Multi-sensor AI early warning system for bulk grain production lines

Abstract

The present invention discloses a multi-sensor AI early warning system for bulk grain systems, belonging to the field of grain transportation. This system addresses the problem of existing early warning systems' difficulty in centralized data processing, which reduces their practicality. By training and testing a neural network model using historical operating data, the system develops a model capable of predicting the operating status of bulk grain equipment. The neural network model is applied to actual bulk grain systems, enabling real-time prediction of the equipment's operating status. The system is then combined with drones to inspect the bulk grain system and its surroundings for abnormalities. When an abnormality is detected, an alarm is promptly issued to a human-computer interface, prompting management personnel to take timely remedial measures. This effectively implements the early warning function of bulk grain equipment, prevents the escalation of bulk grain equipment failures, avoids potential fire and dust explosion hazards in the surrounding environment, and reduces the workload of inspectors.

Description

Technical Field

[0001] The present invention relates to the field of grain transportation technology, and in particular to a multi-sensor equipment AI early warning system applied to a bulk grain system line. Background Art

[0002] The bulk grain system line equipment includes scrapers, bucket elevators, belt conveyors and other major equipment models, which are responsible for transporting bulk grain to bulk grain silos. During the operation, personnel are required to inspect the operating equipment. Due to the uncertainty of manual inspections, abnormal conditions of the equipment cannot be discovered in time, resulting in the expansion of equipment failures and affecting safe production.

[0003] For example, in the prior art: current sensors, vibration sensors, audio sensors, temperature sensors, thermal imaging, flow sensors and drone audio and video special inspection modules are installed on the scraper conveyor, bucket elevator and belt conveyor of the grain conveying equipment, and the signals are collected.

[0004] However, currently, alarm systems such as current alarm systems, vibration alarm systems, fiber optic temperature measurement, and thermal imaging all operate independently, and individual systems often make inaccurate judgments, causing significant trouble for operators and affecting their effectiveness.

[0005] Therefore, the existing needs are not met, so we proposed a multi-sensor equipment AI early warning system for bulk grain system lines. Summary of the Invention

[0006] The purpose of the present invention is to provide a multi-sensor equipment AI early warning system applied to a bulk grain system line. The system predicts the operating status of the bulk grain equipment by adopting a neural network model, and combines it with a drone to inspect the bulk grain system line and the surrounding environment to see if there are any abnormalities. When an abnormal situation is found, an alarm can be promptly issued to the human-computer interaction interface to remind the management personnel to take timely solutions. In this way, the early warning function of the bulk grain equipment is effectively realized, the expansion of bulk grain equipment failures is prevented, and the hidden dangers of fire and dust explosion in the surrounding environment are avoided, the labor intensity of the inspection personnel is reduced, the labor cost is reduced, and the problems raised in the above-mentioned background technology are solved.

[0007] To achieve the above object, the present invention provides the following technical solutions:

[0008] The multi-sensor AI early warning system for bulk grain production lines includes:

[0009] The data acquisition unit collects the working status of the bulk grain equipment in real time through sensor equipment, obtains the historical and real-time operating data of the bulk grain equipment, and feeds back the various operating data to the centralized processing unit for pre-processing;

[0010] The centralized processing unit receives and pre-processes the historical and real-time operating data of the bulk grain equipment, and extracts characteristic information from the pre-processed historical operating data as an indicator reflecting the historical operating status of the bulk grain equipment;

[0011] The model building unit builds a neural network model, using historical operating data and extracted feature information as input sets and the historical operating status of the bulk grain equipment as labels. The neural network model is trained using cross-validation technology and its parameters are adjusted to make it as close as possible to the actual operating status of the bulk grain equipment.

[0012] The intelligent analysis unit uses a trained neural network model to analyze and predict the historical operating data of the bulk grain equipment to obtain an estimated state value of the bulk grain equipment. Based on the estimated state value, it determines whether the bulk grain equipment is in normal working condition and whether there are any abnormal conditions. The judgment result is compared with the historical operating status of the bulk grain equipment to verify the accuracy of the neural network model prediction.

[0013] The model deployment unit deploys the verified neural network model into the actual bulk grain production system to predict the real-time operating status of bulk grain equipment;

[0014] The environmental inspection unit uses drones to regularly inspect the bulk grain system and its surroundings, obtaining image information of the system and its surroundings to determine whether there are any abnormal hazards.

[0015] The human-computer interaction interface visually displays the operating status of the bulk grain equipment and image information of the bulk grain system line and the surrounding environment.

[0016] Furthermore, the data acquisition unit includes:

[0017] Equipment layout module, installs multiple sensor devices on the scraper, bucket elevator and belt conveyor of bulk grain equipment, including: weight sensor, vibration sensor, temperature sensor, humidity sensor, particle size sensor and dust sensor;

[0018] The timing feedback module sets a timing feedback mechanism for each sensor device, so that after the sensor device enters the real-time collection state, it regularly feeds back the real-time operation data of the bulk grain equipment to the centralized processing unit.

