Sprinkling machine early fault identification and early warning method based on intelligent diagnosis technology

Abstract

The invention provides an early failure recognition and early warning method for a sprinkling machine based on an intelligent diagnosis technology, belongs to the technical field of computer data processing, and is used for acquiring high-quality data by adopting a multi-node synchronous acquisition and dynamic sensing strategy and realizing high-precision, self-adaptive and low-false-alarm early warning for various failures of the sprinkling machine. According to the method, multi-node synchronous acquisition and cross-modal fusion solve the problem of feature distortion and decision conflict caused by data space-time inconsistency, time alignment and feature level deep fusion are performed on high-frequency vibration signals and low-frequency attitude signals, so that composite faults and early weak fault symptoms are accurately identified, and the fault diagnosis accuracy is improved. And perception resources are allocated as required according to a transmission path of a potential fault, a deep learning model is guided to pay attention to key information, complex and changeable field working conditions and an equipment aging process are effectively adapted, and reliability and interpretability of a diagnosis decision are ensured while system power consumption and data redundancy are reduced.

Description

Technical Field

[0001] This invention relates to the field of computer data processing technology, and in particular to a method for early fault identification and warning of sprinkler irrigation machines based on intelligent diagnostic technology. Background Technology

[0002] Currently, the alarm mechanisms widely used in sprinkler system fault detection are typically based on numerical thresholds. This method sets a normal operating range for sensor values; when a sensor detects a value exceeding the threshold, the system triggers an alarm and issues a fault warning. However, this threshold-based approach has significant limitations. It usually only triggers an alarm when a fault occurs or is about to occur, failing to provide timely warnings in the early stages of a fault or when potential problems arise. For example, when a sliding sprinkler slips due to insufficient ground friction during operation, causing a positional shift, the alarm system is often not triggered immediately. Only when the shift reaches a certain level will the system recognize the fault, making preventative maintenance difficult and potentially causing missed optimal adjustment opportunities.

[0003] Neural network methods primarily focus on fault identification using a single data source, especially one-dimensional data, such as using current detection to determine the health status of a motor. While this method can effectively identify faults in certain specific situations, it cannot fully leverage the advantages of multiple data sources for comprehensive analysis. Faults in sprinkler irrigation machines may involve multiple aspects, such as the motor, transmission system, and control system, and these different data sources often have close interrelationships. Traditional neural network methods fail to integrate this diverse sensor data, resulting in insufficient dynamic adaptability during fault identification. They cannot track various changes in the equipment in complex operating environments in real time, and their robustness and accuracy are insufficient when dealing with complex faults.

[0004] Furthermore, to achieve comprehensive fault monitoring, sprinkler systems typically require the installation of multiple different types of sensors, including those for monitoring voltage, current, temperature, vibration, and position. The installation of these sensors usually necessitates invasive modifications to the sprinkler system, increasing its complexity. Additionally, installing a large number of sensors incurs high costs and engineering workload, leaving room for improvement. Summary of the Invention

[0005] This invention provides a method for early fault identification and early warning of sprinkler irrigation machines based on intelligent diagnostic technology. It uses a multi-node synchronous acquisition and dynamic perception strategy to obtain high-quality data, and combines a multi-task deep learning model guided by physical knowledge to perform intelligent diagnosis. This method can achieve high-precision, adaptive, and low-false-alarm early warning of various faults in sprinkler irrigation machines.

[0006] To achieve the above objectives, the present invention adopts the following technical solution:

[0007] Firstly, a method for early fault identification and warning of sprinkler irrigation machines based on intelligent diagnostic technology is provided. The method is applied to a warning terminal and includes:

[0008] The early warning terminal collects synchronous sensor data from multiple nodes that characterize the fault propagation model of the physical structure of the self-propelled sprinkler irrigation machine, and analyzes the synchronous sensor data based on the fault propagation model to generate a dynamic acquisition strategy.

[0009] The early warning terminal performs feature extraction and cross-modal fusion on the synchronous sensor data according to the dynamic acquisition strategy to generate a fused feature sequence.

[0010] The early warning terminal inputs the fused feature sequence into a multi-task deep learning model, processes the fused feature sequence through the multi-task deep learning model to output a fault probability, and generates an early warning signal based on the fault probability.

[0011] Optionally, the step of the early warning terminal collecting synchronous sensor data from multiple nodes includes:

[0012] Using the clock signal of the Global Navigation Satellite System as a synchronization trigger source, data acquisition devices at multiple nodes are simultaneously activated to collect raw sensor signals.

[0013] The original sensor signal is de-drifted at the edge to generate vibration data and attitude data.

[0014] The vibration and attitude data are written into a circular buffer to form traceable synchronous sensor data.

[0015] Optionally, the step of analyzing the synchronous sensor data based on the fault propagation model to generate a dynamic acquisition strategy includes:

[0016] Based on the fault propagation model, the propagation path and time relationship of the fault between nodes are determined, forming a fault transmission chain;

[0017] Real-time monitoring of the synchronization sensor data of the initial node in the fault transmission chain to identify abnormal features;

[0018] When the abnormal features are identified, the associated nodes are activated and the sampling frequency is increased according to the fault transmission chain to form the dynamic acquisition strategy. The abnormal features include abnormal slipping posture of walking deviation, deterioration of drive motor transmission due to drive failure, and abnormal nozzle blockage.

