Intelligent frequency modulation method and system for dust removal fan of raw material factory

Abstract

The invention belongs to the technical field of intelligent frequency modulation of fans, and provides an intelligent frequency modulation method and system for a dust removal fan of a raw material plant. Raw material image data and humidity data of a raw material stacking and transmission area in the raw material plant are obtained; preprocessing the raw material image data, extracting textural features, color features and shape features of the preprocessed image, and judging the type and the stacking density of the raw material and whether agglomeration exists or not by using a pre-trained image recognition model; analyzing the humidity data, calculating the average humidity and the humidity change rate, and judging the humidity change condition of the raw materials according to the change trend of the humidity data; according to the recognition result of the raw materials and the change condition of the humidity degree, querying a pre-constructed mapping relation among the image features of the raw materials, the humidity data and the optimal frequency of the dust removal fan, and determining the optimal frequency of the dust removal fan under the current condition; and controlling the dust removal fan according to the optimal frequency. The invention can realize energy conservation, consumption reduction and efficient dust removal.

Description

Technical Field

[0001] The present invention belongs to the technical field of intelligent frequency regulation of fans, and in particular relates to an intelligent frequency regulation method and system for a dust removal fan in a raw material plant. Background Art

[0002] The statements in this section merely provide background information related to the present invention and do not necessarily constitute prior art.

[0003] In the ironmaking process of the blast furnace of the ironmaking plant, raw materials such as iron ore and coal are transported remotely from the fixed raw material plant area to the blast furnace by conveyor belts. The conveyor belts will be changed during the transportation process of the raw materials, and the raw materials will fall from the previous conveyor belt to the next conveyor belt when the conveyor belt changes. A large amount of dust will be generated during the falling process of the raw materials, which not only affects the environmental quality, but also poses a threat to the health of the operators.

[0004] Dust removal fans are important equipment for removing dust from these locations, but traditional fan control systems often have the following problems: First, the fan frequency cannot be dynamically adjusted according to factors such as the type of raw material and humidity; second, the fan's operating status is based on fixed settings, lacks real-time adaptability, and easily causes energy waste. At present, although there are dust concentration sensors to monitor smoke, these sensors have a slow response speed and cannot accurately reflect the impact of humidity and type of different raw materials on smoke emissions.

[0005] Therefore, how to accurately distinguish the type and humidity of raw materials based on the image information of the raw materials, and predict the smoke concentration in real time, so as to intelligently adjust the frequency of the dust removal fan, is a key issue in improving fan efficiency, reducing energy consumption and improving the level of environmental protection. Summary of the invention

[0006] In order to solve the above problems, the present invention proposes an intelligent frequency modulation method and system for dust removal fans in raw material plants. The present invention can accurately judge the characteristics of raw materials according to the type and humidity of raw materials through image recognition technology, predict the smoke emission concentration, and intelligently adjust the frequency of the dust removal fan according to the prediction results, thereby achieving energy saving and consumption reduction and efficient dust removal.

[0007] According to some embodiments, the present invention adopts the following technical solutions:

[0008] An intelligent frequency modulation method for a dust removal fan in a raw material plant comprises the following steps:

[0009] Acquire raw material image data and humidity data of raw material stacking and transportation areas in the raw material factory;

[0010] Preprocessing the raw material image data, extracting texture features, color features and shape features of the preprocessed image, and judging the type, stacking density and whether there is agglomeration of the raw materials based on the features using a pretrained image recognition model;

[0011] Analyze the humidity data, calculate the average humidity and humidity change rate, and determine the change in the dryness and wetness of the raw materials according to the change trend of the humidity data;

[0012] According to the identification results of the raw materials and the changes in the dryness and wetness, the mapping relationship between the pre-built raw material image features, humidity data and the optimal frequency of the dust removal fan is queried to determine the optimal frequency of the dust removal fan under the current situation;

[0013] According to the optimal frequency, the dust removal fan is controlled to adjust the fan frequency.

[0014] As an optional implementation method, the collection points of the raw material image data can fully cover the relevant areas of raw material stacking and transmission, and the light conditions and viewing angles are considered in the selection of the collection points to avoid blind spots;

[0015] The moisture data includes moisture information obtained at different depths and locations in the raw material pile and at multiple locations around the conveyor belt.

