Fan frosting recognition and intelligent defrosting system and method based on AI vision

Abstract

The invention discloses a fan frosting recognition and intelligent defrosting system and method based on AI vision, and belongs to the technical field of computer vision. The system comprises a frosting detection module, a defrosting control module, an image acquisition module, a monitoring area setting module, a data processing module and a system feedback module. The method comprises the following steps of system initialization, frosting detection, defrosting control and system feedback and optimization. The method has important significance in the aspects of energy conservation, emission reduction and energy efficiency improvement, meanwhile, a brand new technical solution and an operation mode are brought for refrigeration house management, accurate monitoring and on-demand defrosting of the refrigeration house fan fin frosting condition can be achieved, the energy consumption of a refrigeration house is effectively reduced, the service life of equipment is prolonged, and the operation efficiency of the equipment is improved. The system has good adaptability and flexibility, can be widely applied to various cold chain facilities, and provides advanced technical support for cold storage management.

Description

Technical Field

[0001] The present invention relates to the field of computer vision technology, and in particular to a fan frost recognition and intelligent defrosting system and method based on AI vision. Background Art

[0002] In the operation and management of cold storage, the defrosting process is one of the key operations. Traditional defrosting methods include manual defrosting and automatic defrosting based on a fixed time cycle (such as PLC program setting). These methods have some common problems, such as energy waste and increased equipment loss, because they lack accurate judgment of the actual frosting situation.

[0003] Manual defrosting relies on manual visual judgment. Workers regularly check the frost in the cold storage and perform defrosting operations. However, due to the subjectivity and uncertainty of manual judgment, defrosting is often performed too early or too late, which wastes energy and increases equipment maintenance costs.

[0004] On the other hand, although the automatic defrosting method based on a fixed time cycle solves the time uncertainty problem of manual defrosting, its disadvantage is that it cannot be adjusted according to the real-time frosting situation. This method may result in energy-intensive defrosting operations when the frosting is not serious or defrosting is not required, thereby reducing the energy efficiency ratio and operating efficiency of the cold storage. Summary of the invention

[0005] The purpose of the present invention is to provide a fan frost recognition and intelligent defrosting system and method based on AI vision. Through AI technology, the system can monitor the frosting condition of the fan fin surface in the cold storage in real time and accurately analyze the degree and area of ​​frost. Based on these data, the system intelligently determines whether defrosting operation is required and adjusts the defrosting strategy according to actual needs.

[0006] To achieve the above-mentioned objectives, the present invention provides a fan frost recognition and intelligent defrost system based on AI vision, including a frost detection module and a defrost control module, the frost detection module and the defrost control module are both connected to a data processing module, the data processing module and the defrost control module are both connected to a system feedback module, the frost detection module is connected to a monitoring area setting module, and the monitoring area setting module is connected to an image acquisition module.

[0007] Preferably, the frost detection module is used to analyze and segment the image of the fan fins, accurately identify areas with different degrees of frost, and receive real-time image data from the image acquisition module inside the cold storage to detect and analyze the frost situation and provide real-time frost degree information;

[0008] The defrost control module is used to determine whether to start the defrost operation according to the real-time data provided by the frosting detection module;

[0009] The image acquisition module is used to obtain real-time image data from the camera inside the cold storage and transmit it to the frost detection module for analysis;

[0010] The monitoring area setting module is used to adjust the size and position of the monitoring area as needed, allowing the user to draw and set the monitoring area of ​​the fan fins;

[0011] The data processing module processes and analyzes the recognition result of the image acquisition module, generates real-time frost degree data, and transmits it to the defrost control module;

[0012] The system feedback module is responsible for recording and analyzing data during the system operation process, providing feedback and historical records of defrosting operations, and being able to monitor the system operation status in real time.

