Grain transportation chain bucket intelligent monitoring and conveying method, device and medium

Abstract

The application discloses a kind of grain transport chain bucket intelligent monitoring conveying method, device and medium, comprising: in response to conveying start instruction set, the initial configuration parameter of grain transport chain bucket is acquired, and monitoring mode is determined based on initial configuration parameter, wherein, monitoring mode includes: routine monitoring mode and depth monitoring mode;In response to monitoring mode, monitoring instruction set is generated, monitoring instruction set is used to control the multi-dimensional real-time operation data and grain quality data of multiple sensors deployed in the key position of conveying system are synchronously collected, and the data collected is preprocessed to extract multiple abnormal features;In response to the multiple abnormal features extracted, multiple abnormal features are input into the multimodal fusion intelligent diagnosis model trained in advance, obtain diagnosis result, and generate control instruction set set, control instruction set set is used to adjust the operating parameter of grain transport chain bucket or execute fault linkage disposal.

Description

Technical Field

[0001] This invention discloses an intelligent monitoring and conveying method, device, and medium for grain transport chain buckets, belonging to the field of grain transport chain bucket technology. Background Technology

[0002] Grain conveyor chains are among the most widely used continuous conveying equipment in grain storage, processing, and logistics. They achieve continuous grain transport by driving multiple hoppers in a closed chain. Traditional grain conveyor chain systems generally rely on manual inspection for condition monitoring and fault diagnosis, which has significant time intervals and cannot monitor sudden faults such as hopper jamming, chain breakage, and hopper damage in real time. This can easily lead to large amounts of grain spillage, serious equipment damage, and even safety accidents. At the same time, manual inspections rely on the operator's experience and are difficult to accurately identify hidden problems such as grain clumping, foreign object contamination, and abnormal temperature and humidity inside the hoppers. In addition, traditional conveying systems operate at a fixed speed and cannot adaptively adjust according to grain flow, equipment status, and downstream process requirements, resulting in energy waste and low conveying efficiency. Furthermore, they lack full-process data traceability capabilities and cannot perform long-term statistical analysis and predictive maintenance of quality changes and equipment operating status during grain transport.

[0003] There is currently no effective solution to the above problems. Summary of the Invention

[0004] This invention provides a method, device, and medium for intelligent monitoring and conveying of grain transport chain buckets, which at least solves the technical problem that existing grain transport chain bucket conveying systems cannot monitor the operation status of the chain buckets and the quality of grain in real time, comprehensively, and accurately due to reliance on manual inspection.

[0005] According to one aspect of the present invention, in order to achieve the above-mentioned objective, a method for intelligent monitoring and conveying of grain transport chain buckets is provided, comprising:

[0006] In response to the delivery start command set, the initial configuration parameters of the grain transport chain bucket are obtained, and the monitoring mode is determined based on the initial configuration parameters. The monitoring modes include: regular monitoring mode and deep monitoring mode.

[0007] In response to the monitoring mode, a monitoring instruction set is generated. The monitoring instruction set is used to control multiple sensors deployed at key locations in the conveying system to simultaneously collect multi-dimensional real-time operating data and grain quality data, and to preprocess the collected data to extract various abnormal features.

[0008] In response to the extracted multiple abnormal features, the multiple abnormal features are input into a pre-trained multimodal fusion intelligent diagnostic model to obtain diagnostic results and generate a control instruction set. The control instruction set is used to adjust the operating parameters of the grain transport chain bucket or to perform fault linkage handling.

[0009] Furthermore, based on the initial configuration parameters, the monitoring mode is determined, including:

[0010] Based on the historical operation database, the initial configuration parameters are matched with the historical operation database to obtain the matching results, which include: successful matching and unsuccessful matching.

[0011] In response to a successful match result, the monitoring mode is set to the normal monitoring mode.

[0012] If the matching result is "unsuccessful", the monitoring mode is set to deep monitoring mode.

[0013] Furthermore, in response to the monitoring mode being set to the normal monitoring mode, a monitoring instruction set is generated, including:

[0014] Based on the initial configuration parameters, determine the standard sensor sampling frequency and characteristic parameter thresholds corresponding to the grain transport chain bucket;

[0015] A monitoring instruction set is generated based on the standard sensor sampling frequency and characteristic parameter thresholds.

[0016] Furthermore, in response to the monitoring mode being set to deep monitoring mode, a monitoring instruction set is generated, including:

[0017] Acquire several image data of the grain transport chain bucket, and based on the image data, determine the bucket's appearance integrity information and grain clumping information;

[0018] Based on information about the integrity of the hopper's appearance and information about grain clumping, a high-density sampling scheme was determined.

[0019] A monitoring instruction set is generated based on a high-density sampling scheme.

[0020] Furthermore, based on several image data, information on the integrity of the hopper's appearance and information on grain clumping are determined, including:

[0021] Several image data are filtered to obtain several filtered image data. The filtering processes include: median filtering, Gaussian filtering, and guided filtering.

[0022] Based on several filtered image data, an adaptive histogram equalization method is used to obtain several processed image data.

[0023] Based on several processed image data, several initial hopper breakage feature data and initial agglomeration feature data are determined;

[0024] Based on several initial hopper damage characteristic data and initial agglomeration characteristic data, the majority voting calculation method is used to determine the hopper appearance integrity information and grain agglomeration information.

[0025] Furthermore, various abnormal features are extracted, including:

[0026] Based on the changing trends of weight data, features representing empty, full, overloaded, and grain spillage in the hopper are extracted.

[0027] Based on visual image data, features representing hopper damage, deformation, missing parts, grain clumping, and foreign matter contamination are extracted.

[0028] Based on the spectral analysis of vibration data, features characterizing chain wear, chain bucket jamming, and drive wheel abnormalities are extracted.

[0029] Based on the temperature and humidity data change curves, early characteristics representing grain heating and mold growth are extracted.

[0030] Furthermore, in response to the diagnostic results, a set of control instructions is generated, including:

[0031] The operating speed of the chain bucket is automatically adjusted based on real-time collected grain flow data.

[0032] Based on real-time collected chain tension data, the tensioning device is controlled to automatically adjust the chain tension.

[0033] When minor damage to the hopper is diagnosed, a set of control instructions is generated to reduce the operating speed and mark the location of the hopper.

[0034] When chain bucket jamming or chain breakage is diagnosed, a set of control commands is generated to immediately stop the conveying system and close the feed inlet valve.

[0035] When an abnormal increase in grain temperature and humidity is diagnosed, a set of control instructions is generated to reduce the chain bucket operating speed and activate the ventilation device.

[0036] Furthermore, in response to the diagnostic results, the generation of a control instruction set also includes:

[0037] Based on the fixed pitch and running speed of the chain bucket, a mapping relationship between the theoretical passage time and the conveying position of each bucket is established; the weight values ​​measured in real time for each bucket are arranged in time series to form a three-dimensional map; when the weight value of a bucket at a certain position shows a continuous and regular abnormal decrease, and is not related to the weight values ​​of the upstream and downstream buckets at that position, it is determined that the corresponding bucket has leaked, and the number of the leaking bucket is accurately located.

[0038] Extract the characteristic frequency of the drive wheel and the amplitude ratio of its harmonics; calculate the estimated wear of the current drive wheel teeth based on a pre-established mapping database; when the estimated wear exceeds a preset threshold, issue a warning message to replace the drive wheel.