[0016] As an optional implementation, the preprocessing process includes performing denoising, converting the color image into a grayscale image, and performing image enhancement operations.

[0017] As an optional implementation, the process of extracting texture features, color features and shape features of the preprocessed image includes: extracting texture features of the raw material image using a texture feature extraction method to determine the particle size and shape distribution of the raw material;

[0018] Extract color features. For raw materials with color differences, obtain color feature information through color histogram to assist in determining the type and state of the raw materials.

[0019] Extract shape features and identify the contour shape of raw materials to distinguish raw materials of different forms and analyze the stacking of raw materials.

[0020] As an optional implementation method, through experiments and statistical analysis of historical data, a mapping relationship between raw material image features, humidity data and the optimal frequency of the dust removal fan is established, and the mapping relationship is stored in the form of a database or a rule table.

[0021] As an optional implementation, when determining the optimal frequency of the dust removal fan under the current situation, a certain frequency adjustment step and buffer zone are set in consideration of the operation stability and the continuity of the dust removal effect.

[0022] As an optional implementation method, after the fan frequency is adjusted, the dust removal effect is continuously monitored, and the mapping relationship between image recognition, humidity monitoring and fan frequency is optimized and adjusted based on the feedback of the dust removal effect.

[0023] An intelligent frequency modulation system for dust removal fans in a raw material plant, comprising:

[0024] A data acquisition unit, used to obtain raw material image data and humidity data of raw material stacking and transmission areas in the raw material factory;

[0025] An image recognition module is used to preprocess the raw material image data, extract texture features, color features and shape features of the preprocessed image, and determine the type, stacking density and whether there is agglomeration of the raw materials based on the features using a pre-trained image recognition model;

[0026] A humidity analysis module is used to analyze the humidity data, calculate the average humidity and humidity change rate, and determine the change in the dryness and wetness of the raw materials according to the change trend of the humidity data;

[0027] The intelligent frequency modulation decision module is used to query the mapping relationship between the pre-built raw material image features, humidity data and the optimal frequency of the dust removal fan according to the identification results of the raw materials and the changes in the dryness and wetness, and determine the optimal frequency of the dust removal fan under the current situation;

[0028] The fan frequency control module is used to control the dust removal fan according to the optimal frequency to adjust the fan frequency.

[0029] As an optional implementation, the data acquisition unit includes an image acquisition module and a humidity monitoring module, wherein the image acquisition module is used to acquire images of the raw materials in the raw material factory;

[0030] The humidity monitoring module is used to monitor the humidity data of the raw materials in real time.

[0031] As an optional implementation, the intelligent frequency modulation decision module includes a data analysis processor and a database. The data analysis processor is used to process image features and humidity data, and call a pre-established image recognition model and mapping relationship to determine the optimal operating frequency of the dust removal fan. The database is used to store historical data and decision results.

[0032] As an optional implementation, the fan frequency control module includes a variable frequency driver, a controller and a communication interface, the variable frequency driver is used to adjust the speed of the fan, the controller is used to receive instructions and control the operation of the variable frequency driver, and the communication interface is used to realize information interaction with other modules.

[0033] As an optional implementation, it also includes a human-computer interaction module, which includes a display screen, an operation panel and an alarm device. The display screen is used to display the raw material image, humidity data and fan operating frequency information in real time. The operation panel is used to receive user setting instructions, perform parameter setting and manual control switching, and the alarm device is used to issue a warning message when an abnormal situation occurs in the system.

[0034] Compared with the prior art, the present invention has the following beneficial effects:

[0035] The present invention combines image recognition technology with humidity sensors to intelligently distinguish between different types of raw materials and humidity, thereby predicting smoke emission concentrations and adjusting fan frequency in real time, thus avoiding the fixed setting of traditional fan control methods;

[0036] The present invention can automatically adjust the air volume of the fan according to the actual smoke and dust emission of the raw material, thereby reducing unnecessary energy consumption and achieving energy-saving effect;

[0037] The present invention ensures that the smoke emission concentration is within a reasonable range and improves the dust removal effect by controlling the operating state of the fan in real time; combines image recognition and humidity data with the smoke emission model for prediction, making the smoke concentration prediction more accurate, thereby achieving more efficient fan adjustment;

[0038] The invention can dynamically adapt to changes in raw material types and humidity, and is widely applicable to production environments of various raw material types.