[0013] A method for fan frost recognition and intelligent defrosting based on AI vision, comprising the following steps:

[0014] S1. System initialization. After the system is started, the initial configuration is performed first, including the setting of the cold storage monitoring area, the calibration of the camera, and the loading of the AI ​​model;

[0015] S2, frost detection, after the system enters the running state, the frost situation inside the cold storage is detected in real time through image acquisition, frost recognition and data processing;

[0016] S3, defrost control: according to the frost detection results, the system automatically decides whether to start the defrost operation based on the comparison between the frost degree detected in the monitoring area and the set threshold;

[0017] S4, system feedback and optimization: the system continuously optimizes the defrost strategy and improves system performance by recording and analyzing defrost operation data.

[0018] Preferably, in step S1, the calibration of the camera includes adjusting the position of the camera and correcting image distortion.

[0019] Preferably, in step S2, the specific steps of frost detection are as follows:

[0020] S21, image acquisition, the system obtains real-time image data from the camera inside the cold storage and transmits it to the frost detection module for analysis;

[0021] S22, frost recognition. The frost detection module uses the AI ​​model to segment and analyze the image and identify areas with different degrees of frost in the monitoring area. The system generates a detailed report on the frost situation based on the recognition results, including the distribution of areas with heavy frost, light frost and no frost.

[0022] S23, data processing, the system processes and analyzes the recognition results, generates real-time frost degree data, and transmits it to the defrost control module.

[0023] Preferably, in step S3, the specific steps of defrost control are as follows:

[0024] S31, start defrosting. When it is detected that the degree of frost in the monitoring area exceeds the set threshold, the system automatically starts the defrosting operation. The system starts to perform overall defrosting operation on the cold storage, and heats the fan fins through the heating device to remove the frost;

[0025] S32, real-time monitoring. During the defrosting process, the system continuously monitors the frost condition in the monitoring area and detects the frost change in real time through the camera to determine whether the defrosting effect meets expectations.

[0026] S33, ending the defrost ahead of schedule. When the system detects that the defrost effect reaches the set standard, it automatically ends the defrost operation ahead of schedule, analyzes the data of this defrost operation, and optimizes future defrost strategies.

[0027] Preferably, in step S4, the specific steps of system feedback and optimization are as follows:

[0028] S41, data recording, the system records the detailed data of each defrosting operation, including frosting conditions, defrosting time and energy consumption, and the data is stored in the system database for subsequent analysis and optimization;

[0029] S42. The system analyzes the recorded data to evaluate the effect and energy consumption of the defrost operation. Through data analysis, the system adjusts the defrost strategy and parameters to improve the accuracy of defrost and the energy-saving effect.

[0030] Therefore, the present invention adopts a fan frost recognition and intelligent defrosting system and method based on AI vision with the above structure, which has the following beneficial effects:

[0031] The present invention is not only of great significance in terms of energy conservation and emission reduction and improving energy efficiency, but also brings a new technical solution and operation mode to cold storage management, which can realize accurate monitoring of the frosting of the cold storage fan fins and defrost on demand, effectively reduce the energy consumption of the cold storage, and improve the service life and operation efficiency of the equipment. The system has good adaptability and flexibility, can be widely used in various cold chain facilities, and provides advanced technical support for cold storage management.

[0032] The technical solution of the present invention is further described in detail below through the accompanying drawings and embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0033] Figure 1 A schematic diagram of a fan frost identification and intelligent defrosting system and method based on AI vision according to the present invention;

[0034] Figure 2 It is a flow chart of a fan frost identification and intelligent defrosting system and method based on AI vision of the present invention;

[0035] Figure 3 The present invention is a schematic diagram of the division of areas by a monitoring area setting module in a fan frost recognition and intelligent defrosting system and method based on AI vision. DETAILED DESCRIPTION

[0036] The technical solution of the present invention is further described below through the accompanying drawings and embodiments.