[0039] According to one embodiment of the present invention, a smart monitoring and conveying device for grain transport chain hoppers is also provided, comprising:

[0040] The acquisition module is used to acquire the initial configuration parameters of the grain transport chain bucket in response to the transport start command set, and determine the monitoring mode based on the initial configuration parameters. The monitoring modes include: regular monitoring mode and deep monitoring mode.

[0041] The generation module is used to generate a monitoring instruction set in response to the monitoring mode. The monitoring instruction set is used to control multiple sensors deployed at key locations in the conveying system to simultaneously collect multi-dimensional real-time operating data and grain quality data, and to preprocess the collected data to extract various abnormal features.

[0042] The diagnostic module is used to respond to the extracted multiple abnormal features, input the multiple abnormal features into a pre-trained multimodal fusion intelligent diagnostic model, obtain diagnostic results, and generate a control instruction set. The control instruction set is used to adjust the operating parameters of the grain transport chain bucket or to perform fault linkage handling.

[0043] According to another aspect of the present invention, a computer-readable storage medium is also provided, the computer-readable storage medium including a stored executable program, wherein, when the executable program is executed, it controls the device where the computer-readable storage medium is located to perform the methods of various embodiments of the present invention.

[0044] In this embodiment of the invention, by responding to the conveying start command to obtain initial configuration parameters and intelligently selecting a conventional monitoring mode or a deep monitoring mode, an adaptive monitoring strategy for different working conditions is realized; in response to the monitoring mode, a monitoring instruction set is generated to control multiple sensors to simultaneously collect multi-dimensional data and perform preprocessing to extract abnormal features, thereby realizing comprehensive real-time perception of the chain bucket's operating status and grain quality; in response to the extracted multiple abnormal features, a pre-trained multi-modal fusion intelligent diagnostic model is used for accurate diagnosis, and a control instruction set is generated to adjust operating parameters or perform fault linkage handling, thereby effectively improving the safety, reliability, and efficiency of grain conveying, reducing grain loss and equipment maintenance costs, and realizing full-process data traceability and predictive maintenance. Attached Figure Description

[0045] The accompanying drawings, which are included to provide a further understanding of the invention and form part of this application, illustrate exemplary embodiments of the invention and, together with their description, serve to explain the invention and do not constitute an undue limitation thereof. In the drawings:

[0046] Figure 1 This is a flowchart of a method for intelligent monitoring and conveying of grain transport buckets according to one embodiment of the present invention;

[0047] Figure 2 This is a structural block diagram of an intelligent monitoring and conveying device for a grain transport chain according to one embodiment of the present invention. Detailed Implementation

[0048] To enable those skilled in the art to better understand the present invention, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of the present invention.

[0049] It should be noted that the terms "first," "second," etc., in the specification, claims, and accompanying drawings of this invention are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of the invention described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover a non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0050] According to an embodiment of the present invention, an embodiment of a method for intelligent monitoring and conveying of grain transport chain buckets is provided. It should be noted that the steps shown in the flowchart in the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions. Furthermore, although a logical order is shown in the flowchart, in some cases, the steps shown or described may be executed in a different order than that shown here.

[0051] This method embodiment can be executed in an electronic device or similar computing device that includes a memory and a processor. Taking operation on a vehicle terminal as an example, the vehicle terminal may include one or more processors (processors may include, but are not limited to, central processing units (CPUs), graphics processing units (GPUs), digital signal processing (DSP) chips, microcontroller units (MCUs), field-programmable gate arrays (FPGAs), neural network processors (NPUs), tensor processors (TPUs), artificial intelligence (AI) type processors, etc.) and a memory for storing data. Optionally, the vehicle terminal may also include transmission devices, input / output devices, and display devices for communication functions. Those skilled in the art will understand that the above structural description is merely illustrative and does not limit the structure of the vehicle terminal. For example, the vehicle terminal may include more or fewer components than described above, or have a different configuration than described above.

[0052] The memory can be used to store computer programs, such as application software programs and modules, like the computer program corresponding to the intelligent monitoring and conveying method for grain transport chain buckets in this embodiment of the invention. The processor executes various functional applications and data processing by running the computer program stored in the memory, thereby realizing the aforementioned intelligent monitoring and conveying method for grain transport chain buckets. The memory may include high-speed random access memory, and may also include non-volatile memory, such as one or more magnetic storage devices, flash memory, or other non-volatile solid-state memory. In some instances, the memory may further include memory remotely located relative to the processor, and these remote memories can be connected to a mobile terminal via a network. Examples of such networks include, but are not limited to, the Internet, corporate intranets, local area networks, mobile communication networks, and combinations thereof.

[0053] The transmission device is used to receive or send data via a network. Specific examples of the network mentioned above may include a wireless network provided by the mobile terminal's communication provider. In one example, the transmission device includes a Network Interface Controller (NIC), which can connect to other network devices via a base station to communicate with the Internet. In another example, the transmission device may be a Radio Frequency (RF) module, used for wireless communication with the Internet.

[0054] Display devices can be, for example, touchscreen liquid crystal displays (LCDs) and touch displays (also referred to as "touchscreens" or "touch displays"). The LCD allows users to interact with the user interface of the mobile terminal. In some embodiments, the mobile terminal has a graphical user interface (GUI), which allows users to interact with the GUI through finger contact and / or gestures on a touch-sensitive surface. Optional human-computer interaction functions include: creating web pages, drawing, word processing, creating electronic documents, playing games, video conferencing, instant messaging, sending and receiving emails, call interfaces, playing digital video, playing digital music, and / or web browsing, etc. An executable instruction set for performing the above human-computer interaction functions is configured / stored in one or more processor-executable computer program products or readable storage media.

[0055] Figure 1 According to one embodiment of the present invention, a method for intelligent monitoring and conveying of grain transport chain buckets includes:

[0056] Step S110: In response to the conveying start command set, the initial configuration parameters of the grain transport chain bucket are obtained, and the monitoring mode is determined based on the initial configuration parameters. The monitoring modes include: regular monitoring mode and deep monitoring mode, the details of which are as follows:

[0057] In step S110, when the grain transport chain bucket conveyor is powered on or receives a conveying task instruction from the central dispatch layer, this method first responds to the conveying start instruction set. The conveying start instruction set may include a start button trigger signal, a chain start signal from upstream equipment, or a remote start instruction issued by the operator on the monitoring interface. After capturing the instruction set, the device's built-in controller immediately performs the configuration acquisition operation during the initialization phase.

[0058] Specifically, the controller reads the initial configuration parameters of the grain transport chain bucket from local storage or a cloud database. These initial configuration parameters include the model and size of the chain bucket, the designed operating speed range of the chain, the layout information of the conveying route, and the type and expected conveying volume of grain for this transport task. In some embodiments, the controller can also automatically identify the grain type and associate it with corresponding standard parameter thresholds by scanning the barcode or RFID tag on the grain packaging bag at the feed inlet. For example, the normal temperature and humidity range for wheat is 15 to 25 degrees Celsius and 12% to 14% humidity.