[0019] Optionally, the step of performing feature extraction and cross-modal fusion on the synchronous sensor data according to the dynamic acquisition strategy includes:

[0020] The increase in vibration amplitude and the expansion of bandwidth in the vibration signal of the drive fault analysis indicate that the vibration energy is often concentrated in the mid-to-high frequency range, accompanied by the rise of the main peak in the envelope spectrum.

[0021] For the analysis of walking deviation faults, the attitude sensor was used to detect abnormal fluctuations in angular velocity and pitch angle, whether the low-frequency energy ratio of the attitude signal increased, and whether the change in sway amplitude was symmetrical.

[0022] For nozzle clogging faults, analyze pressure changes, which generate irregular fluctuations in vibration signals, and analyze the energy drift of low-frequency and mid-frequency components in vibration signals, as well as the enhancement of sidebands and frequency drift in the spectrum.

[0023] Based on the dynamic acquisition strategy, time-domain and frequency-domain features are extracted from the synchronous sensor data;

[0024] The time-domain features and the frequency-domain features are used to construct a multi-source feature matrix;

[0025] The multi-source feature matrix is ​​subjected to dynamic time warping and channel splicing to generate the fused feature sequence.

[0026] Optionally, the step of performing dynamic time warping and channel concatenation on the multi-source feature matrix includes:

[0027] Based on the temporal relationships defined in the fault propagation model, the multi-source feature matrix is ​​aligned in the time dimension.

[0028] The aligned multi-source feature matrix is ​​input into a multi-scale convolutional coding network to extract spatial texture;

[0029] The spatial texture and the aligned multi-source feature matrix are channel-integrated and compressed to generate the fused feature sequence.

[0030] Optionally, the step of performing dynamic time warping and channel concatenation on the multi-source feature matrix includes:

[0031] Based on the temporal relationships defined in the fault propagation model, the multi-source feature matrix is ​​aligned in the time dimension.

[0032] The aligned multi-source feature matrix is ​​input into a multi-scale convolutional coding network to extract spatial texture;

[0033] The spatial texture and the aligned multi-source feature matrix are channel-integrated and compressed to generate the fused feature sequence.

[0034] Optionally, the step of inputting the fused feature sequence into the dynamic temporal attention module includes:

[0035] The fused feature sequences are linearly transformed to generate a query matrix, a key matrix, and a value matrix.

[0036] The similarity calculation between the query matrix and the key matrix is ​​guided by the fault propagation model to generate attention weights;

[0037] The attention weights are weighted and summed with the value matrix, and then processed through residual connections and layer normalization to generate the enhanced feature sequence.

[0038] Optionally, the step of generating a warning signal based on the fault probability includes:

[0039] The fault probabilities are processed by an exponentially weighted moving average to generate a smoothed fault score;

[0040] Obtain a trigger threshold and a release threshold to construct a dynamic threshold;

[0041] The smooth fault score is compared with the dynamic threshold. When the smooth fault score is consistently higher than the trigger threshold but not lower than the release threshold, the warning signal is generated.

[0042] Optionally, it also includes steps for online learning and adaptive updates:

[0043] Construct a digital shadow model and monitor the deviation between the output of the multi-task deep learning model and the simulation expectation of the digital shadow model;

[0044] When the deviation exceeds the tolerance, incremental learning is triggered;

[0045] The model parameters and dynamic threshold of the multi-task deep learning model are updated through mini-batch fine-tuning and shadow verification.

[0046] Secondly, a sprinkler irrigation machine early fault identification and warning system based on intelligent diagnostic technology is provided, the system being configured as follows:

[0047] A multi-node data acquisition module is used to acquire fault propagation models and synchronous sensor data, and generate dynamic acquisition strategies;

[0048] The feature processing module is used to process the synchronous sensor data according to the dynamic acquisition strategy to generate a fused feature sequence;

[0049] The intelligent early warning module is equipped with a multi-task deep learning model to process the fused feature sequence to output the fault probability and generate an early warning signal;

[0050] An online update module is used to monitor the performance of the intelligent early warning module and update it adaptively.

[0051] Thirdly, an electronic device is provided, comprising: a processor and a memory; the memory is used to store a computer program, which, when executed by the processor, causes the electronic device to perform the sprinkler machine early fault identification and warning method based on intelligent diagnostic technology described in the first aspect.

[0052] In one possible design, the electronic device described in the third aspect may further include a transceiver. This transceiver may be a transceiver circuit or an interface circuit. The transceiver can be used for communication between the electronic device described in the third aspect and other electronic devices.

[0053] In the embodiments of the present invention, the electronic device described in the third aspect may be a terminal, or a chip (system) or other component or assembly disposed in the terminal, or a system containing the terminal.

[0054] Fourthly, a computer-readable storage medium is provided, comprising: a computer program or instructions; when the computer program or instructions are run on a computer, the computer causes the computer to perform the method for early fault identification and early warning of sprinkler irrigation machines based on intelligent diagnostic technology as described in the first aspect.