[0039] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, preferred embodiments are given below and described in detail with reference to the accompanying drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] The accompanying drawings in the specification, which constitute a part of the present invention, are used to provide a further understanding of the present invention. The exemplary embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute improper limitations on the present invention.

[0041] Figure 1 A schematic diagram of the structure of an intelligent frequency modulation system for dust removal fans in a raw material plant based on raw material image recognition and humidity monitoring according to an embodiment;

[0042] Figure 2 A flow chart of an intelligent frequency modulation method for dust removal fans in a raw material plant based on raw material image recognition and humidity monitoring according to an embodiment;

[0043] Among them, 1. Raw materials, 2. High-definition industrial camera, 3. Humidity online monitoring device, 4. Converter, 5. Fiber optic transceiver, 6. Fiber optic transceiver, 7. Converter, 8. Computer host, 9. Display screen, 10. Controller, 11. Frequency converter, 12. Fan. DETAILED DESCRIPTION

[0044] The present invention will be further described below in conjunction with the accompanying drawings and embodiments.

[0045] It should be noted that the following detailed descriptions are all illustrative and intended to provide further explanation of the present invention. Unless otherwise specified, all technical and scientific terms used herein have the same meanings as those commonly understood by those skilled in the art to which the present invention belongs.

[0046] It should be noted that the terms used herein are only for describing specific embodiments and are not intended to limit exemplary embodiments according to the present invention. As used herein, unless the context clearly indicates otherwise, the singular form is also intended to include the plural form. In addition, it should be understood that when the terms "comprising" and / or "including" are used in this specification, it indicates the presence of features, steps, operations, devices, components and / or combinations thereof.

[0047] In the absence of conflict, the embodiments in this application and the features in the embodiments may be combined with each other.

[0048] Embodiment 1

[0049] An intelligent frequency modulation system for dust removal fans in a raw material plant based on raw material image recognition and humidity monitoring comprises an image acquisition module, a humidity monitoring module, an image recognition module, an intelligent frequency modulation decision module, a fan frequency modulation control module, a communication module and a human-computer interaction module.

[0050] like Figure 1 As shown, in this embodiment, the image acquisition module mainly includes a high-definition industrial camera 2. The raw material 1 can be on the ground or on a conveyor belt. The high-definition industrial camera 2 has a high resolution (e.g., 1080p and above) and a high frame rate (more than 30 frames per second), and can clearly capture the details of the raw materials under different lighting conditions. The lens of the high-definition industrial camera 2 needs to have a suitable focal length and viewing angle to cover the key areas of raw material stacking and transmission.

[0051] In some embodiments, a fixed-focus lens may be used, and a suitable focal length (eg, 5-50 mm) may be selected according to the monitoring distance and range.

[0052] High-definition industrial cameras 2 need to be equipped with protective housings and mounting brackets. The protective housing can protect the camera from harsh environmental factors such as dust and moisture. The mounting bracket should be able to flexibly adjust the angle and height of the camera to ensure the best shooting angle. For example, a rotatable and retractable metal bracket can be used. The material can be stainless steel or aluminum alloy. In addition, auxiliary lighting equipment such as LED light bars or infrared lights can be equipped in dark areas. LED light bars can provide stable visible light illumination, and infrared lights can be used to achieve night vision functions in low light or no light conditions to ensure the continuity of image acquisition.

[0053] Of course, in other embodiments, the image acquisition module may use existing equipment, or adopt other structures and layout methods, as long as a clear image of the raw material can be acquired.

[0054] In this embodiment, the humidity monitoring module mainly includes a humidity online monitoring device 3 and a data acquisition device. The humidity online monitoring device 3 uses a high-precision, high-reliability humidity sensor, such as a capacitive humidity sensor or a resistive humidity sensor. The measurement range of the sensor should meet the requirements of humidity changes in the raw material factory (usually 0-100% RH), and the accuracy can reach ±2% RH or higher. Multiple sensors are evenly distributed and installed at different depths of the raw material pile (such as every 0.5-1 meter), different positions (edges, centers, etc.) and on the conveyor belt to fully reflect the changes in raw material humidity. The data acquisition device should have multiple channels (determined according to the number of sensors) and can simultaneously receive and process data from multiple sensors. It can convert analog signals into digital signals, and perform preliminary filtering and calibration processing on the data to improve the accuracy of the data.