[0059] After reading the initial configuration parameters, the controller further determines the current monitoring mode based on these parameters. In this invention, the monitoring mode is divided into two types: conventional monitoring mode and deep monitoring mode. Conventional monitoring mode is suitable for situations where the same type of grain has been transported multiple times and the equipment's historical operating data is complete. Its characteristic is the use of standardized sampling frequency and characteristic parameter thresholds for monitoring, in order to reduce computational load and energy consumption. Deep monitoring mode is suitable for situations such as the first transport of new types of grain, the first operation of equipment after maintenance, or the presence of abnormal records in the historical operating database. This mode will enable higher frequency data acquisition, more refined image analysis, and stricter anomaly judgment criteria.

[0060] As an optional implementation, the controller determines the monitoring mode by querying the historical operation database. Specifically, the controller compares the currently acquired initial configuration parameters, especially the grain type and equipment model identifier, as keywords with the historical task records stored in the historical operation database. If there is a record in the database that completely matches the current configuration parameters and whose historical operating status is normal, the controller automatically determines the monitoring mode as the regular monitoring mode; otherwise, if there is no matching record, or if the matching record contains fault alarms such as hopper damage or excessive temperature and humidity, the controller automatically determines the monitoring mode as the deep monitoring mode.

[0061] For example, in one application scenario, if the controller detects that the grain being transported is corn, the equipment model is specific, and the historical operation database contains normal operation records of this model of equipment transporting corn over the past thirty days, including standard data ranges from all sensors and no fault alarms, then the controller will set the monitoring mode to normal monitoring mode. In another scenario, the controller identifies that the grain being transported is soybeans, but there are no records of soybean transport in the historical database. In this case, the controller automatically determines it as the first transport and sets the monitoring mode to deep monitoring mode to collect richer sample data and establish a standard model for soybean transport.

[0062] Furthermore, this method allows operators to force the specification of monitoring modes. For example, a forced depth monitoring button can be set on the control panel. When operators determine that the moisture content of the grain is high or the equipment has not been maintained for a long time, they can manually select the depth monitoring mode to enhance the reliability of monitoring. Regardless of the method used to determine the monitoring mode, the controller will store the currently selected mode identifier in the working register after confirmation, which will guide subsequent data acquisition and processing steps.

[0063] Through the above steps, this method can intelligently select the most suitable monitoring strategy based on different transportation tasks and equipment historical status, optimize resource utilization while ensuring monitoring accuracy, and lay the foundation for subsequent multi-sensor collaborative acquisition and intelligent diagnosis.

[0064] Step S120: In response to the monitoring mode, a monitoring instruction set is generated. The monitoring instruction set is used to control multiple sensors deployed at key locations in the conveying system to simultaneously collect multi-dimensional real-time operational data and grain quality data, and to preprocess the collected data to extract various abnormal features. The specific content is as follows:

[0065] In step S120, after determining the monitoring mode and storing the currently selected mode identifier in the working register, this method immediately responds to the monitoring mode by generating a matching monitoring instruction set. This monitoring instruction set is not a fixed set of commands, but rather a dynamically generated set of instructions based on the differences between the conventional monitoring mode and the deep monitoring mode. For the conventional monitoring mode, the monitoring instruction set requires each sensor to use a lower sampling frequency and a more lenient data precision to save computing resources and storage space; for the deep monitoring mode, the monitoring instruction set forces each sensor to operate at the highest sampling frequency, enable continuous image capture, and activate the data streams of all auxiliary sensors, thereby providing the richest information input for subsequent intelligent diagnosis.

[0066] The primary task of the monitoring instruction set is to control multiple sensors deployed at key locations on the conveying equipment to simultaneously collect multi-dimensional real-time operational and grain quality data. These sensors can be functionally categorized into motion status monitoring, material status monitoring, and environmental status monitoring groups. The motion status monitoring group includes speed sensors mounted on the chain bucket drive shaft, tension sensors mounted on the chain tensioning device, and vibration sensors mounted near the chain guide rails and drive wheels. The material status monitoring group includes weight sensors mounted at the bottom of each hopper, high-definition vision sensors mounted on the support above the hopper, and laser rangefinders mounted at the inlet and outlet sections. The environmental status monitoring group includes temperature and humidity sensors installed inside the conveying channel and near the outlet, as well as optional dust concentration sensors.

[0067] Taking a 50-meter-long grain transport chain containing 120 hoppers as an example, the speed sensor outputs a pulse signal per revolution. The controller calculates the real-time running speed of the chain based on the number of pulses per unit time, with an accuracy of 0.1 meters per minute. The tension sensor uses a strain gauge structure and is installed between the spring seat and the fixed base of the chain tensioning device. It can sense tension changes in the range of 0.5 kN to 50 kN, and the output signal directly reflects the chain tension state after analog-to-digital conversion. The vibration sensor uses a piezoelectric accelerometer with a frequency response range of 0.5 Hz to 5 kHz, capable of capturing high-frequency vibrations generated by the meshing of the chain and drive wheel, as well as low-frequency impacts caused by hopper jamming. The weight sensor uses a cantilever beam weighing module, with one independently installed at the bottom of each hopper. The range is 0 to 100 kg, with an accuracy of ±0.05 kg. When a hopper passes through a fixed weighing area, the sensor instantaneously records the weight of the grain in the hopper. The high-definition vision sensor uses a two-megapixel industrial camera equipped with an automatic lens cleaning device, capable of capturing 30 frames per second with a resolution of 1920 x 1080 pixels. The laser rangefinder sensor, employing the phase method measurement principle, is installed above the inlet and outlet to measure the real-time height change of the grain accumulation surface, and calculates the instantaneous volumetric flow rate by combining this with the conveyor belt width.

[0068] To achieve synchronized data acquisition across multiple sensors, the controller sends a synchronization start signal to all sensors simultaneously with the generation of the monitoring command set. This signal is generated by a high-precision real-time clock module within the controller, with an accuracy down to the microsecond level. Upon receiving the synchronization signal, each sensor begins acquiring data at the same time reference point, appending the clock count value at the acquisition moment as a timestamp to each data packet. For smart sensors unable to receive external synchronization signals, the controller employs an interpolation alignment algorithm during subsequent data preprocessing to align the data from different sensors onto a common timeline according to their timestamps.

[0069] After synchronous data acquisition is initiated, all sensors operate continuously at the sampling frequency specified in the monitoring instruction set. In the standard monitoring mode, the speed sensor acquires a speed value once per second, the tension sensor acquires a tension value once per second, each hopper records its weight as it passes the weighing zone, the vibration sensor acquires a one-second vibration waveform every ten seconds, and the temperature and humidity sensor acquires values ​​every five seconds. In the depth monitoring mode, the sampling frequency of the speed sensor increases to ten times per second, the tension sensor to five times per second, the vibration sensor is switched to continuous acquisition and real-time storage, the high-definition vision sensor captures sixty frames per second with infrared illumination, the temperature and humidity sensor to twice per second, and the laser rangefinder to twenty times per second. The amount of data generated in the depth monitoring mode is approximately twenty times that of the standard mode, and the controller temporarily activates a large-capacity cache and determines in real-time whether edge computing compression is needed.

[0070] The acquired multi-dimensional raw data is then sent to the preprocessing stage. The preprocessing operation is performed in the digital signal processing unit inside the controller, including five sub-steps: data cleaning, filtering and noise reduction, time synchronization and alignment, outlier removal, and data normalization.