[0019] Furthermore, the centralized processing unit includes:

[0020] Data cleaning module, which removes missing values, duplicate values, and outliers from the running data, as well as noise and unnecessary redundant information, and converts the format of the running data;

[0021] The feature extraction module analyzes the vibration spectrum, particle velocity and size, particle density and flow rate, and temperature distribution of the bulk grain equipment to extract characteristic information about the bulk grain equipment's operating status;

[0022] The data storage module classifies and organizes the received operation data by category, and stores the organized operation data by time series and category to form a database.

[0023] Furthermore, the model building unit includes:

[0024] The sample partitioning module divides the historical operation data and the extracted feature information into training sets and test sets;

[0025] The tag matching module collects the historical operating status of bulk grain equipment in the same period as the historical operating data; sets a tag for each type of historical operating status, and then matches the tag with each type of historical operating status and historical operating data;

[0026] The model building module takes the training set as the input set, builds and trains the neural network model based on the input set, and establishes the feature layer and learning process of the neural network model.

[0027] Furthermore, the intelligent analysis unit includes:

[0028] The model verification module inputs the test set into the trained neural network model for prediction and outputs the estimated state value of the bulk grain equipment. The estimated state value is then compared with the actual result to verify the accuracy of the neural network model's prediction ability.

[0029] The result generation module determines whether the bulk grain equipment is in normal working condition and whether there are any abnormal conditions based on the estimated state value of the bulk grain equipment, and generates corresponding prediction results.

[0030] Furthermore, the intelligent analysis unit further includes:

[0031] The model optimization module regularly uses real-time operation data to update and iterate the neural network model to optimize the parameters and performance of the neural network model.

[0032] Furthermore, the model deployment unit includes:

[0033] The model deployment module exports the verified neural network model into an executable format and deploys the neural network model to the cloud platform loaded with the human-computer interaction interface based on the requirements of the neural network model and system characteristics;

[0034] The timing prediction module regularly predicts the operating status of bulk grain equipment based on the feedback frequency of real-time operating data;

[0035] The early warning prompt module sends a warning prompt to the human-computer interaction interface in a timely manner when the neural network model predicts that the bulk grain equipment and the drone detects abnormal conditions in the surrounding environment, reminding management personnel to check and repair.

[0036] Furthermore, the model deployment unit further includes:

[0037] The model operation module performs real-time analysis based on the neural network model to determine the first real-time operating status of the bulk grain equipment;

[0038] The monitoring module deploys a monitoring mechanism and calculates the operation status of the bulk grain equipment in real time based on the monitoring mechanism to obtain a second real-time operating status;

[0039] Wherein, obtaining the second real-time operating status includes:

[0040] Collect the real-time measurement values of each sensor in each sensor type, and calculate the comprehensive measurement value of each type of bulk grain equipment based on the real-time measurement values of each sensor;

[0041] ;

[0042] in, Indicates the Comprehensive measurement values of bulk grain equipment for each sensor under different sensor types; Indicates the number of sensors; Indicates the total number of sensors; Indicates the Sensor type The initial weight corresponding to each sensor; Indicates the Sensor type Dynamic weight adjustment factor corresponding to each sensor; Indicates the Sensor type Real-time measurement values of sensors; The serial number value indicating the sensor type;

[0043] Calculate the risk assessment value of bulk grain equipment based on the comprehensive measurement values of the bulk grain equipment.

[0044] ;

[0045] in, Indicates the risk assessment value of bulk grain equipment; Indicates the total number of sensor types; Indicates taking the maximum value; Indicates the Safety measurement thresholds corresponding to different sensor types; Indicates the Risk weights corresponding to sensor types;

[0046] Determine the second real-time operating status based on the risk assessment value of the bulk grain equipment. hour, , it means that the current bulk grain equipment is within the safe range; when When , it means that the current bulk grain equipment is in the abnormal range;

[0047] The report generating module is used to generate an evaluation report based on the first real-time operating status and the second real-time operating status, and transmit the evaluation report to the user terminal.

[0048] Furthermore, the environmental inspection unit includes:

[0049] Equipment configuration module, configures corresponding drone equipment and image acquisition equipment according to the patrol area and frequency;

[0050] The planning module formulates an inspection plan based on the location and environmental characteristics of the bulk grain system line, including the inspection time, route and key inspection contents;

[0051] The image acquisition module uses drones to conduct regular inspections of the bulk grain system line and its surrounding environment to obtain corresponding image information;

[0052] The image analysis module analyzes the image information to determine whether there are any abnormalities in the bulk grain system line and the surrounding environment.

[0053] Furthermore, the human-computer interaction interface includes:

[0054] A parameter acquisition module acquires the operating status of the bulk grain equipment and the image information of the bulk grain system line and the surrounding environment, sets the operating status of the bulk grain equipment and the image information of the bulk grain system line and the surrounding environment as visual display elements, and determines the display weight of the visual display elements based on the visual display requirements;

[0055] A display configuration module extracts element features of visual display elements and allocates a display interface area for each visual display element in the human-computer interaction interface based on the element features and the display weight of the visual display element. At the same time, an element name is generated for each display interface area based on the element features, and the display interface area is marked based on the element name;

[0056] Determine the positional relationship between each bulk grain equipment and its surrounding environment and the bulk grain system line, and add visual association paths for each display interface area in the human-computer interaction interface based on the positional association relationship;

[0057] Based on the visual association path, the operating status of the bulk grain equipment and the image information of the surrounding environment are visualized with the image information of the bulk grain system line at the same location and frequency;

[0058] The human-computer interaction configuration module adds preset interactive actions to the human-computer interaction interface on the background configuration end, and determines the response state of the visual display screen under each preset interactive action;

[0059] Based on the response status of the visualization display screen under each preset interactive action, a multi-touch interaction mechanism is added to the visualization display results of the same position and frequency. Based on the multi-touch interaction mechanism, independent visualization display configuration is performed in the background configuration section for the operating status of the bulk grain equipment, the image information of the surrounding environment, and the image information of the bulk grain system line;

[0060] Based on the independent visualization display configuration, the operating status of the bulk grain equipment and the image information of the bulk grain system line and the surrounding environment are visualized.