[0055] Similarly, in other embodiments, the humidity monitoring module may use existing equipment or sensors, or adopt other structures and layout methods, as long as humidity monitoring data that meets the accuracy requirements can be collected.

[0056] In this embodiment, the image recognition module mainly includes a computer host 8. The computer host 8 is equipped with a dedicated graphics processing unit (GPU) or a high-performance central processing unit (CPU). The GPU has a strong parallel computing capability, is suitable for processing large-scale image data, and can accelerate the operation of the image recognition algorithm. For example, some GPU models of NVIDIA have a large number of CUDA cores, which can process images efficiently.

[0057] The computer host 8 also needs to install image recognition algorithm software, which is developed based on machine learning or deep learning algorithms.

[0058] For machine learning algorithms, support vector machines, random forests, etc. may be included; for deep learning algorithms, convolutional neural networks (CNNs) are commonly used, such as AlexNet, VGG, ResNet and other models. These algorithm software needs to be run on a server or a high-performance computer, and the software needs to have model training, parameter adjustment and real-time image recognition functions. The storage device in the computer host 8 is a large-capacity hard disk (such as a mechanical hard disk or a solid-state hard disk) for storing training image data, trained model parameters and intermediate calculation results. At the same time, memory (RAM) may be required to temporarily store the image data being processed, and the memory capacity should be large enough (16GB and above) to ensure the efficient operation of the image recognition algorithm.

[0059] Of course, in other embodiments, the image recognition module may also be implemented using other processors or modules, which are not exhaustively listed here.

[0060] like Figure 1 As shown, the intelligent frequency modulation decision module is also mainly executed by the computer host 8. The computer host 8 has a data analysis processor with high-performance computing capabilities for processing a large amount of image feature data and humidity data. The processor should have a multi-core architecture (quad-core, octa-core, etc.) and can quickly perform data analysis and algorithm operations.

[0061] The computer host 8 is installed with intelligent algorithm software, which can establish intelligent algorithms based on historical data and experimental results, such as fuzzy logic algorithms, neural network algorithms, etc. These algorithm software take the raw material state and humidity data as input and output the optimal frequency value of the dust removal fan. The software should have self-learning and self-adaptive capabilities, and be able to continuously optimize the algorithm parameters according to the system operation conditions. The computer host 8 is installed with database software for storing historical raw material image data, humidity data, fan operation frequency data and decision results. The database can be a relational database (such as MySQL, Oracle, etc.) or a non-relational database (such as MongoDB, etc.), with large-capacity storage and fast query functions to support long-term storage and real-time retrieval of data.

[0062] Of course, in other embodiments, the intelligent frequency modulation decision module may also be implemented using other processors or modules, which are not exhaustively listed here.

[0063] like Figure 1 As shown, the fan frequency control module includes a controller 10, a frequency converter 11 and a fan 12. The frequency converter 11 receives the frequency control signal from the intelligent frequency control decision module and converts it into the voltage and frequency for driving the motor of the fan 12, thereby changing the speed of the fan 12. The frequency converter 11 should have a wide frequency speed regulation range (such as 0-50Hz or wider), high-precision frequency control (frequency resolution can reach 0.01Hz or higher) and good overload protection capability.

[0064] The controller 10 is used to coordinate the communication and control logic between the inverter 11 and other modules (such as the computer host 8). It can be a programmable logic controller (PLC) or a dedicated microcontroller. The controller 10 receives the instructions of the intelligent frequency modulation decision module and transmits them to the inverter 11. At the same time, it monitors the operating status of the inverter 11 and the fan 12, such as current, voltage, speed, etc., to ensure the safe and stable operation of the system.

[0065] Similarly, in other embodiments, the fan frequency modulation control module may also adopt other structural forms or be implemented by other devices.