[0071] The first step in data cleaning is to remove data points that are clearly beyond the sensor's range or physically impossible. For example, when the weight sensor records a negative value or exceeds the maximum range of 120 kg, the controller discards the data point and records a sensor malfunction event. For missing data points, the controller uses linear interpolation between adjacent time points to fill in the gaps and ensure the continuity of the time series.

[0072] Different filters are used for different types of data to reduce noise. For weight and tensile data, the controller uses a moving average filtering algorithm with a window width of five consecutive sampling points. This effectively eliminates high-frequency noise caused by mechanical vibration or electromagnetic interference while retaining the low-frequency trend reflecting changes in grain flow. For vibration data, the controller first applies a bandpass filter with a lower cutoff frequency set at 10 Hz to filter out low-frequency environmental vibrations and an upper cutoff frequency set at 2 kHz to retain the characteristic frequencies of sprocket meshing and hopper collisions. A Hanning window is then used for a fast Fourier transform to obtain the vibration spectrum. For temperature and humidity data, the controller uses a median filtering algorithm with a window width of seven sampling points. This effectively removes isolated outliers caused by occasional sensor drift while maintaining the true trend of temperature and humidity changes. For laser ranging data, the controller uses a Gaussian filtering algorithm to smooth measurement fluctuations caused by irregular reflections from the grain surface.

[0073] Time synchronization alignment is a crucial step in preprocessing. Since the sampling frequencies of the speed, vibration, and laser rangefinder sensors are much higher than those of the temperature and humidity sensors, the controller needs to unify all data onto the same time reference. Specifically, the time axis of the sensor with the highest sampling frequency is used as the main axis; for example, the exposure time of each frame of an image from a high-definition vision sensor in depth monitoring mode is used as the reference time point. Then, cubic spline interpolation is performed on the data from the low-speed sensors to calculate the theoretical value of that sensor at each reference time point. For the sensor data from the motion state monitoring group, the controller also binds the fixed spatial position of each hopper to a time point based on the mapping relationship between the chain's running speed and the hopper's position, generating a three-dimensional data cube of "time-hopper number-measured value".

[0074] Outlier removal employs a statistical control chart method. For weight data, the controller first calculates the mean and standard deviation of the weights of the past 100 hoppers. Then, it compares the current hopper weight with the mean. If the deviation exceeds three times the standard deviation, the data point is marked as a suspected outlier. The controller does not directly discard suspected outliers but cross-validates them by combining weight data from adjacent hoppers and visual image data. For example, if the current hopper weight is abnormally low, but the visual image shows the hopper is intact and the weights of the hoppers before and after it are normal, it is determined to be a grain spillage feature and the data is retained. If the visual image also shows obvious damage to the hopper, it is also retained and marked as a damage feature.

[0075] Data normalization maps all physical quantities with different dimensions to a unified numerical range, preparing them for subsequent input into deep learning models. The controller pre-stores the minimum and maximum values ​​of feature parameters for each sensor; these extreme values ​​are derived from factory calibration or statistical analysis of historical operating data. For example, the minimum value for a weight sensor is 0 kg (no load), and the maximum value is 50 kg (full load); the minimum value for a temperature sensor is -10°C (cold start in winter), and the maximum value is 50°C (high summer temperature or grain heating); the minimum value for a humidity sensor is 5% (extremely dry environment), and the maximum value is 95% (saturated humid air); the minimum value for vibration acceleration is 0 m / s² (stationary state), and the maximum value is 50 m / s² (severe impact). The normalization formula is: the normalized value equals the original value minus the minimum value, divided by the difference between the maximum and minimum values. After normalization, all data are distributed between zero and one, eliminating the impact of dimensional differences on model training.

[0076] After completing the above preprocessing steps, the controller extracts various anomaly features from the processed data. The extraction of anomaly features is performed separately based on the characteristics of each type of data.

[0077] For weight data, the controller maintains a sliding window with a length equal to the total number of hoppers (120). As each hopper passes the weighing zone, the controller calculates the difference between the current hopper weight and the historical average weight of hoppers at the same position within the window, as well as the difference between the current hopper weight and the average weights of the five hoppers immediately before and after it. If the current hopper weight is consistently below 30% of the historical average for three consecutive cycles, and the deviation from the average of adjacent hoppers also exceeds 30%, a "grain spillage" feature is extracted, and the hopper is numbered. If the current hopper weight exceeds 150% of the historical average, an "overload" feature is extracted. If the weight is close to zero and the visual image shows no grain in the hopper, an "empty" feature is extracted.

[0078] For visual image data, the controller invokes image processing algorithms based on edge detection and morphological analysis. First, the color image is converted to grayscale, then the Canny edge detection operator is applied to extract the hopper's edge contour. The current contour is then matched against a template contour. If the contour's missing area exceeds 10%, the "hopper damaged" feature is extracted; if the contour is significantly offset or distorted, the "hopper deformed" feature is extracted; if the contour lacks a necessary edge segment, the "hopper missing" feature is extracted. For the grain image inside the hopper, the controller applies a texture analysis algorithm to calculate the grayscale co-occurrence matrix. If the texture uniformity is below a threshold and irregular blocky areas appear, the "grain clumping" feature is extracted; if a contour significantly different from the grain texture is detected, such as metal fragments, stones, or plastic pieces, the "foreign object intrusion" feature is extracted.

[0079] For vibration data, the controller extracts features in the frequency domain. After performing a Fast Fourier Transform on the time-domain vibration signal, a frequency and amplitude spectrum is obtained. The normal meshing frequency of the chain is equal to the number of teeth on the drive wheel multiplied by the rotational speed divided by sixty. If an abnormal increase in amplitude occurs near this frequency, exceeding twice the normal value, the "chain wear" feature is extracted. Chain bucket jamming generates low-frequency impacts, resulting in a wide-band energy increase in the spectrum between 10 Hz and 50 Hz. The controller extracts the ratio of the total energy in this frequency band to the normal value; if it exceeds a threshold, the "chain bucket jamming" feature is extracted. Abnormalities in the drive wheel manifest as abnormal drive wheel rotation frequency and its harmonic amplitudes. The controller extracts the ratio of the first-order rotation frequency amplitude to the second-order harmonic amplitude; if this ratio exceeds the normal range, the "drive wheel wear or eccentricity" feature is extracted.

[0080] For temperature and humidity data, the controller calculates the rate of temperature change over time. If the temperature rises continuously within a short period and the rate of change exceeds 0.5 degrees Celsius per minute, while humidity also rises abnormally, the early characteristic of "grain heating" is extracted. If the humidity remains above 75% and the temperature is within the normal range, the characteristic of "grain dampness" is extracted, indicating a potential risk of mold growth. If the humidity drops sharply within a short period while the temperature remains unchanged, it may indicate that the drying equipment is overworking or that the grain has become moldy and clumped together, preventing moisture from escaping.

[0081] All extracted abnormal features are organized into a feature vector in chronological order. Each feature vector contains the current hopper number, feature type, feature strength confidence level, and a summary of the relevant raw data. The controller temporarily stores these feature vectors in a local circular buffer and triggers the next stage of the intelligent diagnostic process. Through the generation of the aforementioned monitoring instructions, simultaneous acquisition by multiple sensors, and refined preprocessing and feature extraction, this method provides high-quality, multi-dimensional input data for subsequent deep learning diagnostics, significantly improving the accuracy and real-time performance of fault detection.