[0061] Compared with the prior art, the present invention has the following beneficial effects:

[0062] 1. The present invention uses a neural network model to predict the operating status of bulk grain equipment, and combines it with drones to inspect the bulk grain system line and the surrounding environment to determine whether there are any abnormalities. When an abnormality is found, an alarm can be promptly sent to the human-computer interaction interface to remind management personnel to take timely remedial measures. This effectively realizes the early warning function of bulk grain equipment, prevents the expansion of bulk grain equipment failures, and avoids the hidden dangers of fire and dust explosion in the surrounding environment. It also reduces the labor intensity of inspection personnel, reduces labor costs, and improves the intelligence of existing agricultural development.

[0063] 2. By determining the display weights of the operating status of the bulk grain equipment and the image information of the bulk grain system line and the surrounding environment, corresponding display interface areas are allocated to different display contents in the human-computer interaction interface according to the display weights, and corresponding element names are generated and area labels are performed based on the element characteristics of each display interface area, so that different display interface areas can display corresponding content. Secondly, the positional association relationship between each bulk grain equipment and the surrounding environment and the bulk grain system line is determined, and different display interface areas are associated according to the positional association relationship. This facilitates the same-position and same-frequency visualization of the operating status of the bulk grain equipment and the image information of the surrounding environment and the image information of the bulk grain system line during drone interaction, facilitating effective understanding and viewing of different dimensional information at the same location through the human-computer interaction interface. Finally, preset interactive actions are added to the human-computer interaction interface on the background configuration end, so that the display content to be viewed can be retrieved and independently visualized through the preset interactive actions, improving the effect of human-computer interaction and facilitating users to quickly view the corresponding display content as needed.

[0064] 3. By obtaining the first real-time operating status and the second real-time operating status respectively, the comprehensiveness and accuracy of the bulk grain equipment monitoring results can be effectively guaranteed, thereby improving the accuracy of the early warning, and further ensuring the objectivity, accuracy and reliability of the generation of the assessment report. BRIEF DESCRIPTION OF THE DRAWINGS

[0065] Figure 1 This is a diagram showing the module composition of the multi-sensor AI early warning system applied to the bulk grain system line of the present invention. DETAILED DESCRIPTION

[0066] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0067] In order to solve the technical problem that the current alarm system, vibration alarm system, optical fiber temperature measurement and thermal imaging alarm systems in the existing technology are all operated independently, and the individual systems always make inaccurate judgments, which brings great trouble to the operators and affects the use effect, please refer to Figure 1 , this embodiment provides the following technical solutions:

[0068] The multi-sensor AI early warning system for bulk grain production lines includes:

[0069] The data acquisition unit collects the working status of the bulk grain equipment in real time through sensor equipment, obtains historical and real-time operating data of the bulk grain equipment, such as pressure, vibration, temperature, humidity, particle size, and dust content in the air, and feeds back various operating data to the centralized processing unit for pre-processing; the data acquisition unit includes:

[0070] The equipment layout module installs a variety of sensor devices on the scraper conveyor, bucket elevator and belt conveyor of the bulk grain equipment. The sensor devices include: weight sensors, vibration sensors, temperature sensors, humidity sensors, particle size sensors and dust sensors. Specifically, the appropriate sensor type and quantity are selected according to the parameters to be detected, and the installation location is determined. In this embodiment, for example, the weight sensor can be installed at the bucket elevator port, discharge port and other places, the temperature sensor can be installed at the feed port, discharge port and other places, and the particle size sensor can be installed at the feed port, discharge port and other places. Then, the corresponding interface is designed according to the sensor type and quantity to connect the sensor to the early warning system. After installing and configuring the sensor equipment, the system needs to be tested and verified to ensure that the early warning system can accurately detect and predict the operating status of the bulk grain equipment. The test includes: testing under normal working conditions and abnormal conditions to ensure the stability and reliability of the early warning system in actual application.

[0071] The timing feedback module sets a timing feedback mechanism for each sensor device, so that after the sensor device enters the real-time collection state, it regularly feeds back the real-time operation data of the bulk grain equipment to the centralized processing unit; specifically, by designing a timing feedback mechanism based on the data sampling frequency of the sensor, the sensor device automatically enters the real-time collection state after starting or running for a period of time, and regularly sends data feedback to the centralized processing unit; for example: according to actual needs, the feedback frequency of the sensor device is set to feedback every 60 minutes; if subsequent analysis finds that there is an abnormality in the operation state of the bulk grain equipment, the originally set feedback frequency can be adjusted to feedback every 30 minutes, until the bulk grain equipment fault is resolved, and then the data feedback frequency is adjusted to the originally set feedback every 60 minutes.