[0066] like Figure 1As shown, the human-computer interaction module includes a display screen 9, an operation panel, a keyboard and a mouse. The display screen 9 can be an industrial-grade liquid crystal display (LCD) or a light-emitting diode display (LED), and the size is determined according to the operating distance and visibility requirements. The display screen should have high definition, high contrast and a good viewing angle to clearly display various parameters, images and alarm information of the system. The operation panel includes input devices such as buttons and touch screens, allowing users to perform operations such as parameter setting and manual control. Buttons can be used for emergency operations, parameter adjustment and other functions, and the touch screen provides a more intuitive and convenient operation method. Users can start, stop, set parameters and other functions of the system through touch operations.

[0067] Some embodiments also include an alarm device. When an abnormal situation occurs in the system, such as excessive humidity, image recognition failure, fan failure, etc., the alarm device sends out an audible and visual alarm to remind the user to handle it in time.

[0068] like Figure 1 As shown, the communication module mainly includes a converter 4, a converter 7, a fiber optic transceiver 5, a fiber optic transceiver 6, an optical fiber and a network cable. First, the number of devices that need to be connected in the system must be determined. Including high-definition industrial cameras, humidity sensors (connected through a data collector), image processors, servers in the intelligent frequency modulation decision module, devices in the fan frequency modulation control module (such as variable frequency drives, controllers), and devices in the human-computer interaction module (such as display screens, host computers, etc.). Generally speaking, a certain amount of port redundancy needs to be reserved to meet possible future equipment expansion needs. For example, if there are currently 20 devices that need to be connected, it is recommended to select a switch with at least 24 ports. For devices such as cameras, servers, and controllers, the RJ45 port can provide a stable wired network connection.

[0069] Fiber optic ports are very useful if some devices in the system are far away (for example, more than 100 meters) or if data transmission requires high anti-interference performance. These devices can be connected to the fiber optic ports of the switch through fiber optic patch cords to achieve high-speed, stable and long-distance data transmission.

[0070] In the complex electromagnetic environment of the raw material plant, fiber optic connection can effectively reduce the impact of electromagnetic interference on data transmission.

[0071] For high-definition cameras in the image acquisition module, if high resolution and high frame rate are used for image acquisition, in order to ensure that image data can be transmitted in a timely and smooth manner, it is recommended to choose a Gigabit switch, because a 100M switch may have a data transmission bottleneck, resulting in image delay or frame loss.

[0072] For other devices, such as humidity sensor data collectors, if the data volume is small, a 100M port may be sufficient to meet the needs. If there is a large amount of image data in the system that needs to be transmitted simultaneously, or it may be upgraded to an application with higher bandwidth requirements in the future (such as adding more high-definition cameras or performing real-time video analysis), a 10G switch can be considered. A 10G switch can provide a higher transmission rate and meet the high-speed transmission requirements of large data volumes, but the cost is relatively high.

[0073] Embodiment 2

[0074] A working method of an intelligent frequency modulation system for dust removal fans in a raw material plant based on raw material image recognition and humidity monitoring is provided in an embodiment 1, such as Figure 2 As shown, the following steps are included:

[0075] Step 1: Data collection and sensor installation;

[0076] Step 2: Establishment of raw material image recognition model;

[0077] Step 3: Real-time monitoring and acquisition of raw material humidity;

[0078] Step 4: Establishment of intelligent frequency modulation decision-making mechanism;

[0079] Step 5: Realize intelligent frequency control of the fan.

[0080] Specifically, step 1, data collection and sensor installation, includes the following steps:

[0081] Step 101: Install high-definition cameras at key locations in the raw material factory to ensure that they can fully cover relevant areas such as raw material stacking and transmission to obtain image data of the raw materials. The installation location of the camera should take into account light conditions and viewing angles to avoid blind spots.

[0082] Step 102: Install humidity sensors at different depths and locations of the raw material pile, near the conveyor belt, etc. The sensors need to be able to accurately measure humidity changes in the raw materials and the surrounding environment, and transmit the data in real time.

[0083] Step 2: Establishing a raw material image recognition model, including the following steps:

[0084] Step 201: De-noise the collected raw material image, using algorithms such as median filtering and Gaussian filtering to remove interference factors such as salt and pepper noise and Gaussian noise in the image. Grayscale the image to convert the color image into a grayscale image to reduce the amount of data while retaining key information such as image texture. Perform image enhancement operations such as histogram equalization to enhance the contrast of the image and make the characteristics of the raw material more obvious.

