Intelligent control system of material leveling device

Abstract

The invention relates to the technical field of leveling device control systems, and discloses an intelligent control system of a leveling device, which comprises a material identification module, a parameter processing module, a grade division module and the like. The material identification module divides a flattening area and a to-be-processed area, and the parameter processing module controls starting of material flattening operation. The state monitoring module collects material basic form data and a real-time movement track of the device, and the storage unit stores historical feature data, distribution difference features and fluctuation data. The grade division module judges the material feature grade and the flatness grade, and the strategy generation module generates a dynamic adjustment strategy according to the material feature grade and the flatness grade and the operation starting point distance. The difference judgment module identifies the adjustment area, and the distribution detection module judges whether the track deviates or not and triggers correction. The system also realizes surface data analysis and visualization through a surface analysis module and an operation terminal, improves the intelligence, accuracy and efficiency of material leveling operation, and is suitable for industrial material leveling scenes.

Description

Technical Field

[0001] The present invention relates to the technical field of a leveling device control system, and particularly to an intelligent control system for a leveling device. Background Art

[0002] In the field of industrial production, material stacking is a common production process, such as the stacking of goods in warehousing logistics, the accumulation of raw materials in chemical production, etc. The flatness of material stacking directly affects the efficiency and quality of subsequent production processes. For example, during the material transportation process, uneven stacking may cause the materials to scatter, increasing the transportation risks and costs; in the processing link, insufficient flatness of the materials will affect the positioning accuracy of the processing equipment, resulting in increased processing errors and even equipment damage. Traditional leveling operations mainly rely on manual operation or simple mechanical control, which have many defects.

[0003] The manual leveling method has a high labor intensity, low efficiency, and is significantly affected by the experience and skill level of the operators, making it difficult to ensure the consistency of leveling quality. For example, in a large-scale warehousing scenario, manually leveling a large amount of materials requires a lot of time and manpower, and it is easy to have omissions or uneven leveling. And simple mechanical control leveling devices usually adopt fixed operation strategies and cannot be dynamically adjusted according to the actual situation of the material stack. The physical properties (such as humidity, thickness, particle size, etc.) of different materials vary greatly, and the stacking states of the same material in different areas may also be different. Traditional mechanical control cannot perceive these changes in real time, resulting in poor leveling effects.

[0004] With the development of industrial automation and intelligence, higher requirements are put forward for the intelligent control of leveling devices. In the prior art, although some leveling devices introduce sensors for status monitoring, most of them can only collect single parameters, such as only monitoring the thickness or humidity of the materials, and cannot comprehensively evaluate the material stack based on multi-dimensional data. At the same time, the existing control systems lack the effective utilization of historical data and cannot predict the possible problems in the current leveling operation based on the historical characteristics of the materials, resulting in the lack of scientificity and pertinence in the formulation of leveling strategies.

[0005] In addition, during the operation of the leveling device, how to adjust the movement trajectory in real time to adapt to the differences in material distribution is also a difficult point in the prior art. Traditional systems are difficult to accurately judge whether the leveling device enters the area that needs to be adjusted, nor can they dynamically correct the movement trajectory according to the characteristics of the material distribution differences, easily causing the leveling device to deviate from the optimal path during the operation, affecting the leveling efficiency and quality. For example, when there are areas with higher local humidity or larger thickness in the material stack, traditional devices may not be able to adjust the operation intensity and trajectory in time, resulting in incomplete leveling of this area. Summary of the Invention

[0006] The purpose of the present invention is to provide an intelligent control system for a material leveling device to solve the problems raised in the above-mentioned background technology.

[0007] To achieve the above object, the present invention provides the following technical solution: An intelligent control system for a material leveling device, the system includes: A material identification module, a parameter processing module, a grade classification module, a storage unit, a strategy generation module, a state monitoring module, a distribution detection module, and a difference determination module; The material identification module is used to divide the flat area and the area to be processed of the material stack; the parameter processing module is used to control the material leveling device to start the leveling operation after receiving an instruction; when performing the initial leveling operation, the state monitoring module is used to collect the basic form data of the material stack and transmit it to the grade classification module; the storage unit is used to store the historical feature data of the material stack and transmit it to the grade classification module; the grade classification module is used to determine the feature grade and flatness grade of the material stack, generate the historical feature grade and flatness grade and transmit them to the strategy generation module; the strategy generation module is used to generate a dynamic adjustment strategy during the operation of the material leveling device according to the flatness grade and historical feature grade, and transmit the dynamic adjustment strategy to the parameter processing module; When the material leveling device performs the leveling operation according to the dynamic adjustment strategy, the state monitoring module is also used to collect the real-time movement trajectory of the material leveling device and transmit it to the difference determination module; the storage unit is also used to store the distribution difference features in the material stack and the fluctuation data corresponding to the distribution difference features and transmit them to the difference determination module; the difference determination module is used to determine whether the material leveling device enters the adjustment area, and if a trajectory correction signal is generated, transmit the real-time movement trajectory of the material leveling device in the adjustment area to the distribution detection module; the distribution detection module is used to judge whether the material leveling device deviates from the reference path during the operation in the adjustment area according to the real-time movement trajectory, and if it deviates, send a movement trajectory correction instruction to the parameter processing module.

[0008] Preferably, the system further includes a surface analysis module and an operation terminal. The state monitoring module is also used to collect the real-time surface data of the material stack and transmit it to the surface analysis module. The surface analysis module is used to analyze the real-time surface data of the material stack in real time. If a defect signal or a deformation signal is generated by the analysis, the coordinate mark and the corresponding real-time surface data are transmitted to the operation terminal, and the operation terminal is used for visual processing of the real-time surface data of the material stack.