[0055] In summary, the above methods and systems have the following technical effects:

[0056] This invention solves the problems of feature distortion and decision conflict caused by spatiotemporal inconsistency of data by multi-node synchronous acquisition and cross-modal fusion. By aligning high-frequency vibration signals with low-frequency attitude signals in time and deeply fusing them at the feature level, it can accurately identify complex faults and early weak fault signs, improve the accuracy and sensitivity of early warning, allocate sensing resources as needed according to the transmission path of potential faults, guide deep learning models to focus on key information, effectively adapt to complex and ever-changing field conditions and equipment aging processes, and ensure the reliability and interpretability of diagnostic decisions while reducing system power consumption and data redundancy. Attached Figure Description

[0057] Figure 1 This is a flowchart illustrating the method for early fault identification and early warning of sprinkler irrigation machines based on intelligent diagnostic technology provided in an embodiment of the present invention.

[0058] Figure 2 A logic block diagram of the sprinkler irrigation machine early fault identification and early warning method based on intelligent diagnostic technology provided in an embodiment of the present invention;

[0059] Figure 3 A sensor structure diagram of the sprinkler machine early fault identification and early warning method based on intelligent diagnostic technology provided in an embodiment of the present invention;

[0060] Figure 4This is a sensor installation diagram of the sprinkler machine early fault identification and early warning method based on intelligent diagnostic technology provided in an embodiment of the present invention. Detailed Implementation

[0061] The following will be combined with the appendix Figure 1-4 The technical solutions in this invention will be described below.

[0062] In this embodiment of the invention, "instruction" can include direct and indirect instructions, as well as explicit and implicit instructions. The information indicated by a certain piece of information is called the information to be instructed. In specific implementation, there are many ways to instruct the information to be instructed, such as, but not limited to, directly instructing the information to be instructed, such as the information to be instructed itself or its index. It can also indirectly instruct the information to be instructed by instructing other information, where there is a correlation between the other information and the information to be instructed. It can also instruct only a part of the information to be instructed, while the other parts are known or pre-agreed upon. For example, the instruction of specific information can be achieved by using a pre-agreed (e.g., protocol-defined) arrangement of various pieces of information, thereby reducing instruction overhead to some extent. Simultaneously, common parts of various pieces of information can be identified and uniformly indicated to reduce the instruction overhead caused by individually indicating the same information.

[0063] Furthermore, the specific indication method can also be any existing indication method, such as, but not limited to, the above-mentioned indication methods and their various combinations. Specific details of various indication methods can be found in existing technologies, and will not be elaborated upon here. As described above, for example, when multiple pieces of information of the same type need to be indicated, the indication methods for different pieces of information may differ. In specific implementation, the required indication method can be selected according to specific needs. This embodiment of the invention does not limit the selected indication method; therefore, the indication methods involved in this embodiment of the invention should be understood to cover various methods that enable the party to be indicated to obtain the information to be indicated.

[0064] It should be understood that the information to be indicated can be sent as a whole or divided into multiple sub-information messages sent separately, and the sending period and / or timing of these sub-information messages can be the same or different. The specific sending method is not limited in this embodiment of the invention. The sending period and / or timing of these sub-information messages can be predefined, for example, according to a protocol, or configured by the sending device by sending configuration information to the receiving device.

[0065] "Predefined" or "pre-configured" can be achieved by pre-saving corresponding codes, tables, or other means that can be used to indicate relevant information in the device. This embodiment of the invention does not limit the specific implementation method. "Saving" can refer to saving in one or more memories. These memories can be separate installations or integrated into the encoder, decoder, processor, or electronic device. Alternatively, some memories can be separately installed, while others are integrated into the decoder, processor, or electronic device. The type of memory can be any form of storage medium, and this embodiment of the invention does not limit this.

[0066] In the embodiments of this invention, the “protocol” may refer to a protocol family in the field of communication, a standard protocol with a similar protocol family frame structure, or a related protocol applied to a future sprinkler irrigation machine early fault identification and warning method system based on intelligent diagnostic technology. The embodiments of this invention do not specifically limit this.

[0067] In this embodiment of the invention, descriptions such as "when," "under the circumstances," "if," and "if" all refer to the device making corresponding processing under certain objective circumstances, and are not limited to a specific time. They do not require the device to make a judgment action during implementation, nor do they imply any other limitations.

[0068] In the description of the embodiments of the present invention, unless otherwise stated, " / " indicates that the objects before and after are in an "or" relationship. For example, A / B can represent A or B. "And / or" in the embodiments of the present invention is merely a description of the relationship between the related objects, indicating that three relationships can exist. For example, A and / or B can represent: A alone, A and B simultaneously, and B alone, where A and B can be singular or plural. Furthermore, in the description of the embodiments of the present invention, unless otherwise stated, "multiple" refers to two or more. "At least one of the following" or similar expressions refer to any combination of these items, including any combination of single or plural items. For example, at least one of a, b, or c can represent: a, b, c, ab, ac, bc, or abc, where a, b, and c can be single or multiple. Additionally, to facilitate a clear description of the technical solutions of the embodiments of the present invention, the terms "first" and "second" are used in the embodiments of the present invention to distinguish identical or similar items with essentially the same function and effect. Those skilled in the art will understand that the terms "first," "second," etc., do not limit the quantity or order of execution, and that "first," "second," etc., are not necessarily different. Furthermore, in the embodiments of this invention, words such as "exemplary" or "for example" are used to indicate that something is being described as an example, illustration, or description. Any embodiment or design scheme described as "exemplary" or "for example" in the embodiments of this invention should not be construed as being more preferred or advantageous than other embodiments or design schemes. Specifically, the use of words such as "exemplary" or "for example" is intended to present the relevant concepts in a concrete manner for ease of understanding.