[0072] The centralized processing unit receives and pre-processes the historical and real-time operating data of the bulk grain equipment. Pre-processing includes cleaning, denoising, de-spurious processing, and formatting to facilitate subsequent data analysis and modeling. Feature information is extracted from the pre-processed historical operating data, such as vibration spectrum, particle velocity, particle size, particle shape, and temperature distribution, as indicators reflecting the historical operating status of the bulk grain equipment. The centralized processing unit includes:

[0073] The data cleaning module removes missing values, duplicate values, and outliers from the operating data, as well as noise and unnecessary redundant information from the operating data. For example, a low-pass filter is used to remove noise from the operating data, and the operating data is formatted for conversion. This effectively removes outliers and noise from the operating data, improving the accuracy and reliability of the operating data. At the same time, it facilitates the extraction of effective and useful feature information to improve the quality of the operating data analysis results.

[0074] The feature extraction module analyzes the vibration spectrum, particle speed and size, particle density and flow rate, and temperature distribution of the bulk grain equipment to extract characteristic information of the operating status of the bulk grain equipment. By analyzing the vibration signal generated by the bulk grain equipment during operation, information such as the vibration frequency, intensity, and phase of the equipment can be obtained, thereby reflecting the operating status and stability of the bulk grain equipment. By measuring and analyzing the particles generated by the bulk grain equipment during operation, information such as the speed, size, and shape of the particles can be obtained, thereby reflecting the conveying capacity and particle characteristics of the bulk grain equipment. By analyzing the temperature signal generated by the bulk grain equipment during operation, information such as the temperature distribution and change trend of the bulk grain equipment can be obtained, thereby reflecting the operating status and thermal balance of the bulk grain equipment. By analyzing the particle density and flow rate signals generated by the bulk grain equipment during operation, information such as the material conveying rate and conveying efficiency of the bulk grain equipment can be obtained, thereby reflecting the operating status and production capacity of the bulk grain equipment.

[0075] The data storage module classifies and organizes the received operating data by category, and stores the organized operating data by time series and category to form a database; specifically, when collecting the operating data of the bulk grain equipment, the timestamp of the data is recorded synchronously; the pre-processed data is classified by category, such as temperature, humidity, and weight; the data after classification and time series analysis is stored in the database, and different tables are established to store different data, such as temperature table, humidity table and weight table, etc., and each table includes the timestamp of the data and the corresponding category information; at the same time, in order to facilitate later analysis and query, functions such as indexes and views can also be added to the database to improve the effectiveness of the database.

[0076] The model building unit builds a neural network model. Based on the scale of the operating data, it selects the appropriate network architecture, activation function, loss function, and optimizer. It uses historical operating data and extracted feature information as input sets and the historical operating status of the bulk grain equipment as labels. It uses cross-validation technology to train the neural network model and adjusts the parameters of the neural network model to make the neural network model as close to the actual operating status of the bulk grain equipment as possible. The model building unit includes:

[0077] The sample partitioning module divides the historical operation data and the extracted feature information into a training set and a test set. The data in the training set will be used to train the neural network model, while the data in the test set will be used to verify the neural network model, such as indicators such as accuracy, precision, and recall rate. If the performance of the model does not meet the requirements, it is necessary to adjust the parameters of the training model or add new features.

[0078] The label matching module collects the historical operating status of bulk grain equipment in the same period as the historical operating data; sets a label for each type of historical operating status, and then matches the label with each type of historical operating status and historical operating data respectively; specifically, according to the time information and operating status information in the historical operating data, corresponding labels are set for different operating states, so that each operating state corresponds to a unique label, such as "normal operation", "fault shutdown" or "maintenance", etc.; by traversing the historical operating data to match the content of each label and the location where it appears, the correlation between the label and the location where it appears is analyzed; for example: all bulk grain equipment operations or other events associated with a specific label are recorded to better understand the status and behavior patterns of the bulk grain equipment, thereby predicting future operating status.

[0079] The model building module takes the training set as the input set, builds and trains the neural network model based on the input set, and establishes the feature layer and learning process of the neural network model; specifically, the samples in the training set are input into the neural network for forward propagation to obtain the output result of the neural network; and the output result is used as the input of the loss function to calculate the loss value; after calculating the loss value, the backpropagation algorithm is used to calculate the updated value of the neural network model parameters, so that the output result of the neural network model is closer to the real result.

[0080] The intelligent analysis unit uses a trained neural network model to analyze and predict the historical operating data of the bulk grain equipment to obtain an estimated state value of the bulk grain equipment. Based on the estimated state value, it determines whether the bulk grain equipment is in normal working condition and whether there are any abnormal conditions. The judgment result is compared with the historical operating status of the bulk grain equipment to verify the accuracy of the neural network model prediction. The intelligent analysis unit includes:

[0081] The model verification module inputs the test set into the trained neural network model for prediction and outputs the state estimation value of the bulk grain equipment; and compares the state estimation value with the actual result to verify whether the prediction ability of the neural network model is accurate; specifically, the data of the test set is input into the trained neural network model for prediction, and then the prediction result output by the neural network model is compared with the actual result. If the prediction result is completely consistent with the actual result, it means that the prediction ability of the neural network model is relatively accurate; otherwise, it means that there is an error in the neural network model, and the source of the error needs to be further analyzed, and the neural network model needs to be corrected or improved; after multiple iterations and improvements, until the error between the prediction result of the neural network model and the actual result reaches an acceptable level.