[0085] Step 202: Use texture feature extraction methods, such as grayscale co-occurrence matrix, to extract texture features of the raw material image to determine the particle size, shape distribution, etc. of the raw material, because the texture of different raw material states is different. Extract color features. For raw materials with color differences, color feature information can be obtained through methods such as color histograms to assist in determining the type and state of the raw material. Perform shape feature extraction to identify the contour shape of the raw material. For raw materials in different forms such as block and powder, their contour features help analyze the accumulation of the raw materials.

[0086] Step 203, use machine learning or deep learning algorithms to establish an image recognition model. For machine learning, algorithms such as support vector machines and random forests can be used to train the model using the extracted features as input. If deep learning is used, a convolutional neural network (CNN) such as ResNet, VGG and other network structures can be selected, and a large amount of labeled raw material image data can be used for training. The collected real-time images are classified and identified to determine the type of raw materials, bulk density, whether there is agglomeration, etc.

[0087] Step 3: Real-time monitoring and acquisition of raw material humidity, including the following steps:

[0088] Step 301: The humidity sensor continuously collects raw material humidity data and transmits the data to the control system. The humidity sensor is calibrated regularly to ensure the accuracy of the measured data. The sensor can be calibrated using a standard humidity source to adjust the sensor parameters.

[0089] Step 302: Analyze the collected humidity data in real time, including calculating parameters such as average humidity and humidity change rate. According to the change trend of humidity data, determine the change of dryness and wetness of the raw material. For example, if the humidity change rate suddenly increases, it may indicate that the raw material has a tendency to be damp.

[0090] Step 4: Establishment of intelligent frequency modulation decision-making mechanism, including the following steps:

[0091] Step 401: Establish a mapping relationship between raw material image features, humidity data and the optimal frequency of the dust removal fan through experiments and historical data statistical analysis. For example, when the raw material image shows that the particles are small and densely packed, and the humidity is high, a higher fan frequency is required to effectively remove dust. This mapping relationship is stored in the control system in the form of a database or rule table.

[0092] Step 402: Integrate the real-time data obtained by the image recognition module and the humidity monitoring module, query the mapping table, and determine the optimal frequency of the dust removal fan under the current situation. Considering the stability of the system and the continuity of the dust removal effect, a certain frequency adjustment step and buffer zone can be set to avoid frequent and large fluctuations in the fan frequency.

[0093] Step 5: Implement intelligent frequency control of the fan, including the following steps:

[0094] Step 501: According to the frequency value determined by the intelligent frequency regulation decision module, the control system outputs a corresponding control signal to the variable frequency driver of the dust removal fan to adjust the fan frequency.

[0095] Step 502: After the fan frequency is adjusted, the dust removal effect is continuously monitored, and real-time data can be obtained by installing a dust concentration sensor at the discharge port.

[0096] Step 503: According to the feedback of dust removal effect, the mapping relationship between image recognition, humidity monitoring and fan frequency is optimized and adjusted to improve the accuracy and efficiency of the whole system.

[0097] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. For those skilled in the art, the present invention may have various modifications and variations. Any modification, equivalent replacement, improvement, etc. made by those skilled in the art within the spirit and principle of the present invention without creative labor shall be included in the protection scope of the present invention.

Claims

1. An intelligent frequency modulation method for dust removal fans in raw material plants, characterized in that: The following steps are involved: Acquire raw material image data and humidity data of raw material stacking and transportation areas in the raw material factory; Preprocessing the raw material image data, extracting texture features, color features and shape features of the preprocessed image, and judging the type, stacking density and whether there is agglomeration of the raw materials based on the features using a pretrained image recognition model; Analyze the humidity data, calculate the average humidity and humidity change rate, and determine the change in the dryness and wetness of the raw materials according to the change trend of the humidity data; According to the identification results of the raw materials and the changes in the dryness and wetness, the mapping relationship between the pre-built raw material image features, humidity data and the optimal frequency of the dust removal fan is queried to determine the optimal frequency of the dust removal fan under the current situation; According to the optimal frequency, the dust removal fan is controlled to adjust the fan frequency.

2. The intelligent frequency modulation method for dust removal fans in raw material plants as claimed in claim 1 is characterized in that: The collection points of raw material image data can fully cover the relevant areas of raw material stacking and transmission, and the light conditions and viewing angles are considered in the selection of collection points to avoid blind spots; The moisture data includes moisture information obtained at different depths and locations in the raw material pile and at multiple locations around the conveyor belt.