[0069] The network architecture and business scenarios described in the embodiments of this invention are for the purpose of more clearly illustrating the technical solutions of the embodiments of this invention, and do not constitute a limitation on the technical solutions provided by the embodiments of this invention. As those skilled in the art will know, with the evolution of network architecture and the emergence of new business scenarios, the technical solutions provided by the embodiments of this invention are also applicable to similar technical problems.

[0070] The early warning terminal can be a terminal device with transmitting, receiving, and processing functions, or a chip or chip system that can be installed in the terminal device. This terminal device can also be referred to as user equipment (UE), access terminal equipment, subscriber unit, user station, mobile station (MS), mobile station, remote station, remote terminal equipment, mobile device, user terminal equipment, terminal equipment, wireless communication equipment, user agent, or user device. The terminal devices in the embodiments of this application may be mobile phones, cellular phones, smartphones, tablets, wireless data cards, personal digital assistants (PDAs), wireless modems, handsets, laptop computers, machine-type communication (MTC) terminal devices, computers with wireless transceiver capabilities, virtual reality (VR) terminal devices, augmented reality (AR) terminal devices, wireless terminal devices in industrial control, wireless terminal devices in self-driving, wireless terminal devices in remote medical care, wireless terminal devices in smart grids, wireless terminal devices in transportation safety, wireless terminal devices in smart cities, wireless terminal devices in smart homes, vehicle-mounted terminal devices, roadside units (RSUs) with terminal device functions, etc. The terminal device in this application may also be an on-board module, on-board unit, on-board component, on-board chip, or on-board unit that is built into a vehicle as one or more components or units. Alternatively, the terminal device may also be a customer-premises equipment (CPE).

[0071] The early warning terminal communicates with corresponding synchronization sensors, which can be intelligent sensors with signal receiving and transmitting capabilities, such as... Figure 3 As shown, 1 is a sensor unit (including attitude and vibration sensors), and 2 is a data acquisition unit with a data transceiver module, used to receive and transmit the data from the sensor unit to the early warning terminal.

[0072] Figure 1 This is a flowchart illustrating the method provided in an embodiment of the present invention. A method for early fault identification and warning of a sprinkler irrigation machine based on intelligent diagnostic technology, the method being applied to an early warning terminal, the method comprising:

[0073] The early warning terminal collects synchronous sensor data from multiple nodes that characterize the fault propagation model of the physical structure of the self-propelled sprinkler irrigation machine, and analyzes the synchronous sensor data based on the fault propagation model to generate a dynamic acquisition strategy.

[0074] The early warning terminal performs feature extraction and cross-modal fusion on the synchronous sensor data according to the dynamic acquisition strategy to generate a fused feature sequence.

[0075] The early warning terminal inputs the fused feature sequence into a multi-task deep learning model, processes the fused feature sequence through the multi-task deep learning model to output a fault probability, and generates an early warning signal based on the fault probability.

[0076] In this multi-task deep learning model, spatial feature encoding uses a multi-scale convolutional coding network as a shared backbone. The coding network consists of several parallel convolutional branches, each configured with a different receptive field and stride to capture multi-scale structures ranging from mid-to-high frequency modes such as tooth meshing and bearing impact to low frequency modes such as slow posture fluctuations. The outputs of each branch are channel-joined and pointwise convolutionally compressed, and then residual connections and random dropout are superimposed to suppress overfitting. When the input is a time-frequency map, the convolution is performed on the time-frequency plane, thereby simultaneously modeling spectral texture and temporal evolution.

[0077] To enhance the discriminability of time dependencies and reduce the impact of irrelevant perturbations, a dynamic temporal attention module is set up before entering the temporal unit. This module uses the encoded feature sequence as the key, value, and query to calculate multi-head attention weights, enabling the model to assign higher weights to segments at key moments (such as 1× order enhancement, GMF sideband appearance, low-frequency energy surge, or frequency centroid drift) based on the context. The attention output is stabilized through gating residuals and layer normalization training, and the attention weights are used as a side-channel supervision signal to explain the model's decisions.

[0078] The temporal modeling part adopts a bidirectional long short-term memory network (Bi-LSTM). Compared with simple recurrent units, the input gate, forget gate and output gate structure of LSTM can effectively alleviate the gradient vanishing problem in long sequences, thus preserving both the low-frequency trend caused by slow-changing conditions and the transient mode caused by intermittent impacts. The model uses two to three stacked Bi-LSTM layers to perform forward and backward modeling on the attention-enhanced feature sequence, and finally performs pooling or extracts the terminal state in the time dimension to form a temporal representation vector for the current window.