[0037] Unless otherwise defined, the technical terms or scientific terms used in the present invention should be understood by people with ordinary skills in the field to which the present invention belongs. The words "first", "second" and similar words used in the present invention do not indicate any order, quantity or importance, but are only used to distinguish different components. "Include" or "comprise" and similar words mean that the elements or objects appearing before the word include the elements or objects listed after the word and their equivalents, without excluding other elements or objects. "Connect" or "connected" and similar words are not limited to physical or mechanical connections, but may include electrical connections, whether direct or indirect. "Up", "down", "left", "right" and the like are only used to indicate relative positional relationships. When the absolute position of the described object changes, the relative positional relationship may also change accordingly.

[0038] Example

[0039] like Figure 1 As shown, the present invention provides a fan frost recognition and intelligent defrost system and method based on AI vision, including a frost detection module and a defrost control module, the frost detection module and the defrost control module are both connected to the data processing module, the data processing module and the defrost control module are both connected to the system feedback module, the frost detection module is connected to the monitoring area setting module, and the monitoring area setting module is connected to the image acquisition module.

[0040] The frost detection module is used to analyze and segment the image of the fan fins, accurately identify areas with different degrees of frost, and receive real-time image data from the image acquisition module inside the cold storage to detect and analyze the frost situation and provide real-time frost degree information;

[0041] The defrost control module is used to determine whether to start the defrost operation based on the real-time data provided by the frost detection module;

[0042] The image acquisition module is used to obtain real-time image data from the camera inside the cold storage and transmit it to the frost detection module for analysis;

[0043] The monitoring area setting module is used to adjust the size and position of the monitoring area according to the needs, allowing the user to draw and set the monitoring area of ​​the fan fins;

[0044] The data processing module processes and analyzes the recognition results of the image acquisition module, generates real-time frost degree data, and transmits it to the defrost control module;

[0045] The system feedback module is responsible for recording and analyzing data during system operation, providing feedback and historical records of defrost operations, and monitoring the system operation status in real time.

[0046] like Figure 2 , Figure 3 As shown, a method for fan frost recognition and intelligent defrosting based on AI vision includes the following steps:

[0047] S1. System initialization. After the system is started, the initial configuration is performed first, including the setting of the cold storage monitoring area, the calibration of the camera, and the loading of the AI ​​model;

[0048] S2, frost detection, after the system enters the running state, the frost situation inside the cold storage is detected in real time through image acquisition, frost recognition and data processing;

[0049] S3, defrost control: according to the frost detection results, the system automatically decides whether to start the defrost operation based on the comparison between the frost degree detected in the monitoring area and the set threshold;

[0050] S4, system feedback and optimization: the system continuously optimizes the defrost strategy and improves system performance by recording and analyzing defrost operation data.

[0051] In step S1 , the calibration of the camera includes adjusting the position of the camera and correcting image distortion.

[0052] In step S2, the specific steps of frost detection are as follows:

[0053] S21, image acquisition, the system obtains real-time image data from the camera inside the cold storage and transmits it to the frost detection module for analysis;

[0054] S22, frost recognition. The frost detection module uses the AI ​​model to segment and analyze the image and identify areas with different degrees of frost in the monitoring area. The system generates a detailed report on the frost situation based on the recognition results, including the distribution of areas with heavy frost, light frost and no frost.

[0055] S23, data processing, the system processes and analyzes the recognition results, generates real-time frost degree data, and transmits it to the defrost control module.

[0056] In step S3, the specific steps of defrost control are as follows:

[0057] S31, start defrosting. When it is detected that the degree of frost in the monitoring area exceeds the set threshold, the system automatically starts the defrosting operation. The system starts to perform overall defrosting operation on the cold storage, and heats the fan fins through the heating device to remove the frost;

[0058] S32, real-time monitoring. During the defrosting process, the system continuously monitors the frost condition in the monitoring area and detects the frost change in real time through the camera to determine whether the defrosting effect meets expectations.

[0059] S33, ending the defrost ahead of schedule. When the system detects that the defrost effect reaches the set standard, it automatically ends the defrost operation ahead of schedule, analyzes the data of this defrost operation, and optimizes future defrost strategies.