3. The intelligent frequency modulation method for dust removal fans in a raw material plant as claimed in claim 1 is characterized in that: The preprocessing process includes performing denoising, converting the color image into a grayscale image, and performing image enhancement operations.

4. The intelligent frequency modulation method for dust removal fans in a raw material plant as claimed in claim 1 is characterized in that: The process of extracting texture features, color features and shape features of the preprocessed image includes: extracting texture features of the raw material image using a texture feature extraction method to determine the particle size and shape distribution of the raw material; Extract color features. For raw materials with color differences, obtain color feature information through color histogram to assist in determining the type and state of the raw materials. Extract shape features and identify the contour shape of raw materials to distinguish raw materials of different forms and analyze the stacking of raw materials.

5. The intelligent frequency modulation method for dust removal fans in raw material plants as claimed in claim 1 is characterized in that: Through experiments and statistical analysis of historical data, a mapping relationship between raw material image features, humidity data and the optimal frequency of the dust removal fan is established, and the mapping relationship is stored in the form of a database or rule table.

6. The intelligent frequency modulation method for dust removal fans in a raw material plant as claimed in claim 1 is characterized in that: When determining the optimal frequency of the dust removal fan under the current circumstances, consider the operating stability and continuity of the dust removal effect, and set a certain frequency adjustment step and buffer zone.

7. The intelligent frequency modulation method for dust removal fans in a raw material plant as claimed in claim 1 is characterized in that: After the fan frequency is adjusted, the dust removal effect is continuously monitored, and the mapping relationship between image recognition, humidity monitoring and fan frequency is optimized and adjusted based on the feedback of the dust removal effect.

8. An intelligent frequency modulation system for dust removal fans in a raw material plant, characterized in that: include: A data acquisition unit, used to obtain raw material image data and humidity data of raw material stacking and transmission areas in the raw material factory; An image recognition module is used to preprocess the raw material image data, extract texture features, color features and shape features of the preprocessed image, and determine the type, stacking density and whether there is agglomeration of the raw materials based on the features using a pre-trained image recognition model; A humidity analysis module is used to analyze the humidity data, calculate the average humidity and humidity change rate, and determine the change in the dryness and wetness of the raw materials according to the change trend of the humidity data; The intelligent frequency modulation decision module is used to query the mapping relationship between the pre-built raw material image features, humidity data and the optimal frequency of the dust removal fan according to the identification results of the raw materials and the changes in the dryness and wetness, and determine the optimal frequency of the dust removal fan under the current situation; The fan frequency control module is used to control the dust removal fan according to the optimal frequency to adjust the fan frequency.

9. The intelligent frequency modulation system for dust removal fans in a raw material plant as claimed in claim 8, characterized in that: The data acquisition unit includes an image acquisition module and a humidity monitoring module, wherein the image acquisition module is used to acquire images of the raw materials in the raw material factory; The humidity monitoring module is used to monitor the humidity data of the raw materials in real time.

10. The intelligent frequency modulation system for dust removal fans in a raw material plant as claimed in claim 8, characterized in that: The intelligent frequency modulation decision module includes a data analysis processor and a database. The data analysis processor is used to process image features and humidity data, and call a pre-established image recognition model and mapping relationship to determine the optimal operating frequency of the dust removal fan. The database is used to store historical data and decision results.

11. The intelligent frequency modulation system for dust removal fans in a raw material plant as claimed in claim 8, characterized in that: The fan frequency modulation control module includes a variable frequency driver, a controller and a communication interface. The variable frequency driver is used to adjust the speed of the fan, the controller is used to receive instructions and control the operation of the variable frequency driver, and the communication interface is used to realize information exchange with other modules.

12. The intelligent frequency modulation system for dust removal fans in a raw material plant as claimed in claim 8, characterized in that: It also includes a human-computer interaction module, which includes a display screen, an operation panel and an alarm device. The display screen is used to display raw material images, humidity data and fan operating frequency information in real time. The operation panel is used to receive user setting instructions, perform parameter settings and manual control switching, and the alarm device is used to issue early warning information when an abnormal situation occurs in the system.

Images