shown, the training module 200 includes a training data acquisition unit 210, a training data normalization unit 220, a training parameter temporal pattern feature extraction unit 230, a training parameter temporal fusion unit 240, a training feature association unit 250, a classification loss unit 260, and a training unit 270.

[0098] The training data acquisition unit 210 is configured to acquire training data and a true value of whether the bucket wheel machine has a fault; wherein the training data comprises a time sequence of bucket wheel machine operation training parameters, and the bucket wheel machine operation training parameters comprise a training motor temperature value, a training motor speed value and a training hydraulic system pressure value of the bucket wheel machine.

[0099] The training data normalization unit 220 is configured to normalize the time sequence of bucket wheel machine operation training parameters in the time dimension and the sample dimension to obtain a training motor temperature time sequence input vector, a training motor speed time sequence input vector and a training hydraulic system pressure time sequence input vector.

[0100] The training parameter time sequence mode feature extraction unit 230 is configured to input the training motor temperature time sequence input vector, the training motor speed time sequence input vector and the training hydraulic system pressure time sequence input vector into a bucket wheel machine operation parameter time sequence mode feature extractor based on a causal convolution network respectively to obtain a corresponding training motor temperature time sequence mode feature vector, a training motor speed time sequence mode feature vector and a training hydraulic system pressure time sequence mode feature vector.

[0101] The training parameter time sequence fusion unit 240 is configured to input the training motor temperature time sequence mode feature vector and the training motor speed time sequence mode feature vector into a motor operation state capturer based on a motor operation parameter time sequence fusion module to obtain a training motor temperature-speed time sequence mode fusion representing motor operation state feature vector.

[0102] The training feature correlation unit 250 is configured to perform dependency scale calculation on the training motor temperature-speed time sequence mode fusion representing motor operation state feature vector and the training hydraulic system pressure time sequence mode feature vector to obtain a training bucket wheel machine operation state representation matrix.

[0103] The classification loss unit 260 is configured to input the training bucket wheel machine operation state representation matrix into a bucket wheel machine fault identifier based on a classifier to obtain a predicted value of whether the bucket wheel machine has a fault, and determine a classification loss function value according to the predicted value of whether the bucket wheel machine has a fault and the true value of whether the bucket wheel machine has a fault.

[0104] The training unit 270 is configured to train the bucket wheel machine operation parameter time sequence mode feature extractor based on the causal convolution network, the motor operation state capturer based on the motor operation parameter time sequence fusion module and the bucket wheel machine fault identifier based on the classifier based on the classification loss function value and through back propagation of gradient descent.

[0105] For example, when the training unit 270 trains the bucket wheel machine operation parameter time sequence pattern feature extractor based on the causal convolution network, the motor operation state capturer based on the motor operation parameter time sequence fusion module, and the bucket wheel machine fault identifier based on the classifier based on the classification loss function value and through the back propagation of gradient descent, the distribution noise loss function value can be further introduced, wherein the distribution noise loss function value can be calculated by the following steps: inputting the training bucket wheel machine operation state representation matrix into the bucket wheel machine fault identifier based on the classifier to obtain the class probability value, which can be used as the prediction value of whether the bucket wheel machine has a fault, taking the true value of whether the bucket wheel machine has a fault as 1, calculating the absolute value of the difference between the class probability value and 1 to obtain the probability difference value; determining the feature mean and feature variance of each feature value of the training bucket wheel machine operation state representation matrix as the feature set; determining the first product of each feature value in the training bucket wheel machine operation state representation matrix and the class probability value and the second product of the probability difference value and the feature mean; summing the square of the difference between the first product and the second product corresponding to each feature value in the training bucket wheel machine operation state representation matrix and dividing by the length of the training bucket wheel machine operation state representation matrix to obtain the division result; subtracting the product of the square sum of the class probability value and the probability difference value and the feature variance from the division result to obtain the distribution noise loss function value. After obtaining the distribution noise loss function value, the training unit 270 can take the weighted sum of the classification loss function value and the multi-distribution noise loss function value as the loss function value, and train the bucket wheel machine operation parameter time sequence pattern feature extractor based on the causal convolution network, the motor operation state capturer based on the motor operation parameter time sequence fusion module, and the bucket wheel machine fault identifier based on the classifier through the back propagation of gradient descent.