[0079] At the task level, the model has multiple parallel decision branches, corresponding to fault types such as drive failure and nozzle blockage. Each branch consists of one or two layers of lightweight fully connected networks and a Sigmoid or Softmax output, providing the fault probability or anomaly score for that window. To adapt to early warning requirements, the model adds a temporal smoothing and hysteresis mechanism to the output of each branch: the scores of multiple consecutive windows are compared with an adaptive threshold after being exponentially weighted by a moving average. When the score continuously exceeds the trigger threshold but does not fall below the release threshold, it is determined to be a valid alarm, thereby reducing false triggers caused by jitter.

[0080] Multi-task training employs a Dynamic Weighted Average (DWA) strategy to adaptively balance the losses of each task, avoiding performance degradation caused by single-task-dominated optimization. The overall loss is composed of the binary cross-entropy or focus loss of each branch, attention sparsity regularization, and temporal consistency constraints. Among these, the consistency constraints encourage adjacent window predictions to remain smooth when there are no drastic changes in operating conditions, and allow rapid transitions at fault edges. During training, online data augmentation is also employed to improve generalization ability, including amplitude random scaling, small frequency drift, band-limited noise injection, and random masking of missing sample segments to simulate installation differences and communication jitter in the field.

[0081] In terms of multi-node fusion, the model utilizes the unified timestamp of GNSS to stitch together inputs from windows of the same time from different installation points along the channel dimension to share the backbone. Alternatively, attention-based cross-node aggregation can be set after the time-series layer, enabling the model to learn the spatial association patterns between the drive end and the cantilever center. When operating parameters such as rotational speed are available, the system can also perform order resampling during the preprocessing stage, standardizing the frequency axis to the order domain to improve the alignment robustness of feature lines (such as 1× and GMF) under variable speed conditions.

[0082] Optionally, the step of the early warning terminal collecting synchronous sensor data from multiple nodes includes:

[0083] Using the clock signal of the Global Navigation Satellite System as a synchronization trigger source, data acquisition devices at multiple nodes are simultaneously activated to collect raw sensor signals.

[0084] The original sensor signal is de-drifted at the edge to generate vibration data and attitude data.

[0085] The vibration and attitude data are written into a circular buffer to form traceable synchronous sensor data.

[0086] Optionally, the step of analyzing the synchronous sensor data based on the fault propagation model to generate a dynamic acquisition strategy includes:

[0087] Based on the fault propagation model, the propagation path and time relationship of the fault between nodes are determined, forming a fault transmission chain;

[0088] Real-time monitoring of the synchronization sensor data of the initial node in the fault transmission chain to identify abnormal features;

[0089] When the abnormal features are identified, the associated nodes are activated and the sampling frequency is increased according to the fault transmission chain to form the dynamic acquisition strategy. The abnormal features include abnormal slipping posture of walking deviation, deterioration of drive motor transmission due to drive failure, and abnormal nozzle blockage.

[0090] Optionally, the step of performing feature extraction and cross-modal fusion on the synchronous sensor data according to the dynamic acquisition strategy includes:

[0091] The increase in vibration amplitude and the expansion of bandwidth in the vibration signal of the drive fault analysis indicate that the vibration energy is often concentrated in the mid-to-high frequency range, accompanied by the rise of the main peak in the envelope spectrum.

[0092] For the analysis of walking deviation faults, the attitude sensor was used to detect abnormal fluctuations in angular velocity and pitch angle, whether the low-frequency energy ratio of the attitude signal increased, and whether the change in sway amplitude was symmetrical.

[0093] For nozzle clogging faults, analyze pressure changes, which generate irregular fluctuations in vibration signals, and analyze the energy drift of low-frequency and mid-frequency components in vibration signals, as well as the enhancement of sidebands and frequency drift in the spectrum.

[0094] Based on the dynamic acquisition strategy, time-domain and frequency-domain features are extracted from the synchronous sensor data;

[0095] The time-domain features and the frequency-domain features are used to construct a multi-source feature matrix;

[0096] The multi-source feature matrix is ​​subjected to dynamic time warping and channel splicing to generate the fused feature sequence.

[0097] Specifically, in order to accurately identify these faults, feature extraction will comprehensively analyze the vibration and attitude signals from multiple perspectives, including the time domain, frequency domain, and time-frequency domain, and construct a sliding window (length T, step size Δ) to calculate features window by window.

[0098] Temporal feature extraction primarily focuses on the amplitude fluctuations and temporal variations of the signal. Temporal features (such as...)

[0099] Variance, kurtosis, and mean, etc. These time-domain characteristics can be represented as follows:

[0100]

[0101] Where, xi Let μ be the i-th sampling point, μ be the mean, and σ be the standard deviation.

[0102] Frequency domain feature extraction is crucial for detecting nozzle blockage, as it reveals the energy distribution of vibration signals across different frequency ranges. This is especially important because blockage can weaken the signal strength. Frequency domain features (such as spectrum, power spectral density, frequency center, and spectral flatness) can be extracted using methods like Fourier transform and by calculating relevant statistics to capture the frequency components of the signal.

[0103]

[0104] Where: X(f): is the spectral value at frequency f.

[0105] P(f): is the power spectral density at frequency f.

[0106] Csc: is the frequency center, reflecting the frequency "center of gravity" of the signal.

[0107] Fsf: Spectral flatness, which measures the flatness of the signal spectrum.