[0060] In step S4, the specific steps of system feedback and optimization are as follows:

[0061] S41, data recording, the system records the detailed data of each defrosting operation, including frosting conditions, defrosting time and energy consumption, and the data is stored in the system database for subsequent analysis and optimization;

[0062] S42. The system analyzes the recorded data to evaluate the effect and energy consumption of the defrost operation. Through data analysis, the system adjusts the defrost strategy and parameters to improve the accuracy of defrost and the energy-saving effect.

[0063] The present invention builds a target detection and segmentation model, and the design of the AI ​​module includes the following contents:

[0064] Input Module:

[0065] Image preprocessing: Normalize, scale, and augment the input image to fit the model’s input size.

[0066] Data loading: Load image data and convert it into tensor format for model processing.

[0067] Feature extraction module:

[0068] Convolutional layer: Extracts the spatial features and semantic information of the image through a series of convolution operations.

[0069] Pooling layer: The pooling operation reduces the dimension of the feature map, reduces the amount of calculation, and retains important features.

[0070] Feature pyramid module:

[0071] FPN: Combines feature maps from different scales to capture multi-scale information, thereby improving the detection ability of small and large objects.

[0072] Prediction Module:

[0073] Anchor point generation: Generate anchor points on the feature map to locate the position of potential targets.

[0074] Classification and regression: predict the class probability and bounding box regression value for each anchor point.

[0075] Loss function module:

[0076] VFL Loss (Improved Cross Entropy Loss)

[0077]

[0078] Post-processing module:

[0079] Non-maximum suppression: remove overlapping prediction boxes and keep only the box with the highest score.

[0080] Bounding box adjustment: Adjust the position and size of the bounding box according to the regression value output by the model.

[0081] Segmentation module:

[0082] Segmentation head: A segmentation head is added based on target detection to generate segmentation masks of the target.

[0083] Mask prediction: The feature map is further processed to generate a pixel-level segmentation mask for each object.

[0084] Loss function: Calculate the loss of the segmentation mask using binary cross entropy loss.

[0085] Training and optimization module:

[0086] Optimizer: Adam (Adaptive Moment Estimation Algorithm) is used to update the model parameters.

[0087] Model evaluation module:

[0088] Evaluation indicators: accuracy, recall and F1 score are used to evaluate model performance.

[0089] Through the collaborative work of these sub-modules, target detection and segmentation tasks can be completed efficiently, and real-time image processing and analysis can be achieved.

[0090] The AI ​​model building process includes the following steps:

[0091] Data collection and preprocessing:

[0092] Data source: Collect image data of fan fins from a high-resolution camera inside the cold storage. Ensure that the images cover various frost conditions, including no frost, light frost, medium frost, and heavy frost.

[0093] Data annotation: Through manual annotation, the frosted area and degree of frosting are marked for each image.

[0094] Data augmentation: Perform operations such as rotating, flipping, cropping, and color adjustment on images to expand the dataset and increase the robustness of the model.

[0095] Model selection and training:

[0096] Model selection: Use self-designed object detection and segmentation models

[0097] Model architecture: The model consists of several convolutional layers, pooling layers, fully connected layers, and Softmax layers. The convolutional layers extract image features, the pooling layers reduce feature dimensions, the fully connected layers perform classification, and the Softmax layers output the probability distribution of the degree of frost.

[0098] Training process: Use the labeled data set for training, and optimize the model parameters by improving the cross entropy loss function and the Adam (adaptive moment estimation algorithm) optimizer. The training process includes forward propagation, loss calculation, back propagation, and parameter update.

[0099] Parameter formula:

[0100] Improved cross entropy loss formula

[0101]

[0102] Adam formula

[0103] m t =β1m t-1 +(1-β1)g t

[0104]

[0105] Model validation and optimization:

[0106] Validation set: Use the validation set to evaluate the performance of the model and calculate the precision, recall, and F1 score.