[0082] The result generation module determines whether the bulk grain equipment is in normal working condition and whether there are any abnormal conditions based on the state estimation value of the bulk grain equipment, and generates corresponding prediction results. Specifically, the working condition of the bulk grain equipment is judged based on the predicted state estimation value of the bulk grain equipment, combined with predefined normal working condition rules and abnormal condition judgment criteria. If the state estimation value of the bulk grain equipment meets the normal working condition rules and no abnormal conditions are detected, it can be considered that the bulk grain equipment is in normal working condition and the corresponding prediction result is output; otherwise, the opposite is true.

[0083] The model optimization module regularly uses real-time operation data to update and iterate the neural network model to optimize the parameters and performance of the neural network model.

[0084] The model deployment unit deploys the verified neural network model into the actual bulk grain production system to predict the real-time operating status of the bulk grain equipment. The model deployment unit includes:

[0085] The model deployment module exports the verified neural network model into an executable format, such as ONNX, TensorFlow, or Caffe. Based on the requirements of the neural network model and the characteristics of the system, the neural network model is deployed to the cloud platform where the human-computer interaction interface is installed.

[0086] The timing prediction module regularly predicts the operating status of the bulk grain equipment according to the feedback frequency of the real-time operating data. Specifically, following the above embodiment, for example, the feedback frequency of the sensor equipment is set to be fed back once every 60 minutes, and the prediction frequency of the neural network model is also set to be predicted once every 60 minutes. For another example, if analysis shows that the operating status of the bulk grain equipment is abnormal, the originally set feedback frequency can be adjusted to be fed back once every 30 minutes, and the prediction frequency of the neural network model is synchronously adjusted to be predicted once every 30 minutes. After the failure of the bulk grain equipment is resolved, the data feedback frequency and the prediction frequency of the neural network model are adjusted to their original positions.

[0087] The early warning prompt module promptly sends an early warning prompt to the human-computer interaction interface when the neural network model predicts that the bulk grain equipment and the drone detects an abnormal situation in the surrounding environment, reminding management personnel to check and repair. Specifically, if the estimated state value of the bulk grain equipment does not meet the normal working state rules, or any abnormal situation is detected, such as machine failure, environmental pollution, etc., it is necessary to intervene in the state of the bulk grain equipment and output the corresponding early warning results. The early warning results can be output in different ways according to different situations, such as: issuing an alarm to the human-computer interaction interface, recording a log, or sending an alarm message.

[0088] In one embodiment, the multi-sensor AI early warning system applied to the bulk grain system line, the model deployment unit, further includes:

[0089] The model operation module performs real-time analysis based on the neural network model to determine the first real-time operating status of the bulk grain equipment;

[0090] The monitoring module deploys a monitoring mechanism and calculates the operation status of the bulk grain equipment in real time based on the monitoring mechanism to obtain a second real-time operating status;

[0091] Wherein, obtaining the second real-time operating status includes:

[0092] Collect the real-time measurement values of each sensor in each sensor type, and calculate the comprehensive measurement value of each type of bulk grain equipment based on the real-time measurement values of each sensor;

[0093] ;

[0094] in, Indicates the Comprehensive measurement values of bulk grain equipment for each sensor under different sensor types; Indicates the number of sensors; Indicates the total number of sensors; Indicates the Sensor type The initial weight corresponding to each sensor; Indicates the Sensor type Dynamic weight adjustment factor corresponding to each sensor; Indicates the Sensor type Real-time measurement values of sensors; The serial number value indicating the sensor type;

[0095] Calculate the risk assessment value of bulk grain equipment based on the comprehensive measurement values of the bulk grain equipment.

[0096] ;

[0097] in, Indicates the risk assessment value of bulk grain equipment; Indicates the total number of sensor types; Indicates taking the maximum value; Indicates the Safety measurement thresholds corresponding to different sensor types; Indicates the Risk weights corresponding to sensor types;

[0098] Determine the second real-time operating status based on the risk assessment value of the bulk grain equipment. hour, , it means that the current bulk grain equipment is within the safe range; when When , it means that the current bulk grain equipment is in the abnormal range;

[0099] The report generating module is used to generate an evaluation report based on the first real-time operating status and the second real-time operating status, and transmit the evaluation report to the user terminal.

[0100] In this embodiment, the dynamic weight adjustment factor can be dynamically adjusted based on factors such as the historical accuracy and real-time stability of the sensor. For example, for a sensor with a small historical measurement error and stable recent data fluctuations, its It can take a positive value to increase the weight of the sensor data in the fusion; otherwise, it takes a negative value.

[0101] In this embodiment, the first real-time operating state is the operating state of the bulk grain equipment obtained after real-time analysis based on the neural network model; the second real-time operating state is the operating state of the bulk grain equipment obtained after calculation based on the sensor data collected in real time.

[0102] In this embodiment, the user terminal includes but is not limited to a mobile phone, a computer, etc.