[0106] In particular, when performing dependency scale calculation on the training motor temperature-rotation speed time sequence pattern fusion representation motor operation state feature vector and the training hydraulic system pressure time sequence pattern feature vector, the correlation difference of the time domain distribution of different source data can cause distribution external noise of the correlation representation set, produce class distribution noise of the whole training bucket wheel machine operation state representation matrix, and thus affect the accuracy of the identification result. The above distribution noise loss function value sets the approximate variance expectation characteristics of the distribution external noise of the feature set of the training bucket wheel machine operation state representation matrix relative to the Bernoulli sampling instances of the feature set through the class probability distribution characteristics expressed by the class probability value and the probability difference value obtained by inputting the training bucket wheel machine operation state representation matrix into the bucket wheel machine fault identifier based on the classifier.

[0107] Therefore, by using the statistical noise distribution based on the class probability value and the probability difference value expressed by the class probability distribution characteristics of the distributed noise loss function value, the possible sampling noise features can be degraded based on the probability cross random selection characteristics of the expected variance when the class probability distribution of each feature value of the training bucket wheel machine operation state representation matrix in the classification process, so as to realize the class distribution noise reduction of self-supervised learning in the model parameter updating process, thereby improving the accuracy of the identification result.

[0108] The bucket wheel machine automatic monitoring and alarm system based on the Internet of Things provided by the embodiment of the present disclosure can automatically judge the operation state of the bucket wheel machine according to the characteristic features of the time sequence of each operation parameter of the bucket wheel machine, and can timely issue a warning prompt according to the operation state of the bucket wheel machine, thereby effectively improving the accuracy of fault detection of the bucket wheel machine, and improving the reliability and working efficiency of the equipment.

[0109] For example, the bucket wheel machine automatic monitoring and alarm system based on the Internet of Things 100 provided by the embodiment of the present disclosure can be implemented in various wireless terminals, for example, the wireless terminal can be a server with a bucket wheel machine automatic monitoring and alarm algorithm based on the Internet of Things, etc. In one possible implementation, the bucket wheel machine automatic monitoring and alarm system based on the Internet of Things 100 provided by the embodiment of the present disclosure can be integrated into the wireless terminal as a software module and / or a hardware module. For example, the bucket wheel machine automatic monitoring and alarm system based on the Internet of Things 100 can be a software module in the operating system of the wireless terminal, or can be an application program developed for the wireless terminal; of course, the bucket wheel machine automatic monitoring and alarm system based on the Internet of Things 100 can also be one of the many hardware modules of the wireless terminal.

[0110] For example, the bucket wheel machine automatic monitoring and alarm system based on the Internet of Things 100 and the wireless terminal can also be separate devices, and the bucket wheel machine automatic monitoring and alarm system based on the Internet of Things 100 can be connected to the wireless terminal through a wired and / or wireless network, and transmit interactive information in a predetermined data format.

[0111] Another embodiment of the present disclosure relates to a bucket wheel machine automatic monitoring and alarm method based on the Internet of Things, as shown in Figure 6 The method comprises the following steps:

[0112] In step S110, the time sequence of the bucket wheel machine operation parameters is obtained, and the bucket wheel machine operation parameters include the motor temperature value, the motor speed value and the hydraulic system pressure value of the bucket wheel machine.

[0113] Step S120, the time series of the bucket wheel machine operation parameter is data regularized according to the time dimension and the sample dimension, to obtain the motor temperature time sequence input vector, the motor speed time sequence input vector and the hydraulic system pressure time sequence input vector.

[0114] Step S130, the motor temperature time sequence input vector, the motor speed time sequence input vector and the hydraulic system pressure time sequence input vector are respectively subjected to the bucket wheel machine operation parameter time sequence mode feature extraction, to obtain the motor temperature time sequence mode feature vector, the motor speed time sequence mode feature vector and the hydraulic system pressure time sequence mode feature vector.