[0082] Step S140: In response to the extracted multiple abnormal features, the multiple abnormal features are input into a pre-trained multimodal fusion intelligent diagnostic model to obtain diagnostic results and generate a control instruction set. The control instruction set is used to adjust the operating parameters of the grain transport chain bucket or to execute fault linkage handling. The specific content is as follows:

[0083] In step S140, after extracting various abnormal features and organizing them into a feature vector sequence in chronological order, this method immediately responds to these extracted abnormal features by inputting them into a pre-trained multimodal fusion intelligent diagnostic model. This model is a deep learning network, specifically employing a hybrid architecture of convolutional neural networks and long short-term memory networks. The convolutional neural network part is used to extract the intrinsic correlations between different abnormal features from a spatial dimension, such as whether weight abnormalities, visual abnormalities, and vibration abnormalities point to the same fault source; the long short-term memory network part is used to capture the evolution trend of abnormal features from a temporal dimension, such as the rate of continuous increase in temperature and humidity, and the pattern of gradually increasing vibration amplitude. Before being put into practical use, the model has undergone offline training. The training samples are derived from a large amount of multi-dimensional data collected from historical operating databases under normal operating conditions and various fault conditions. Each sample is manually labeled, with annotations including fault type, fault location, fault severity, and grain quality abnormality type.

[0084] After the extracted abnormal feature vector sequence is fed into the model, the model calculates layer by layer during forward propagation, ultimately outputting a diagnostic result. The diagnostic result includes four elements: fault type, fault location, fault severity, and confidence level. Fault types include, but are not limited to: hopper damage, hopper deformation, missing hopper, grain spillage, chain wear, chain jamming, drive wheel malfunction, grain clumping, foreign object contamination, grain overheating, and grain mold. Fault location is precise down to the specific hopper number, chain segment, or drive wheel position. Fault severity is divided into three levels: mild, moderate, and severe. For example, less than 5% of the hopper's damaged area is considered mild, 5% to 15% is moderate, and greater than 15% is severe. The confidence level is expressed as a percentage, indicating the model's certainty about the diagnostic result. When the confidence level is below 70%, the controller triggers a review process, requiring the high-definition vision sensor in depth monitoring mode to perform secondary confirmation of the suspected area.

[0085] Taking grain spillage as an example, when the abnormal features received by the model include: the weight of a specific numbered hopper is lower than 40% of the historical average for three consecutive cycles, the dust concentration sensor value in the conveying channel below the hopper increases, and the vibration spectrum of the hopper shows abnormal impact when it passes through a fixed position, the model integrates these multimodal features and outputs the following diagnostic results: the fault type is grain spillage, the fault location is hopper number 37, the severity is moderate, and the confidence level is 92%.

[0086] After obtaining the diagnostic results, this method further generates a control instruction set. The control instruction set is a set of commands directed to different actuators, and its specific content depends on the type and severity of the diagnostic results. The control instruction set is used to adjust the operating parameters of the grain transport chain hoppers or to execute fault-based linkage measures, specifically including speed adjustment commands, tension adjustment commands, fault shutdown commands, hopper marking commands, and environmental adjustment commands.

[0087] For minor malfunctions that do not affect continued operation, such as a hopper experiencing minor damage with a confidence level below 85%, the controller generates a speed reduction command, decreasing the chain bucket's operating speed from the rated value to 60% of the rated value. Simultaneously, a hopper marking command is generated, storing the hopper's number and location information in memory for replacement by maintenance personnel after the current conveying task is completed. During the speed reduction period, the controller also increases the monitoring frequency of the hopper, raising the sampling frequency of its weight and image data to the standard of depth monitoring mode.

[0088] For moderate or high-confidence minor faults, such as abnormal tension due to chain wear or a continuous increase in grain temperature and humidity, the controller generates adaptive adjustment commands. Specifically, when the chain tension fed back by the tension sensor is lower than the preset lower limit, the controller sends a tensioning command to the tensioning device, driving the hydraulic telescopic rod or electric push rod to increase the chain preload until the tension returns to the normal range. When the diagnostic results indicate that the grain is overheating, the controller generates commands to reduce the operating speed and activate the ventilation system, reducing the chain bucket speed by 30%, simultaneously activating the axial flow fan installed on the side wall of the conveyor channel to increase air circulation and reduce the grain temperature, and increasing the sampling frequency of the temperature and humidity sensors to once per second, continuously monitoring until the temperature returns to normal.

[0089] In the event of a severe malfunction or an emergency involving safety, such as a jammed bucket chain, broken chain, or severely worn drive wheel detected by the model, the controller immediately generates an emergency stop command and a command to close the feed inlet valve. The emergency stop command directly cuts off the power supply to the drive motor via a hardware interrupt, and the braking device simultaneously activates to stop the bucket chain within the shortest possible distance. The command to close the feed inlet valve is sent to the pneumatic gate valve below the feed hopper, and the valve closes completely within 0.5 seconds, preventing grain from continuing to fall into the bucket chain and preventing the malfunction from escalating. Simultaneously, the controller also sends an alarm signal to the monitoring center, triggering an audible and visual alarm and automatically pushing fault information to the handheld terminals of on-site operators.

[0090] In this embodiment, the control instruction set also includes precise positioning instructions based on three-dimensional map analysis and wear prediction instructions based on vibration spectrum. Specifically, after the controller extracts spillage characteristics from the weight data, it further performs the following operations: Based on the fixed pitch of the chain bucket and the current operating speed, a mapping relationship is established between the theoretical transit time and the conveying position of each bucket. The fixed pitch of the chain bucket is 500 mm. Each rotation of the drive wheel moves eight buckets. The speed sensor monitors the angular velocity of the drive wheel in real time, and the controller calculates the theoretical time for each bucket to reach each sensor position from the feed inlet. The weight values ​​measured in real time for each bucket are arranged in a time sequence to form a three-dimensional map. In this map, the horizontal axis represents time, the vertical axis represents the bucket number, and the color depth represents the weight value. When the weight of a hopper at a certain location shows a continuous and regular abnormal decrease over multiple consecutive cycles—for example, the weight of hopper 37 is below 30% of the average in eight out of the last ten cycles—and this abnormal decrease is unrelated to the weight values ​​of upstream and downstream hoppers 36 and 38, the controller determines that the corresponding hopper has experienced an independent spillage fault and precisely identifies the spilling hopper as hopper 37. Based on this precise identification, the control instruction set generates the following specific instructions: mark hopper 37 as requiring maintenance, and simultaneously reduce the operating speed to 50% of the rated value, but do not stop the entire conveyor line, thereby maximizing production continuity.

[0091] For abnormal drive wheel diagnosis, when the spectral data collected by the vibration sensor outputs the drive wheel wear characteristics after model analysis, the controller further extracts the characteristic frequency of the drive wheel and the amplitude ratio of its harmonics. The characteristic frequency of the drive wheel is equal to the number of teeth multiplied by the rotational frequency, and the harmonic amplitude ratio refers to the ratio of the second harmonic amplitude to the fundamental frequency amplitude. The controller pre-stores an "amplitude ratio-wear amount" mapping table in the database. This table is obtained by fitting a large amount of experimental data and can convert the amplitude ratio into tooth surface wear amount in millimeters. The controller substitutes the currently calculated amplitude ratio into this mapping table to calculate the estimated wear amount of the current drive wheel teeth. When the estimated wear amount exceeds a preset threshold, such as 1.5 millimeters, the controller generates a warning command to replace the drive wheel. This command is not executed in the form of an emergency shutdown, but rather a notification is displayed on the monitoring interface with highlighted text prompting the operator to arrange a planned shutdown for replacement. At the same time, the remaining lifespan estimate information is also displayed for maintenance decision-making reference. This wear prediction method based on vibration spectrum upgrades traditional qualitative diagnosis to quantitative prediction, significantly improving the ability of predictive maintenance.