[0103] The working principle and beneficial effect of the above technical solution are: by obtaining the first real-time operating status and the second real-time operating status respectively, the comprehensiveness and accuracy of the bulk grain equipment monitoring results can be effectively guaranteed, thereby improving the accuracy of the early warning, and then ensuring the objectivity, accuracy and reliability of the generation of the evaluation report.

[0104] The environmental inspection unit uses drones to regularly inspect the bulk grain system and its surroundings, obtains image information of the bulk grain system and its surroundings, and determines whether there are any abnormal hazards in the bulk grain system and its surroundings. The environmental inspection unit includes:

[0105] The equipment configuration module configures corresponding drone equipment and image acquisition equipment, such as high-definition cameras, infrared detectors, and laser scanners, according to the patrol area and frequency, to ensure that the acquired image information is accurate.

[0106] The planning module formulates a detailed inspection plan based on the location and environmental characteristics of the bulk grain system line, including: the time, route and key inspection content of the inspection; secondly, the scope and frequency of the inspection are determined based on factors such as the location, length and environmental characteristics of the bulk grain system line to be inspected; for example: if the bulk grain system line is too long, it can be divided into several sections for inspection; if the environment around the bulk grain system line is complex and changeable, the frequency of inspection can be increased; and during the actual inspection process, the inspection results are recorded.

[0107] The image acquisition module uses drones to conduct regular inspections of the bulk grain system line and its surrounding environment to obtain corresponding image information. The image information includes: the appearance status of the equipment in the bulk grain system line and the meteorological information of the surrounding environment.

[0108] The image analysis module analyzes image information to determine whether there are any abnormalities in the bulk grain system line and the surrounding environment. For example, based on the image information, it can determine the degree of damage to the equipment in the bulk grain system line, the risk of leakage, and environmental pollution. If an abnormality is detected, relevant personnel should be notified in a timely manner to take measures to ensure production safety and environmental protection. Based on this, image information of the bulk grain system line can be effectively obtained, so as to promptly discover and respond to abnormal situations and ensure production safety and environmental protection.

[0109] The human-computer interaction interface visually displays the operating status of the bulk grain equipment and image information of the bulk grain system line and the surrounding environment, so that managers can intervene and manage in a timely manner.

[0110] The beneficial effects achieved by the above content: Through the above operations, the early warning function of the bulk grain system line is effectively realized, the expansion of bulk grain equipment failures is prevented, and the hidden dangers of fire and dust explosion in the surrounding environment are avoided, the labor intensity of patrol personnel is reduced, labor costs are reduced, and the intelligence of existing agricultural development is improved.

[0111] Working Principle: By using historical operating data to train and test the neural network model, a model that can predict the operating status of bulk grain equipment is obtained; the neural network model is applied to the actual bulk grain system line to make real-time predictions on the operating status of the bulk grain equipment, and combined with drone inspections of the bulk grain system line and the surrounding environment to check for abnormalities; when an abnormal situation is found, an alarm can be promptly sent to the human-computer interaction interface to remind management personnel to take timely solutions.

[0112] In one embodiment, a multi-sensor AI early warning system and human-computer interaction interface for a bulk grain production line includes:

[0113] a parameter acquisition module for acquiring the operating status of the bulk grain equipment and the image information of the bulk grain system line and the surrounding environment, setting the operating status of the bulk grain equipment and the image information of the bulk grain system line and the surrounding environment as visual display elements, and determining the display weight of the visual display elements based on the visual display requirements;

[0114] Display configuration module for:

[0115] Extracting element features of the visual display elements, and assigning a display interface area to each visual display element in the human-computer interaction interface based on the element features and the display weight of the visual display elements; generating an element name for each display interface area based on the element features, and marking the display interface area based on the element name;

[0116] Determine the positional relationship between each bulk grain equipment and its surrounding environment and the bulk grain system line, and add visual association paths for each display interface area in the human-computer interaction interface based on the positional association relationship;

[0117] Based on the visual association path, the operating status of the bulk grain equipment and the image information of the surrounding environment are visualized with the image information of the bulk grain system line at the same location and frequency;

[0118] Human-computer interaction configuration module, used for:

[0119] Add preset interactive actions to the human-computer interaction interface on the background configuration side, and determine the response state of the visual display screen under each preset interactive action;

[0120] Based on the response status of the visualization display screen under each preset interactive action, a multi-touch interaction mechanism is added to the visualization display results of the same position and frequency. Based on the multi-touch interaction mechanism, independent visualization display configuration is performed in the background configuration section for the operating status of the bulk grain equipment, the image information of the surrounding environment, and the image information of the bulk grain system line;

[0121] Based on the independent visualization display configuration, the operating status of the bulk grain equipment and the image information of the bulk grain system line and the surrounding environment are visualized.

[0122] In this embodiment, the visual display elements are the operating status of the bulk grain equipment and the image information of the bulk grain system line and the surrounding environment, that is, the content that needs to be displayed on the human-computer interaction interface.

[0123] In this embodiment, the display weight of the visual display element refers to the importance of different visual display elements when displaying each determined visual display element. The larger the value, the more important it is in the visual display. It is related to the position and area allocated when allocating the display interface area.

[0124] In this embodiment, the element feature refers to the type of the visual display element and the object that needs to be displayed during the visual display.