[0107] Hyperparameter tuning: Adjust hyperparameters such as learning rate, batch size, and model depth to optimize model performance.

[0108] Model testing: Perform a final evaluation on the test set to ensure the generalization ability of the model.

[0109] The technical effects brought by the present invention are as follows:

[0110] (1) Innovative frost recognition model:

[0111] The present invention has developed an advanced AI model that can accurately segment the image of the fan fins and identify in detail the areas of heavy frost, light frost and no frost. By drawing a specific monitoring area, the system can focus on the frost situation in the area, thereby avoiding the analysis of irrelevant areas and improving the efficiency and accuracy of detection.

[0112] (2) Intelligent monitoring area:

[0113] The system can draw and adjust the monitoring area according to actual needs, focusing on the key areas that need defrosting. This regional monitoring function avoids indiscriminate detection of the entire irrelevant area, ensuring the pertinence and effectiveness of frost detection.

[0114] (3) On-demand defrost strategy:

[0115] By using the AI ​​model to accurately identify the degree of frost in the monitored area, the system can intelligently decide whether to start the defrost operation and optimize the defrost time and frequency. This on-demand defrost strategy avoids the unnecessary operation of the traditional fixed-time defrost method, greatly reducing energy consumption and operating costs.

[0116] (4) Significantly reduce energy consumption:

[0117] By accurately detecting the frost in the monitoring area and initiating defrosting on demand, the system effectively reduces unnecessary defrosting operations. Ending the defrosting process early reduces the heat input in the cold storage, significantly improves energy efficiency, and reduces overall power consumption and refrigeration burden.

[0118] (5) Improve cold storage operation efficiency:

[0119] The application of the ΑΙ model makes the frost detection and defrosting process of the cold storage highly intelligent and automated, reducing manual intervention. It focuses on frost detection in the monitoring area, improves the accuracy and efficiency of cold storage operations, and optimizes the management process of cold storage.

[0120] (6) Extend equipment service life:

[0121] Through intelligent defrost control and specific area monitoring, the wear and thermal stress caused by excessive defrosting on the equipment is reduced. The system can accurately control the defrost operation to avoid frequent interference with the equipment, thereby extending the service life of the equipment and reducing maintenance costs.

[0122] (7) Technological innovation and market potential:

[0123] AI vision-based intelligent defrosting technology demonstrates significant technological innovation in cold storage management through innovative regional monitoring and on-demand defrosting strategies. This technology solves the problems of high energy consumption and low efficiency in traditional defrosting methods, and has broad market application prospects and commercial value, especially in the cold chain industry where energy conservation, emission reduction and intelligent management requirements are increasing, showing huge commercial potential.

[0124] (8) Environmental protection and sustainable development:

[0125] Through intelligent monitoring areas and on-demand defrosting strategies, the system effectively reduces the energy consumption and carbon emissions of cold storage, meeting the requirements of green environmental protection and sustainable development. This technology is of great significance in reducing carbon footprint and promoting the environmentally friendly management transformation of the cold chain industry.

[0126] (9) Convenient operation and optimized management:

[0127] The system's automated design and regional monitoring function simplify cold storage management operations and reduce the need for frequent manual inspections and adjustments. Through precise monitoring and intelligent adjustment, the system optimizes cold storage defrost management, improves operational stability and efficiency, and reduces management costs.

[0128] Finally, it should be noted that the above embodiments are only used to illustrate the technical solution of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that they can still modify or replace the technical solution of the present invention with equivalents, and these modifications or equivalent replacements cannot cause the modified technical solution to deviate from the spirit and scope of the technical solution of the present invention.

Claims

1. A fan frost recognition and intelligent defrosting system based on AI vision, characterized in that: It includes a frost detection module and a defrost control module, both of which are connected to a data processing module, both of which are connected to a system feedback module, the frost detection module is connected to a monitoring area setting module, and the monitoring area setting module is connected to an image acquisition module.