[0092] The controller is also responsible for storing all diagnostic results, control commands, and corresponding raw monitoring data in a blockchain database in real time, establishing a full-process data traceability system. The blockchain database uses distributed ledger technology; each data block contains a timestamp, a hash value for the diagnostic result, a hash value for the control command, and the hash value of the previous block, forming an immutable chain structure. Any modification to historical data will disrupt the consistency of the hash chain and will be detected immediately. Based on historical data in the blockchain database, the controller calls the training module during idle periods to continuously train and optimize the multimodal fusion intelligent diagnostic model. Specifically, it extracts manually verified data samples from the blockchain database, uses them as incremental training sets, updates the model's weight parameters using a mini-batch gradient descent algorithm, and then writes the updated model version number into the blockchain, forming a trusted closed loop of "data collection-diagnosis-control-evidence storage-model iteration." This mechanism not only ensures the authenticity and traceability of grain transportation data but also enables the intelligent diagnostic model to continuously evolve with accumulated operational experience, gradually improving diagnostic accuracy.

[0093] Through the above steps, this method achieves full-process automation from abnormal feature input to intelligent diagnosis, adaptive control, and fault linkage handling. The multimodal fusion model based on convolutional neural networks and long short-term memory networks can comprehensively utilize multiple features such as weight, vision, vibration, temperature, and humidity, significantly improving the accuracy and reliability of fault diagnosis and reducing false alarms and missed alarms. The generated closed-loop control instruction set can execute differentiated handling strategies according to the fault type and severity. For minor faults, continuous production is maintained and the problematic area is marked; for severe faults, emergency shutdown is implemented to protect the equipment. Furthermore, wear prediction based on vibration spectrum and leak location based on 3D mapping further enhance the precision of diagnosis and predictive maintenance capabilities. The introduction of a blockchain database provides a reliable data storage and continuous model optimization foundation for the entire monitoring and transmission process.

[0094] Based on steps S110 to S140 above, in this embodiment of the invention, by acquiring initial configuration parameters in response to the conveying start command and intelligently selecting a conventional monitoring mode or a deep monitoring mode, an adaptive monitoring strategy for different working conditions is realized; in response to the monitoring mode, a monitoring instruction set is generated to control multiple sensors to simultaneously collect multi-dimensional data and perform preprocessing to extract abnormal features, thereby realizing comprehensive real-time perception of the chain bucket's operating status and grain quality; in response to the extracted multiple abnormal features, a pre-trained multi-modal fusion intelligent diagnostic model is used for accurate diagnosis, and a control instruction set is generated to adjust operating parameters or perform fault linkage handling, thereby effectively improving the safety, reliability, and efficiency of grain conveying, reducing grain loss and equipment maintenance costs, and realizing full-process data traceability and predictive maintenance.

[0095] The method of an embodiment of the present invention determines a monitoring mode based on initial configuration parameters, including: matching the initial configuration parameters with a historical operation database to obtain a matching result, the matching result including: successful matching and unsuccessful matching; in response to a successful matching result, determining the monitoring mode as a regular monitoring mode; in response to an unsuccessful matching result, determining the monitoring mode as a deep monitoring mode.

[0096] In this embodiment, the solution can automatically match the monitoring mode based on historical operating data. For regular operating conditions with existing records, low-power conventional monitoring is used to save resources, while for new operating conditions without records, it automatically switches to high-precision deep monitoring to ensure data integrity and diagnostic reliability.

[0097] Furthermore, in response to the monitoring mode being the conventional monitoring mode, a monitoring instruction set is generated, including: determining the standard sensor sampling frequency and characteristic parameter thresholds corresponding to the grain transport chain bucket based on the initial configuration parameters; and generating the monitoring instruction set based on the standard sensor sampling frequency and characteristic parameter thresholds.

[0098] In this embodiment, the solution can quickly generate monitoring instructions adapted to normal working conditions based on the preset standard sensor sampling frequency and characteristic parameter thresholds, minimizing the consumption of data acquisition and processing resources while ensuring basic monitoring accuracy.

[0099] Furthermore, in response to the monitoring mode being deep monitoring mode, a monitoring instruction set is generated, including: acquiring several image data of the grain transport chain hopper; determining the hopper's appearance integrity information and grain clumping information based on the several image data; determining a high-density sampling scheme based on the hopper's appearance integrity information and grain clumping information; and generating a monitoring instruction set based on the high-density sampling scheme.

[0100] In this embodiment, the scheme can identify the integrity of the hopper's appearance and the state of grain clumping in real time through visual images, and dynamically generate a high-density sampling scheme accordingly, thereby strengthening the monitoring of abnormal areas and significantly improving the targeting of data collection and the ability to detect faults early in the deep monitoring mode.

[0101] Furthermore, based on several image data sets, the hopper's appearance integrity information and grain agglomeration information are determined, including: filtering several image data sets to obtain several filtered image data sets, the filtering processes including: median filtering, Gaussian filtering, and guided filtering; based on the several filtered image data sets, an adaptive histogram equalization method is used to obtain several processed image data sets; based on the several processed image data sets, several initial hopper damage feature data sets and initial agglomeration feature data sets are determined; based on the several initial hopper damage feature data sets and initial agglomeration feature data sets, a majority voting calculation method is used to determine the hopper's appearance integrity information and grain agglomeration information.

[0102] In this embodiment, the scheme preprocesses the image by combining multiple filtering methods with adaptive histogram equalization, and uses a majority voting method to fuse multiple initial feature data. This effectively suppresses noise interference and eliminates single misjudgments, significantly improving the accuracy and robustness of identifying hopper damage and grain agglomeration.

[0103] Furthermore, multiple abnormal features are extracted, including: features representing empty, full, and overloaded hoppers, as well as grain spillage, based on weight data trends; features representing hopper damage, deformation, missing parts, grain clumping, and foreign matter contamination, based on visual image data; features representing chain wear, chain jamming, and drive wheel malfunctions, based on vibration data spectral analysis; and early features representing grain heating and mold growth, based on temperature and humidity data curves. In this embodiment, the scheme extracts multi-dimensional abnormal features of equipment status and grain quality by fusing four types of data: weight, visual, vibration, and temperature and humidity, achieving comprehensive coverage and accurate identification of mechanical failures and grain quality problems.

[0104] Furthermore, in response to the diagnostic results, a set of control instructions is generated, including: automatically adjusting the running speed of the conveyor belt based on real-time collected grain flow data; controlling the tensioning device to automatically adjust the chain tension based on real-time collected chain tension data; generating a set of control instructions to reduce the running speed and mark the location of the conveyor belt when minor damage to the hopper is diagnosed; generating a set of control instructions to immediately stop the conveying system and close the inlet valve when the conveyor belt is diagnosed as jammed or the chain is broken; and generating a set of control instructions to reduce the running speed of the conveyor belt and activate the ventilation device when abnormal increases in grain temperature and humidity are diagnosed.