[0108] Furthermore, the construction of the feature matrix includes:

[0109] After extracting the time-domain and frequency-domain features, they are constructed into a feature matrix for subsequent model training and analysis. Each row of the feature matrix represents a pipeline leak signal sample, and each column represents a feature. The feature matrix includes the following:

[0110] Time-domain characteristics: such as the signal's mean, variance, kurtosis, etc.

[0111] Frequency domain characteristics: such as the dominant frequency of the spectrum, power spectral density, frequency center, etc.

[0112] Other statistics include energy distribution and spectral flatness of the spectrum.

[0113] Finally, all extracted features will be combined into a matrix, represented as follows:

[0114] f i = [x1,x2,…,x n ]

[0115] Optionally, the step of performing dynamic time warping and channel concatenation on the multi-source feature matrix includes:

[0116] Based on the temporal relationships defined in the fault propagation model, the multi-source feature matrix is ​​aligned in the time dimension.

[0117] The aligned multi-source feature matrix is ​​input into a multi-scale convolutional coding network to extract spatial texture;

[0118] The spatial texture and the aligned multi-source feature matrix are channel-integrated and compressed to generate the fused feature sequence.

[0119] Optionally, the step of performing dynamic time warping and channel concatenation on the multi-source feature matrix includes:

[0120] Based on the temporal relationships defined in the fault propagation model, the multi-source feature matrix is ​​aligned in the time dimension.

[0121] The aligned multi-source feature matrix is ​​input into a multi-scale convolutional coding network to extract spatial texture;

[0122] The spatial texture and the aligned multi-source feature matrix are channel-integrated and compressed to generate the fused feature sequence.

[0123] Optionally, the step of inputting the fused feature sequence into the dynamic temporal attention module includes:

[0124] The fused feature sequences are linearly transformed to generate a query matrix, a key matrix, and a value matrix.

[0125] The similarity calculation between the query matrix and the key matrix is ​​guided by the fault propagation model to generate attention weights;

[0126] The attention weights are weighted and summed with the value matrix, and then processed through residual connections and layer normalization to generate the enhanced feature sequence.

[0127] Optionally, the step of generating a warning signal based on the fault probability includes:

[0128] The fault probabilities are processed by an exponentially weighted moving average to generate a smoothed fault score;

[0129] Obtain a trigger threshold and a release threshold to construct a dynamic threshold;

[0130] The smooth fault score is compared with the dynamic threshold. When the smooth fault score is consistently higher than the trigger threshold but not lower than the release threshold, the warning signal is generated.

[0131] Optionally, it also includes steps for online learning and adaptive updates:

[0132] Construct a digital shadow model and monitor the deviation between the output of the multi-task deep learning model and the simulation expectation of the digital shadow model;

[0133] When the deviation exceeds the tolerance, incremental learning is triggered;

[0134] The model parameters and dynamic threshold of the multi-task deep learning model are updated through mini-batch fine-tuning and shadow verification.

[0135] Secondly, a sprinkler irrigation machine early fault identification and warning system based on intelligent diagnostic technology is provided, the system being configured as follows:

[0136] A multi-node data acquisition module is used to acquire fault propagation models and synchronous sensor data, and generate dynamic acquisition strategies;

[0137] The feature processing module is used to process the synchronous sensor data according to the dynamic acquisition strategy to generate a fused feature sequence;

[0138] The intelligent early warning module is equipped with a multi-task deep learning model to process the fused feature sequence to output the fault probability and generate an early warning signal;

[0139] An online update module is used to monitor the performance of the intelligent early warning module and update it adaptively.

[0140] The following is a detailed introduction to the various components of the electronic device:

[0141] In this context, the processor is the control center of the electronic device. It can be a single processor or a collective term for multiple processing elements. For example, a processor can be one or more central processing units (CPUs), an application-specific integrated circuit (ASIC), or one or more integrated circuits configured to implement embodiments of the present invention, such as one or more digital signal processors (DSPs), or one or more field-programmable gate arrays (FPGAs).

[0142] Alternatively, the processor can perform various functions of the electronic device by running or executing software programs stored in memory and calling data stored in memory, such as executing the multi-node synchronous monitoring system for early fault alarm of sprinkler irrigation machines as shown above.

[0143] In a specific implementation, as one example, the processor may include one or more CPUs, such as CPU0 and CPU1.

[0144] In a specific implementation, as one example, the electronic device may also include multiple processors. Each of these processors may be a single-core processor (single-CPU) or a multi-core processor (multi-CPU). Here, a processor may refer to one or more devices, circuits, and / or processing cores for processing data (e.g., computer program instructions).

[0145] The memory is used to store the software program that executes the solution of the present invention, and the execution is controlled by the processor. The specific implementation method can be referred to the above method embodiment, and will not be repeated here.

[0146] Optionally, the memory can be read-only memory (ROM) or other types of static storage devices capable of storing static information and instructions, random access memory (RAM) or other types of dynamic storage devices capable of storing information and instructions, or electrically erasable programmable read-only memory (EEPROM), compact disc read-only memory (CD-ROM) or other optical disc storage, optical disc storage (including compressed optical discs, laser discs, optical discs, digital universal optical discs, Blu-ray discs, etc.), magnetic disk storage media or other magnetic storage devices, or any other medium capable of carrying or storing desired program code in the form of instructions or data structures and accessible by a computer, but not limited thereto. The memory can be integrated with the processor or exist independently and coupled to the processor through an interface circuit of an electronic device; this embodiment of the invention does not specifically limit this.