[0115] Step S140, the motor temperature time sequence mode feature vector and the motor speed time sequence mode feature vector are subjected to the motor operation state capture fusion, to obtain the motor temperature-speed time sequence mode fusion representing the motor operation state feature vector.

[0116] Step S150, the motor temperature-speed time sequence mode fusion representing the motor operation state feature vector and the hydraulic system pressure time sequence mode feature vector are subjected to the dependency scale calculation, to obtain the bucket wheel machine operation state representation matrix as the bucket wheel machine operation state representation feature.

[0117] Step S160, based on the bucket wheel machine operation state representation feature, the corresponding recognition result is obtained, and whether to issue the fault warning prompt is determined based on the recognition result.

[0118] The specific steps of the bucket wheel machine automatic monitoring and alarming method based on the Internet of Things provided by the embodiments of the present disclosure can be referred to the specific implementation manners of the various modules in the bucket wheel machine automatic monitoring and alarming system based on the Internet of Things provided by the embodiments of the present disclosure, which will not be described here.

[0119] The bucket wheel machine automatic monitoring and alarming method based on the Internet of Things provided by the embodiments of the present disclosure, compared with the prior art, by acquiring the time series of the bucket wheel machine operation parameter, using the data acquisition, processing and analysis algorithm based on the Internet of Things and artificial intelligence to perform the time sequence analysis and mutual fusion on the bucket wheel machine operation parameter, so as to automatically judge the operation state of the bucket wheel machine according to the representation features of the various operation parameters of the bucket wheel machine in time sequence, and timely issue the warning prompt according to the operation state of the bucket wheel machine, effectively improving the accuracy of the bucket wheel machine fault detection, and improving the reliability and working efficiency of the equipment.

[0120] Those skilled in the art can understand that the above-mentioned embodiments are specific embodiments for implementing the present disclosure, and in actual application, various changes can be made in form and details without departing from the spirit and scope of the present disclosure.

Claims

1. An automatic monitoring and alarm system for a bucket wheel machine based on the Internet of Things, characterized by The automatic monitoring and alarm system of the bucket wheel machine comprises: A bucket wheel machine operation parameter acquisition module is configured to obtain a time sequence of bucket wheel machine operation parameters, wherein the bucket wheel machine operation parameters include a motor temperature value, a motor speed value, and a hydraulic system pressure value of the bucket wheel machine. A data regularization module is configured to perform data regularization on the time sequence of the bucket wheel machine operation parameters according to time dimensions and sample dimensions to obtain a motor temperature time sequence input vector, a motor speed time sequence input vector, and a hydraulic system pressure time sequence input vector. An operation parameter time sequence mode feature extraction module is configured to perform bucket wheel machine operation parameter time sequence mode feature extraction on the motor temperature time sequence input vector, the motor speed time sequence input vector, and the hydraulic system pressure time sequence input vector, respectively, to obtain a motor temperature time sequence mode feature vector, a motor speed time sequence mode feature vector, and a hydraulic system pressure time sequence mode feature vector. A motor operation state feature fusion module is configured to perform motor operation state capture fusion on the motor temperature time sequence mode feature vector and the motor speed time sequence mode feature vector to obtain a motor temperature-speed time sequence mode fusion representing motor operation state feature vector. An operation state representation module is configured to perform dependency scale calculation on the motor temperature-speed time sequence mode fusion representing motor operation state feature vector and the hydraulic system pressure time sequence mode feature vector to obtain a bucket wheel machine operation state representation matrix as a bucket wheel machine operation state representation feature. An identification result generation module is configured to obtain a corresponding identification result based on the bucket wheel machine operation state representation feature and determine whether to issue a fault warning prompt based on the identification result.

2. The automatic monitoring and alarm system for an excavator according to claim 1, wherein The operation parameter time sequence mode feature extraction module is specifically configured to input the motor temperature time sequence input vector, the motor speed time sequence input vector, and the hydraulic system pressure time sequence input vector into a bucket wheel machine operation parameter time sequence mode feature extractor based on a causal convolution network to obtain the motor temperature time sequence mode feature vector, the motor speed time sequence mode feature vector, and the hydraulic system pressure time sequence mode feature vector.