[0105] In this embodiment, the solution can automatically execute multi-level differentiated control commands, including speed adjustment, tension adjustment, fault shutdown, hopper marking, and environmental control, based on the diagnostic results. This enables precise response and coordinated handling of different fault types, maximizing conveying safety and operational efficiency.

[0106] Furthermore, in response to the diagnostic results, a set of control instructions is generated, including: establishing a mapping relationship between the theoretical throughput time and conveying position of each hopper based on the fixed pitch and operating speed of the chain bucket; arranging the real-time measured weight values ​​of each hopper in a time series to form a three-dimensional map; when the weight value of a hopper at a certain position shows a continuous and regular abnormal decrease, and is not related to the weight values ​​of the upstream and downstream hoppers at that position, it is determined that the corresponding hopper has leaked, and the number of the leaking hopper is accurately located; extracting the characteristic frequency of the drive wheel and the amplitude ratio of its harmonics; calculating the estimated wear of the current drive wheel teeth based on the pre-established mapping database; and issuing a warning message to replace the drive wheel when the estimated wear exceeds a preset threshold.

[0107] In this embodiment, the scheme achieves precise positioning of the spill hopper by constructing a three-dimensional map, and achieves quantitative prediction of the wear of the drive wheel by combining the vibration spectrum amplitude ratio and wear amount mapping database. This improves fault diagnosis from qualitative judgment to precise positioning and remaining life prediction, significantly enhancing predictive maintenance capabilities.

[0108] Through the above description of the embodiments, those skilled in the art can clearly understand that the methods according to the above embodiments can be implemented by means of software plus necessary general-purpose hardware platforms. Of course, they can also be implemented by hardware, but in many cases the former is a better implementation method. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product is stored in a storage medium (such as ROM / RAM, magnetic disk, optical disk) and includes several instruction sets to cause a terminal device (which may be a mobile phone, computer, server, or network device, etc.) to execute the methods of the various embodiments of the present invention.

[0109] This invention also provides an intelligent monitoring and conveying device for grain transport chains, which is used to implement the above embodiments and preferred embodiments; details already described will not be repeated. As used below, the term "module" can refer to a combination of software and / or hardware that performs a predetermined function. Although the devices described in the following embodiments are preferably implemented in software, hardware implementations, or a combination of software and hardware, are also possible and contemplated.

[0110] Figure 2 According to one embodiment of the present invention, a grain transport chain bucket intelligent monitoring and conveying device includes:

[0111] The acquisition module 201 is used to acquire the initial configuration parameters of the grain transport chain bucket in response to the transport start command set, and determine the monitoring mode based on the initial configuration parameters. The monitoring modes include: regular monitoring mode and deep monitoring mode.

[0112] The generation module 202 is used to generate a monitoring instruction set in response to the monitoring mode. The monitoring instruction set is used to control multiple sensors deployed at key locations in the conveying system to simultaneously collect multi-dimensional real-time operating data and grain quality data, and to preprocess the collected data to extract multiple abnormal features.

[0113] The diagnostic module 203 is used to respond to the extracted multiple abnormal features, input the multiple abnormal features into the pre-trained multimodal fusion intelligent diagnostic model, obtain the diagnostic results, and generate a control instruction set. The control instruction set is used to adjust the operating parameters of the grain transport chain bucket or to perform fault linkage handling.

[0114] It should be noted that the above modules can be implemented by software or hardware. For the latter, they can be implemented in the following ways, but are not limited to: all the above modules are located in the same processor; or, the above modules are located in different processors in any combination.

[0115] According to one embodiment of the present invention, an electronic device is also provided, comprising: a memory storing an executable program; and a processor for running the program, wherein the program executes the above-described intelligent monitoring and conveying method for grain transport chain buckets when it runs.

[0116] Optionally, in this embodiment, the processor can be configured to perform the following steps via a computer program:

[0117] Step S1: In response to the delivery start command set, obtain the initial configuration parameters of the grain transport chain bucket, and determine the monitoring mode based on the initial configuration parameters. The monitoring modes include: regular monitoring mode and deep monitoring mode.

[0118] Step S2: In response to the monitoring mode, a monitoring instruction set is generated. The monitoring instruction set is used to control multiple sensors deployed at key locations in the conveying system to simultaneously collect multi-dimensional real-time operating data and grain quality data, and to preprocess the collected data to extract various abnormal features.

[0119] Step S3: In response to the extracted multiple abnormal features, the multiple abnormal features are input into the pre-trained multimodal fusion intelligent diagnostic model to obtain diagnostic results and generate a control instruction set. The control instruction set is used to adjust the operating parameters of the grain transport chain bucket or to perform fault linkage handling.

[0120] According to one embodiment of the present invention, a computer-readable storage medium is also provided, the computer-readable storage medium including a stored executable program, wherein, when the executable program is running, it controls the device where the storage medium is located to execute the above-described intelligent monitoring and conveying method for grain transport chain buckets.

[0121] Optionally, in this embodiment, the storage medium may be configured to store a computer program for performing the following steps:

[0122] Step S1: In response to the delivery start command set, obtain the initial configuration parameters of the grain transport chain bucket, and determine the monitoring mode based on the initial configuration parameters. The monitoring modes include: regular monitoring mode and deep monitoring mode.

[0123] Step S2: In response to the monitoring mode, a monitoring instruction set is generated. The monitoring instruction set is used to control multiple sensors deployed at key locations in the conveying system to simultaneously collect multi-dimensional real-time operating data and grain quality data, and to preprocess the collected data to extract various abnormal features.

[0124] Step S3: In response to the extracted multiple abnormal features, the multiple abnormal features are input into the pre-trained multimodal fusion intelligent diagnostic model to obtain diagnostic results and generate a control instruction set. The control instruction set is used to adjust the operating parameters of the grain transport chain bucket or to perform fault linkage handling.

[0125] Optionally, in this embodiment, the storage medium may include, but is not limited to, various media capable of storing computer programs, such as USB flash drives, read-only memory (ROM), random access memory (RAM), portable hard drives, magnetic disks, or optical disks.

[0126] According to one embodiment of the present invention, a computer program product is also provided, including a computer program that, when executed by a processor, implements the above-described intelligent monitoring and conveying method for grain transport chain buckets.

[0127] Optionally, in this embodiment, the above-mentioned computer program product can be configured as a computer program that performs the following steps:

[0128] Step S1: In response to the delivery start command set, obtain the initial configuration parameters of the grain transport chain bucket, and determine the monitoring mode based on the initial configuration parameters. The monitoring modes include: regular monitoring mode and deep monitoring mode.

[0129] Step S2: In response to the monitoring mode, a monitoring instruction set is generated. The monitoring instruction set is used to control multiple sensors deployed at key locations in the conveying system to simultaneously collect multi-dimensional real-time operating data and grain quality data, and to preprocess the collected data to extract various abnormal features.

[0130] Step S3: In response to the extracted multiple abnormal features, the multiple abnormal features are input into the pre-trained multimodal fusion intelligent diagnostic model to obtain diagnostic results and generate a control instruction set. The control instruction set is used to adjust the operating parameters of the grain transport chain bucket or to perform fault linkage handling.