[0125] In this embodiment, the display interface area refers to the display area allocated in the human-computer interaction interface for the operating status of the bulk grain equipment and the image information of the bulk grain system line and the surrounding environment. For example, when allocating, the display interface area of the visual display element with a greater display weight is closer to the center.

[0126] In this embodiment, the element name is used to mark the type of content displayed in each display interface area.

[0127] In this embodiment, the position association relationship is used to characterize the correspondence between the bulk grain equipment and the surrounding environment and each position in the bulk grain system line. The purpose is to display the same position of the three at the same frequency during visual display.

[0128] In this embodiment, the visual association path is used to represent the corresponding relationship between the visual display elements in each display interface area.

[0129] In this embodiment, the same-position and same-frequency visual display means that the operating status of the bulk grain equipment and the image information of the surrounding environment change with the image information of the bulk grain system line. For example, when the image information of the bulk grain system line displays a certain equipment, the display interface area displaying the operating status of the bulk grain equipment displays the operating status of the equipment, and the display interface area displaying the image information of the surrounding environment displays the image information of the surrounding environment of the equipment.

[0130] In this embodiment, the preset interactive action is set in advance, for example, it can be a click or slide operation.

[0131] In this embodiment, the response state of the visual display screen refers to the situation where the display screen needs to change under each preset interactive action. For example, when the user clicks on the bulk grain equipment that needs to be viewed, the screen needs to immediately change to the corresponding equipment screen.

[0132] In this embodiment, the multi-touch interaction mechanism refers to a mechanism for adding a separate display to each display content, that is, associating each display content with a preset interactive action, in order to ensure that the corresponding display content can be called through the preset interactive action.

[0133] In this embodiment, independent visualization display configuration refers to configuring independent display for each display content, that is, displaying the content that needs to be viewed according to the user's interaction needs without affecting the normal display of other screens.

[0134] The working principle and beneficial effects of the above technical solution are as follows: by determining the display weights of the operating status of the bulk grain equipment and the image information of the bulk grain system line and the surrounding environment, corresponding display interface areas are allocated to different display contents in the human-computer interaction interface according to the display weights, and corresponding element names are generated and area markings are performed according to the element characteristics of each display interface area, so that different display interface areas can display corresponding content. Secondly, the positional association relationship between each bulk grain equipment and the surrounding environment and the bulk grain system line is determined, and different display interface areas are associated according to the positional association relationship, so that when the drone is interactively operated, the image information of the operating status of the bulk grain equipment and the surrounding environment and the image information of the bulk grain system line are visually displayed at the same location and frequency, which facilitates the effective understanding and viewing of different dimensional information at the same location through the human-computer interaction interface. Finally, preset interactive actions are added to the human-computer interaction interface on the background configuration end, so that the display content to be viewed can be retrieved and independently visualized through the preset interactive actions, thereby improving the effect of human-computer interaction and facilitating users to quickly view the corresponding display content as needed.

[0135] It should be noted that, in this document, relational terms such as first and second, etc., are used only to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any actual relationship or order between these entities or operations. Moreover, the terms "including," "having," or any other variations thereof are intended to cover non-exclusive possessors, such that a process, method, article, or apparatus that includes a list of elements includes not only those elements but also other elements not explicitly listed, or includes elements that are inherent to such process, method, article, or apparatus.

[0136] While embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that changes, modifications, substitutions, and alterations may be made to the embodiments without departing from the principles and spirit of the invention, and that the scope of the invention is defined by the appended claims and their equivalents.

Claims

1. The multi-sensor AI early warning system applied to the bulk grain system line is characterized by: include: Data collection unit, which obtains historical and real-time operation data of bulk grain equipment; A centralized processing unit pre-processes the acquired historical operation data and real-time operation data and extracts feature information from the historical operation data; The model building unit builds and trains a neural network model based on feature information and adjusts the parameters of the neural network model to make it close to the actual operating status of the bulk grain equipment; The intelligent analysis unit uses a neural network model to predict historical operating data and determine whether the bulk grain equipment is operating normally; And compare the judgment results with the actual results to verify whether the neural network model is accurate; A model deployment unit deploys the verified neural network model to an actual bulk grain production system to predict the real-time operating status of the bulk grain equipment, including a first real-time operating status and a second real-time operating status; The environmental inspection unit uses drone equipment to obtain image information of the bulk grain system line and the surrounding environment; Model deployment unit, including: The model operation module performs real-time analysis based on the neural network model to determine the first real-time operating status of the bulk grain equipment; The monitoring module deploys a monitoring mechanism and calculates the operation status of the bulk grain equipment in real time based on the monitoring mechanism to obtain a second real-time operating status; Wherein, obtaining the second real-time operating status includes: Collect the real-time measurement values of each sensor in each sensor type, and calculate the comprehensive measurement value of each type of bulk grain equipment based on the real-time measurement values of each sensor; the calculation of the comprehensive measurement value is as follows: ； in, Indicates the Comprehensive measurement values of bulk grain equipment for each sensor under different sensor types; Indicates the number of sensors; Indicates the total number of sensors; Indicates the Sensor type The initial weight corresponding to each sensor; Indicates the Sensor type Dynamic weight adjustment factor corresponding to each sensor; Indicates the Sensor type Real-time measurement values of sensors; The serial number value indicating the sensor type; Based on the comprehensive measurement values of bulk grain equipment, the risk assessment value of bulk grain equipment is calculated; the risk assessment value is calculated as follows: ； in, Indicates the risk assessment value of bulk grain equipment; Indicates the total number of sensor types; Indicates taking the maximum value; Indicates the Safety measurement thresholds corresponding to different sensor types; Indicates the Risk weights corresponding to sensor types; Determine the second real-time operating status based on the risk assessment value of the bulk grain equipment. hour, , it means that the current bulk grain equipment is within the safe range; when When , it means that the current bulk grain equipment is in the abnormal range; a report generating module, configured to generate an evaluation report based on the first real-time operating status and the second real-time operating status, and transmit the evaluation report to a user terminal; A human-computer interaction interface visually displays the operating status of the bulk grain equipment and the image information of the bulk grain system line and the surrounding environment; the display configuration process of the operating status of the bulk grain equipment and the image information of the bulk grain system line and the surrounding environment includes: based on a multi-touch interaction mechanism, independent visual display configuration is performed for the operating status of the bulk grain equipment, the image information of the surrounding environment, and the image information of the bulk grain system line in the background configuration segment.