3. The automatic monitoring and alarming system for bucket wheel machines according to claim 1, characterized in that The motor operation state feature fusion module is specifically configured to input the motor temperature time sequence mode feature vector and the motor speed time sequence mode feature vector into a motor operation state capturer based on a motor operation parameter time sequence fusion module to obtain the motor temperature-speed time sequence mode fusion representing motor operation state feature vector.

4. The automatic monitoring and alarm system for bucket wheel machines according to claim 3, characterized in that The motor operation state feature fusion module comprises: A representation feature construction unit is configured to construct a corresponding motor temperature time sequence mode representation feature vector and a motor speed time sequence mode representation feature vector based on the motor temperature time sequence mode feature vector and the motor speed time sequence mode feature vector, respectively. A motor parameter feature interaction unit is configured to add the motor temperature time sequence mode representation feature vector and the motor speed time sequence mode representation feature vector position by position to obtain the motor temperature-speed time sequence mode fusion representing motor operation state feature vector.

5. The automatic monitoring and alarm system for bucket wheel machines according to claim 4, characterized in that The representation feature construction unit is specifically configured to: The motor temperature time sequence mode feature vector is point convolution coded to obtain a convolution motor temperature time sequence mode feature vector, and the convolution motor temperature time sequence mode feature vector is multiplied by a first hyperparameter at positions to obtain the motor temperature time sequence mode representation feature vector; The motor speed time sequence mode feature vector is point convolution coded to obtain a convolution motor speed time sequence mode feature vector, and the convolution motor speed time sequence mode feature vector is multiplied by a second hyperparameter at positions to obtain the motor speed time sequence mode representation feature vector.

6. The automatic monitoring and alarm system for bucket wheel machines according to claim 1, characterized in that The operation state representation module comprises: The motor pressure time sequence correlation unit is configured to calculate a product of a transposed vector of the motor temperature-speed time sequence mode fusion representation motor operation state feature vector and the hydraulic system pressure time sequence mode feature vector to obtain a motor state pressure time sequence correlation input matrix; The correlation value calculation unit is configured to calculate a sum of each element on a diagonal line in the motor state pressure time sequence correlation input matrix to obtain a motor state pressure time sequence correlation value; The operation state representation unit is configured to divide feature values of each position of the motor state pressure time sequence correlation input matrix by the motor state pressure time sequence correlation value respectively to obtain the bucket wheel machine operation state representation matrix composed of a plurality of motor state pressure time sequence correlation trace coefficient groups.

7. The automatic monitoring and alarming system for bucket wheel excavators according to claim 1, characterized in that The identification result generation module comprises: The fault judgment unit is configured to input the bucket wheel machine operation state representation matrix into a classifier-based bucket wheel machine fault identifier to obtain a corresponding identification result, and the identification result is used to indicate whether the bucket wheel machine has a fault; The early warning prompt generation unit is configured to issue the fault early warning prompt in response to the identification result indicating that the bucket wheel machine has a fault.

8. The automatic monitoring and alarming system for bucket wheel machines according to any of claims 2, 3, 7, characterized in that, The bucket wheel machine automatic monitoring and alarm system further comprises a training module; The training module is configured to: When the operation parameter time sequence mode feature extraction module is specifically configured to input the motor temperature time sequence input vector, the motor speed time sequence input vector and the hydraulic system pressure time sequence input vector into a causal convolution network-based bucket wheel machine operation parameter time sequence mode feature extractor, the causal convolution network-based bucket wheel machine operation parameter time sequence mode feature extractor is trained; When the motor operation state feature fusion module is specifically configured to input the motor temperature time sequence mode feature vector and the motor speed time sequence mode feature vector into a motor operation state capturer based on a motor operation parameter time sequence fusion module, the motor operation state capturer based on the motor operation parameter time sequence fusion module is trained; When the identification result generation module comprises a fault judgment unit, the classifier-based bucket wheel machine fault identifier is trained.