[0131] Optionally, specific examples in this embodiment can refer to the examples described in the above embodiments and optional implementations, and will not be repeated here.

[0132] In the above embodiments of the present invention, the descriptions of each embodiment have different focuses. For parts not described in detail in a certain embodiment, please refer to the relevant descriptions of other embodiments.

[0133] In the several embodiments provided in this application, it should be understood that the disclosed technical content can be implemented in other ways. The device embodiments described above are merely illustrative; for example, the division of units can be 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 displayed or discussed mutual couplings, direct couplings, or communication connections may be through some interfaces; indirect couplings or communication connections between units or modules may be electrical or other forms.

[0134] 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 units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0135] Furthermore, 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. The integrated unit can be implemented in hardware or as a software functional unit.

[0136] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this invention, in essence, or the part that contributes to the prior art, or all or 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 instruction sets 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 of the various embodiments of this invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, read-only memory (ROM), random access memory (RAM), portable hard drives, magnetic disks, or optical disks.

[0137] The above are merely preferred embodiments of the present invention. It should be noted that those skilled in the art can make various improvements and modifications without departing from the principle of the present invention, and these improvements and modifications should also be considered within the scope of protection of the present invention.

Claims

1. A method for intelligent monitoring and conveying of grain transport containers, characterized in that, include: In response to the delivery start command set, the initial configuration parameters of the grain transport chain bucket are obtained, and the monitoring mode is determined based on the initial configuration parameters, wherein the monitoring mode includes: a regular monitoring mode and a deep monitoring mode; In response to the monitoring mode, a monitoring instruction set is generated. The monitoring instruction set is used to control multiple sensors deployed at key locations in the conveying system to simultaneously collect multi-dimensional real-time operating data and grain quality data, and to preprocess the collected data to extract multiple abnormal features. In response to the extracted multiple abnormal features, the multiple abnormal features are input into a pre-trained multimodal fusion intelligent diagnostic model to obtain diagnostic results and generate a control instruction set, which is used to adjust the operating parameters of the grain transport chain bucket or to perform fault linkage handling.

2. The intelligent monitoring and conveying method for grain transport chain buckets according to claim 1, characterized in that, Based on the initial configuration parameters, the monitoring mode is determined, including: Based on the historical operation database, the initial configuration parameters are matched with the historical operation database to obtain a matching result, which includes: successful matching and unsuccessful matching; In response to a successful match result, the monitoring mode is determined to be the normal monitoring mode. In response to the matching result being a failure to match, the monitoring mode is determined to be a deep monitoring mode.

3. The intelligent monitoring and conveying method for grain transport chain buckets according to claim 2, characterized in that, In response to the monitoring mode being the normal monitoring mode, the monitoring instruction set is generated, including: Based on the initial configuration parameters, determine the standard sensor sampling frequency and feature parameter thresholds corresponding to the grain transport chain bucket; The monitoring instruction set is generated based on the standard sensor sampling frequency and characteristic parameter thresholds.

4. The intelligent monitoring and conveying method for grain transport chain buckets according to claim 2, characterized in that, In response to the monitoring mode being the deep monitoring mode, the monitoring instruction set is generated, including: Acquire several image data of the grain transport chain hopper, and based on the image data, determine the hopper's appearance integrity information and grain clumping information; Based on the information on the integrity of the hopper's appearance and the information on grain clumping, a high-density sampling scheme is determined; The monitoring instruction set is generated based on the high-density sampling scheme.

5. The intelligent monitoring and conveying method for grain transport chain buckets according to claim 4, characterized in that, Based on several of the aforementioned image data, information on the integrity of the hopper's appearance and information on grain clumping are determined, including: Several image data are filtered to obtain several filtered image data. The filtering process includes: median filtering, Gaussian filtering, and guided filtering. Based on several filtered image data, an adaptive histogram equalization method is used to obtain several processed image data. Based on several processed image data, several initial hopper breakage feature data and initial agglomeration feature data are determined; Based on several initial hopper damage characteristic data and initial agglomeration characteristic data, a majority voting calculation method is used to determine the hopper's appearance integrity information and grain agglomeration information.

6. The intelligent monitoring and conveying method for grain transport chain buckets according to claim 1, characterized in that, Extract the various abnormal features, including: Based on the changing trends of weight data, features representing empty, full, overloaded, and grain spillage in the hopper are extracted. Based on visual image data, features representing hopper damage, deformation, missing parts, grain clumping, and foreign matter contamination are extracted. Based on the spectral analysis of vibration data, features characterizing chain wear, chain bucket jamming, and drive wheel abnormalities are extracted. Based on the temperature and humidity data change curves, early characteristics representing grain heating and mold growth are extracted.

7. The intelligent monitoring and conveying method for grain transport chain buckets according to claim 1, characterized in that, In response to the diagnostic result, the control instruction set is generated, including: The operating speed of the chain bucket is automatically adjusted based on real-time collected grain flow data. Based on real-time collected chain tension data, the tensioning device is controlled to automatically adjust the chain tension. When minor damage to the hopper is diagnosed, a set of control instructions is generated to reduce the operating speed and mark the location of the hopper. When chain bucket jamming or chain breakage is diagnosed, a set of control commands is generated to immediately stop the conveying system and close the feed inlet valve. When an abnormal increase in grain temperature and humidity is diagnosed, a set of control instructions is generated to reduce the chain bucket operating speed and activate the ventilation device.

8. The intelligent monitoring and conveying method for grain transport chain buckets according to claim 7, characterized in that, In response to the diagnostic result, generating the control instruction set further includes: Based on the fixed pitch and running speed of the chain bucket, a mapping relationship between the theoretical passage time and the conveying position of each bucket is established; the weight values ​​measured in real time for each bucket are arranged in time series to form a three-dimensional map; when the weight value of a bucket at a certain position shows a continuous and regular abnormal decrease, and is not related to the weight values ​​of the upstream and downstream buckets at that position, it is determined that the corresponding bucket has leaked, and the number of the leaking bucket is accurately located. Extract the characteristic frequency of the drive wheel and the amplitude ratio of its harmonics; calculate the estimated wear of the current drive wheel teeth based on a pre-established mapping database; when the estimated wear exceeds a preset threshold, issue a warning message to replace the drive wheel.

9. A smart monitoring and conveying device for grain transport chain buckets, characterized in that, include: The acquisition module is used to acquire the initial configuration parameters of the grain transport chain bucket in response to the transport start command set, and determine the monitoring mode based on the initial configuration parameters, wherein the monitoring mode includes: a regular monitoring mode and a deep monitoring mode; The generation module is used to generate a monitoring instruction set in response to the monitoring mode. The monitoring instruction set is used to control multiple sensors deployed at key locations in the conveying system to simultaneously collect multi-dimensional real-time operating data and grain quality data, and to preprocess the collected data to extract multiple abnormal features. The diagnostic module is used to respond to the extracted multiple abnormal features, input the multiple abnormal features into a pre-trained multimodal fusion intelligent diagnostic model, obtain diagnostic results, and generate a control instruction set. The control instruction set is used to adjust the operating parameters of the grain transport chain bucket or to perform fault linkage handling.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the program is executed by the processor, it implements the intelligent monitoring and conveying method for grain transport chain buckets as described in any one of claims 1 to 8.

Images