2. The multi-sensor AI early warning system for bulk grain production lines according to claim 1 is characterized by: The centralized processing unit includes: Data cleaning module, which removes missing values, duplicate values, and outliers from the running data, as well as noise and unnecessary redundant information, and converts the format of the running data; The feature extraction module analyzes the vibration spectrum, particle velocity and size, particle density and flow rate, and temperature distribution of the bulk grain equipment to extract characteristic information about the bulk grain equipment's operating status; The data storage module classifies and organizes the received operation data by category, and stores the organized operation data by time series and category to form a database.

3. The multi-sensor AI early warning system for bulk grain production lines according to claim 1 is characterized by: The model building unit includes: The sample partitioning module divides the historical operation data and the extracted feature information into training sets and test sets; The tag matching module collects the historical operating status of bulk grain equipment in the same period as the historical operating data; sets a tag for each type of historical operating status, and then matches the tag with each type of historical operating status and historical operating data; The model building module takes the training set as the input set, builds and trains the neural network model based on the input set, and establishes the feature layer and learning process of the neural network model.

4. The multi-sensor AI early warning system for bulk grain production lines according to claim 1 is characterized by: The intelligent analysis unit includes: The model verification module inputs the test set into the trained neural network model for prediction and outputs the estimated state value of the bulk grain equipment. The estimated state value is then compared with the actual result to verify the accuracy of the neural network model's prediction ability. The result generation module determines whether the bulk grain equipment is in normal working condition and whether there are any abnormal conditions based on the estimated state value of the bulk grain equipment, and generates corresponding prediction results.

5. The multi-sensor AI early warning system for bulk grain production lines according to claim 4 is characterized by: The intelligent analysis unit further includes: The model optimization module regularly uses real-time operation data to update and iterate the neural network model to optimize the parameters and performance of the neural network model.

6. The multi-sensor AI early warning system for bulk grain production lines according to claim 1 is characterized by: The model deployment unit further includes: The model deployment module exports the verified neural network model into an executable format and deploys the neural network model to the cloud platform loaded with the human-computer interaction interface based on the requirements of the neural network model and system characteristics; The timing prediction module regularly predicts the operating status of bulk grain equipment based on the feedback frequency of real-time operating data; The early warning prompt module sends a warning prompt to the human-computer interaction interface in a timely manner when the neural network model predicts that the bulk grain equipment and the drone detects abnormal conditions in the surrounding environment, reminding management personnel to check and repair.

7. The multi-sensor AI early warning system for bulk grain production line according to claim 1 is characterized in that: Human-computer interaction interface, including: A parameter acquisition module acquires the operating status of the bulk grain equipment and the image information of the bulk grain system line and the surrounding environment, sets the operating status of the bulk grain equipment and the image information of the bulk grain system line and the surrounding environment as visual display elements, and determines the display weight of the visual display elements based on the visual display requirements; A display configuration module extracts element features of visual display elements and allocates a display interface area for each visual display element in the human-computer interaction interface based on the element features and the display weight of the visual display element. At the same time, an element name is generated for each display interface area based on the element features, and the display interface area is marked based on the element name; Determine the positional relationship between each bulk grain equipment and its surrounding environment and the bulk grain system line, and add visual association paths for each display interface area in the human-computer interaction interface based on the positional association relationship; Based on the visual association path, the operating status of the bulk grain equipment and the image information of the surrounding environment are visualized with the image information of the bulk grain system line at the same location and frequency; The human-computer interaction configuration module adds preset interactive actions to the human-computer interaction interface on the background configuration end, and determines the response state of the visual display screen under each preset interactive action; Based on the response status of the visualization display screen under each preset interactive action, a multi-touch interaction mechanism is added to the visualization display results of the same position and frequency. Based on the multi-touch interaction mechanism, independent visualization display configuration is performed in the background configuration section for the operating status of the bulk grain equipment, the image information of the surrounding environment, and the image information of the bulk grain system line; Based on the independent visualization display configuration, the operating status of the bulk grain equipment and the image information of the bulk grain system line and the surrounding environment are visualized.

Images