2. The AI ​​vision-based fan frost identification and intelligent defrosting system according to claim 1, characterized in that: The frost detection module is used to analyze and segment the image of the fan fins, accurately identify areas with different degrees of frost, and receive real-time image data from the image acquisition module inside the cold storage to detect and analyze the frost situation and provide real-time frost degree information; The defrost control module is used to determine whether to start the defrost operation according to the real-time data provided by the frosting detection module; The image acquisition module is used to obtain real-time image data from the camera inside the cold storage and transmit it to the frost detection module for analysis; The monitoring area setting module is used to adjust the size and position of the monitoring area as needed, allowing the user to draw and set the monitoring area of ​​the fan fins; The data processing module processes and analyzes the recognition result of the image acquisition module, generates real-time frost degree data, and transmits it to the defrost control module; The system feedback module is responsible for recording and analyzing data during the system operation process, providing feedback and historical records of defrosting operations, and being able to monitor the system operation status in real time.

3. A method for fan frost identification and intelligent defrosting based on AI vision according to any one of claims 1-2, characterized in that: The following steps are involved: S1. System initialization. After the system is started, the initial configuration is performed first, including the setting of the cold storage monitoring area, the calibration of the camera, and the loading of the AI ​​model; S2, frost detection, after the system enters the running state, the frost situation inside the cold storage is detected in real time through image acquisition, frost recognition and data processing; S3, defrost control: according to the frost detection results, the system automatically decides whether to start the defrost operation based on the comparison between the frost degree detected in the monitoring area and the set threshold; S4, system feedback and optimization: the system continuously optimizes the defrost strategy and improves system performance by recording and analyzing defrost operation data.

4. The method for fan frost identification and intelligent defrosting based on AI vision according to claim 3 is characterized in that: In step S1 , the calibration of the camera includes adjusting the position of the camera and correcting image distortion.

5. The method for fan frost identification and intelligent defrosting based on AI vision according to claim 3 is characterized in that: In step S2, the specific steps of frost detection are as follows: S21, image acquisition, the system obtains real-time image data from the camera inside the cold storage and transmits it to the frost detection module for analysis; S22, frost recognition. The frost detection module uses the AI ​​model to segment and analyze the image and identify areas with different degrees of frost in the monitoring area. The system generates a detailed report on the frost situation based on the recognition results, including the distribution of areas with heavy frost, light frost and no frost. S23, data processing, the system processes and analyzes the recognition results, generates real-time frost degree data, and transmits it to the defrost control module.

6. The method for fan frost identification and intelligent defrosting based on AI vision according to claim 3, characterized in that: In step S3, the specific steps of defrost control are as follows: S31, start defrosting. When it is detected that the degree of frost in the monitoring area exceeds the set threshold, the system automatically starts the defrosting operation. The system starts to perform overall defrosting operation on the cold storage, and heats the fan fins through the heating device to remove the frost; S32, real-time monitoring. During the defrosting process, the system continuously monitors the frost condition in the monitoring area and detects the frost change in real time through the camera to determine whether the defrosting effect meets expectations. S33, ending the defrost ahead of schedule. When the system detects that the defrost effect reaches the set standard, it automatically ends the defrost operation ahead of schedule, analyzes the data of this defrost operation, and optimizes future defrost strategies.

7. The method for fan frost identification and intelligent defrosting based on AI vision according to claim 3, characterized in that: In step S4, the specific steps of system feedback and optimization are as follows: S41, data recording, the system records the detailed data of each defrosting operation, including frosting conditions, defrosting time and energy consumption, and the data is stored in the system database for subsequent analysis and optimization; S42. The system analyzes the recorded data to evaluate the effect and energy consumption of the defrost operation. Through data analysis, the system adjusts the defrost strategy and parameters to improve the accuracy of defrost and the energy-saving effect.

Images