9. The automatic monitoring and alarm system for bucket wheel machines according to claim 8, characterized in that The training module comprises: The training data acquisition unit is configured to acquire training data and a true value of whether the bucket wheel machine has a fault; wherein the training data comprises a time sequence of bucket wheel machine operation training parameters, and the bucket wheel machine operation training parameters comprise a training motor temperature value, a training motor speed value and a training hydraulic system pressure value of the bucket wheel machine; The training data regularization unit is configured to perform data regularization on the time sequence of the bucket wheel machine operation training parameters in the time dimension and the sample dimension to obtain a training motor temperature time sequence input vector, a training motor speed time sequence input vector, and a training hydraulic system pressure time sequence input vector. The training parameter time sequence mode feature extraction unit is configured to input the training motor temperature time sequence input vector, the training motor speed time sequence input vector, and the training hydraulic system pressure time sequence input vector into the bucket wheel machine operation parameter time sequence mode feature extractor based on the causal convolution network to obtain a corresponding training motor temperature time sequence mode feature vector, a training motor speed time sequence mode feature vector, and a training hydraulic system pressure time sequence mode feature vector. The training parameter time sequence fusion unit is configured to input the training motor temperature time sequence mode feature vector and the training motor speed time sequence mode feature vector into the motor operation state capturer based on the motor operation parameter time sequence fusion module to obtain a training motor temperature-speed time sequence mode fusion representing motor operation state feature vector. The training feature correlation unit is configured to perform dependency scale calculation on the training motor temperature-speed time sequence mode fusion representing motor operation state feature vector and the training hydraulic system pressure time sequence mode feature vector to obtain a training bucket wheel machine operation state representation matrix. The classification loss unit is configured to input the training bucket wheel machine operation state representation matrix into the bucket wheel machine fault identifier based on the classifier to obtain a predicted value of whether the bucket wheel machine has a fault, and determine a classification loss function value according to the predicted value of whether the bucket wheel machine has a fault and a true value of whether the bucket wheel machine has a fault. The training unit is configured to train the bucket wheel machine operation parameter time sequence mode feature extractor based on the causal convolution network, the motor operation state capturer based on the motor operation parameter time sequence fusion module, and the bucket wheel machine fault identifier based on the classifier based on the classification loss function value and through back propagation of gradient descent.

10. A bucket wheel machine automatic monitoring and alarm method based on the Internet of Things, characterized in that, The bucket wheel machine automatic monitoring and alarming method comprises: obtaining a time sequence of bucket wheel machine operation parameters, wherein the bucket wheel machine operation parameters include a motor temperature value, a motor speed value, and a hydraulic system pressure value of the bucket wheel machine; performing data regularization on the time sequence of the bucket wheel machine operation parameters in the time dimension and the sample dimension to obtain corresponding motor temperature time sequence input vectors, motor speed time sequence input vectors, and hydraulic system pressure time sequence input vectors; performing bucket wheel machine operation parameter time sequence mode feature extraction on the motor temperature time sequence input vectors, the motor speed time sequence input vectors, and the hydraulic system pressure time sequence input vectors to obtain corresponding motor temperature time sequence mode feature vectors, motor speed time sequence mode feature vectors, and hydraulic system pressure time sequence mode feature vectors; performing motor operation state capturing fusion on the motor temperature time sequence mode feature vectors and the motor speed time sequence mode feature vectors to obtain motor temperature-speed time sequence mode fusion representing motor operation state feature vectors; and performing bucket wheel machine fault identification based on the classifier to obtain a predicted value of whether the bucket wheel machine has a fault, and determining a classification loss function value according to the predicted value of whether the bucket wheel machine has a fault and a true value of whether the bucket wheel machine has a fault. The motor temperature-rotation speed time sequence mode fusion represents a motor operation state feature vector and the hydraulic system pressure time sequence mode feature vector to perform a dependent scale calculation, so as to obtain a bucket wheel machine operation state representation matrix as a bucket wheel machine operation state representation feature; Based on the bucket wheel machine operation state representation feature, a corresponding recognition result is obtained, and it is determined whether to issue a fault early warning prompt based on the recognition result.

Images