[0147] A transceiver is used for communication with other electronic devices. For example, if the electronic device is a terminal, the transceiver can be used to communicate with a network device or with another terminal device. Similarly, if the electronic device is a network device, the transceiver can be used to communicate with a terminal or with another network device.

[0148] Optionally, the transceiver may include a receiver and a transmitter. The receiver is used to implement the receiving function, and the transmitter is used to implement the sending function.

[0149] Optionally, the transceiver can be integrated with the processor or exist independently and coupled to the processor through the interface circuit of the electronic device. This embodiment of the invention does not specifically limit this.

[0150] It is understood that the structure of an electronic device does not constitute a limitation on the electronic device. An actual electronic device may include more or fewer components than shown in the figure, or combine certain components, or have different component arrangements.

[0151] Furthermore, the technical effects of the electronic equipment can be referred to the technical effects of the multi-node synchronous monitoring system for early fault alarm of sprinkler irrigation machines described in the above method embodiments, and will not be repeated here.

[0152] It should be understood that the processor in the embodiments of the present invention can be a central processing unit (CPU), or it can be other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor can be a microprocessor or any conventional processor.

[0153] It should also be understood that the memory in the embodiments of the present invention can be volatile memory or non-volatile memory, or may include both volatile and non-volatile memory. The non-volatile memory can be read-only memory (ROM), programmable read-only memory (PROM), erasable programmable read-only memory (EPROM), electrically erasable programmable read-only memory (EEPROM), or flash memory. The volatile memory can be random access memory (RAM), which is used as an external cache. By way of example, but not limitation, many forms of random access memory (RAM) are available, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate synchronous DRAM (DDR SDRAM), enhanced synchronous DRAM (ESDRAM), synchronous linked DRAM (SLDRAM), and direct rambus RAM (DR RAM).

[0154] The above embodiments can be implemented, in whole or in part, by software, hardware (such as circuits), firmware, or any other combination thereof. When implemented using software, the above embodiments can be implemented, in whole or in part, as a computer program product. The computer program product includes one or more computer instructions or computer programs. When the computer instructions or computer programs are loaded or executed on a computer, all or part of the processes or functions described in the embodiments of the present invention are generated. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, the computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via wired (e.g., infrared, wireless, microwave, etc.) means. The computer-readable storage medium can be any available medium that a computer can access or a data storage device such as a server or data center that includes one or more sets of available media. The available medium can be a magnetic medium (e.g., floppy disk, hard disk, magnetic tape), an optical medium (e.g., DVD), or a semiconductor medium. A semiconductor medium can be a solid-state drive.

[0155] It should be understood that the term "and / or" in this article is merely a description of the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A existing alone, A and B existing simultaneously, or B existing alone. A and B can be singular or plural. Additionally, the character " / " in this article generally indicates an "or" relationship between the preceding and following related objects, but it can also represent an "and / or" relationship. Please refer to the context for a more accurate understanding.

[0156] In this invention, "at least one" means one or more, and "more than one" means two or more. "At least one of the following" or similar expressions refer to any combination of these items, including any combination of a single item or a plurality of items. For example, at least one of a, b, or c can represent: a, b, c, ab, ac, bc, or abc, where a, b, and c can be a single item or multiple items.

[0157] It should be understood that, in various embodiments of the present invention, the sequence number of each process does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present invention.

[0158] Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementations should not be considered beyond the scope of this invention.

[0159] Those skilled in the art will understand that, for the sake of convenience and brevity, the specific working processes of the systems, devices, and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0160] In the embodiments provided by this invention, it should be understood that the disclosed systems, apparatuses, and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between devices or units may be electrical, mechanical, or other forms.

[0161] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0162] In addition, the functional units in the various embodiments of the present invention can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit.

[0163] If the aforementioned functions are implemented as 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 a part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0164] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in the present invention should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. A method for early fault identification and warning of sprinkler irrigation machines based on intelligent diagnostic technology, characterized in that, The method is applied to an early warning terminal, and the method includes: The early warning terminal collects synchronous sensor data from multiple nodes that characterize the fault propagation model of the physical structure of the self-propelled sprinkler irrigation machine, and analyzes the synchronous sensor data based on the fault propagation model to generate a dynamic acquisition strategy. The early warning terminal performs feature extraction and cross-modal fusion on the synchronous sensor data according to the dynamic acquisition strategy to generate a fused feature sequence. The early warning terminal inputs the fused feature sequence into a multi-task deep learning model, processes the fused feature sequence through the multi-task deep learning model to output a fault probability, and generates an early warning signal based on the fault probability.

2. The method for early fault identification and early warning of sprinkler irrigation machines based on intelligent diagnostic technology according to claim 1, characterized in that, The steps for the early warning terminal to collect synchronous sensor data from multiple nodes include: Using the clock signal of the Global Navigation Satellite System as a synchronization trigger source, data acquisition devices at multiple nodes are simultaneously activated to collect raw sensor signals. The original sensor signal is de-drifted at the edge to generate vibration data and attitude data. The vibration and attitude data are written into a circular buffer to form traceable synchronous sensor data.

3. The method for early fault identification and early warning of sprinkler irrigation machines based on intelligent diagnostic technology according to claim 2, characterized in that, The step of analyzing the synchronous sensor data based on the fault propagation model to generate a dynamic acquisition strategy includes: Based on the fault propagation model, the propagation path and time relationship of the fault between nodes are determined, forming a fault transmission chain; Real-time monitoring of the synchronization sensor data of the initial node in the fault transmission chain to identify abnormal features; When the abnormal features are identified, the associated nodes are activated and the sampling frequency is increased according to the fault transmission chain to form the dynamic acquisition strategy. The abnormal features include abnormal slipping posture of walking deviation, deterioration of drive motor transmission due to drive failure, and abnormal nozzle blockage.

4. The method for early fault identification and early warning of sprinkler irrigation machines based on intelligent diagnostic technology according to claim 3, characterized in that, The step of performing feature extraction and cross-modal fusion of the synchronous sensor data according to the dynamic acquisition strategy includes: The increase in vibration amplitude and the expansion of bandwidth in the vibration signal of the drive fault analysis indicate that the vibration energy is often concentrated in the mid-to-high frequency range, accompanied by the rise of the main peak in the envelope spectrum. For the analysis of walking deviation faults, the attitude sensor was used to detect abnormal fluctuations in angular velocity and pitch angle, whether the low-frequency energy ratio of the attitude signal increased, and whether the change in sway amplitude was symmetrical. For nozzle clogging faults, analyze pressure changes, which generate irregular fluctuations in vibration signals, and analyze the energy drift of low-frequency and mid-frequency components in vibration signals, as well as the enhancement of sidebands and frequency drift in the spectrum. Based on the dynamic acquisition strategy, time-domain and frequency-domain features are extracted from the synchronous sensor data; The time-domain features and the frequency-domain features are used to construct a multi-source feature matrix; The multi-source feature matrix is ​​subjected to dynamic time warping and channel splicing to generate the fused feature sequence.

5. The method for early fault identification and early warning of sprinkler irrigation machines based on intelligent diagnostic technology according to claim 2, characterized in that, The steps of performing dynamic time warping and channel concatenation on the multi-source feature matrix include: Based on the temporal relationships defined in the fault propagation model, the multi-source feature matrix is ​​aligned in the time dimension. The aligned multi-source feature matrix is ​​input into a multi-scale convolutional coding network to extract spatial texture; The spatial texture and the aligned multi-source feature matrix are channel-integrated and compressed to generate the fused feature sequence.

6. The method for early fault identification and early warning of sprinkler irrigation machines based on intelligent diagnostic technology according to claim 1, characterized in that, The step of inputting the fused feature sequence into a multi-task deep learning model includes: The fused feature sequence is input into the dynamic temporal attention module to calculate attention weights and generate an enhanced feature sequence. The enhanced feature sequence is input into a bidirectional long short-term memory network for temporal modeling to generate a temporal representation vector. The time-series representation vector is input to multiple parallel task branches to output the failure probability.

7. The method for early fault identification and early warning of sprinkler irrigation machines based on intelligent diagnostic technology according to claim 6, characterized in that, The step of inputting the fused feature sequence into the dynamic temporal attention module includes: The fused feature sequences are linearly transformed to generate a query matrix, a key matrix, and a value matrix. The similarity calculation between the query matrix and the key matrix is ​​guided by the fault propagation model to generate attention weights; The attention weights are weighted and summed with the value matrix, and then processed through residual connections and layer normalization to generate the enhanced feature sequence.

8. The method for early fault identification and early warning of sprinkler irrigation machines based on intelligent diagnostic technology according to claim 7, characterized in that, The steps for generating a warning signal based on the fault probability include: The fault probabilities are processed by an exponentially weighted moving average to generate a smoothed fault score; Obtain a trigger threshold and a release threshold to construct a dynamic threshold; The smooth fault score is compared with the dynamic threshold. When the smooth fault score is consistently higher than the trigger threshold but not lower than the release threshold, the warning signal is generated.

9. The method for early fault identification and early warning of sprinkler irrigation machines based on intelligent diagnostic technology according to claim 7, characterized in that, It also includes steps for online learning and adaptive updates: Construct a digital shadow model and monitor the deviation between the output of the multi-task deep learning model and the simulation expectation of the digital shadow model; When the deviation exceeds the tolerance, incremental learning is triggered; The model parameters and dynamic threshold of the multi-task deep learning model are updated through mini-batch fine-tuning and shadow verification.

10. The method for early fault identification and early warning of sprinkler irrigation machines based on intelligent diagnostic technology according to any one of claims 1-9, characterized in that, It also includes an early fault identification and warning system for sprinkler irrigation machines based on intelligent diagnostic technology, the system being configured as follows: A multi-node data acquisition module is used to acquire fault propagation models and synchronous sensor data, and generate dynamic acquisition strategies; The feature processing module is used to process the synchronous sensor data according to the dynamic acquisition strategy to generate a fused feature sequence; The intelligent early warning module is equipped with a multi-task deep learning model to process the fused feature sequence to output the fault probability and generate an early warning signal; An online update module is used to monitor the performance of the intelligent early warning module and update it adaptively.

Images