[0009] Preferably, the basic form data includes the thickness deviation value and humidity fluctuation value of the material stack; The historical feature data is the historical thickness fluctuation times of the material stack and the adjustment duration corresponding to each fluctuation.

[0010] Preferably, the determination process of the grade classification module is specifically as follows: Obtain the historical thickness fluctuation times of each material stack, and then obtain the adjustment duration corresponding to each fluctuation of each material stack. Accumulate the adjustment durations corresponding to each fluctuation and calculate the mean value to obtain the average adjustment duration of each material stack; Calculate the characteristic coefficient of each material stack based on the historical thickness fluctuation times and the average adjustment duration; Compare the characteristic coefficient with the preset characteristic threshold to determine the historical characteristic level of the material stack as the first-level characteristic, the second-level characteristic or the third-level characteristic; Synchronously obtain the thickness deviation value and the humidity fluctuation value of each material stack, and calculate the flatness coefficient of each material stack; Compare the flatness coefficient with the preset flatness threshold to determine the flatness level of the material stack as the first-level flatness, the second-level flatness or the third-level flatness.

[0011] Preferably, the complexity corresponding to the third-level characteristic is higher than that of the second-level characteristic, and the complexity corresponding to the second-level characteristic is higher than that of the first-level characteristic; The finishing requirement corresponding to the third-level flatness is higher than that of the second-level flatness, and the finishing requirement corresponding to the second-level flatness is higher than that of the first-level flatness.

[0012] Preferably, the working process of the strategy generation module is specifically as follows: The leveling device takes the flatness level of the material stack as the first-priority factor, the historical characteristic level of the material stack as the second-priority factor, and the distance between the current position of the leveling device and the starting point of the operation of the material stack as the third-priority factor; The leveling device starts operating from the current starting point of the operation. If there are multiple material stacks with the same flatness level near the starting point, the material stack with a higher historical characteristic level is preferably selected for processing; If the flatness level and the historical characteristic level of the material stack are the same, the material stack closest to the starting point of the operation is selected for processing; When the leveling device reaches the demarcation point: If the demarcation point is only connected to a single material stack, the leveling device directly processes the material stack; If the demarcation point is connected to two or more material stacks, the material stack with a higher flatness level is preferably processed; when the flatness levels of the multiple material stacks connected by the demarcation point are the same, the material stack with a higher historical characteristic level is preferably processed; when the flatness levels and the historical characteristic levels of the material stacks connected by the demarcation point are the same, a material stack is randomly selected for processing; If the demarcation point is not connected to other material stacks and there is an unprocessed area, the unprocessed area closest to it is preferably processed; when the leveling device reaches any demarcation point, if all other areas have been processed, it returns to the initial position and stops. Generate a dynamic adjustment strategy for the leveling device during operation according to the above rules; When the leveling device operates according to the dynamic adjustment strategy, different operation intensities are adopted for operations in areas corresponding to different flatness levels or three-level features.

[0013] Preferably, the fluctuation data are the core coordinate points, the reference fluctuation amplitude, and the attenuation change rate of the distribution difference features.

[0014] Preferably, the determination process of the difference determination module is specifically as follows: Calculate the displacement difference between the real-time movement trajectory of the leveling device and the core coordinate points of the distribution difference features; Obtain the fluctuation data corresponding to the distribution difference features in the material stack, extract the core coordinate points, the reference fluctuation amplitude, and the attenuation change rate, and obtain the adjustment radius of the distribution difference features by dividing the reference fluctuation amplitude by the attenuation change rate; Construct an adjustment area of the distribution difference features with the core coordinate points as the center point and the adjustment radius as the coverage range; When the displacement difference is greater than the adjustment radius, the leveling device operates outside the adjustment area and maintains the current operating state; When the displacement difference is less than or equal to the adjustment radius, the leveling device enters the adjustment area to operate and generates a trajectory correction signal.

[0015] Preferably, the judgment process of the distribution detection module is specifically as follows: Set the standard pressure range and the allowable offset threshold for the operation of the leveling device; If the value feedback by the pressure sensor is within the allowable offset threshold and the traveling speed of the leveling device is within the standard pressure range, keep the current operation parameters; If the value feedback by the pressure sensor exceeds the allowable offset threshold or the traveling speed of the leveling device deviates from the standard pressure range, generate a movement trajectory correction instruction.

[0016] Preferably, the analysis process of the surface analysis module is specifically as follows: Capture the three-dimensional point cloud data of the material stack; Convert the point cloud data into a two-dimensional projection image, use median filtering to eliminate noise points, set a filtering threshold according to the height difference between the surface form of the material and the background, and extract the effective surface area of the material stack; apply morphological closing operation to fill small holes and retain the complete surface contour features; detect straight edges through Hough transform, match the geometric structure features of the material stack, and filter non-structural interference areas; Synchronously obtain the infrared thermal imaging data of the material stack, set the temperature gradient range to identify abnormal heating areas, use the pseudo-color enhancement technology to strengthen the display of temperature differences, and combine with the edge detection algorithm to extract the temperature mutation boundary to identify abnormal internal structures of the material; If the three-dimensional point cloud data and the infrared thermal imaging data both show abnormalities, generate a defect signal and mark the corresponding coordinates; If neither of the two data shows abnormalities, maintain the normal operation process; If any one of the data shows abnormalities, generate a deformation signal and mark the corresponding coordinates.

[0017] Compared with the prior art, the beneficial effects of the present invention are: The intelligent control system of the leveling device provided by the present invention significantly improves the intelligent level and operation quality of the leveling operation through the collaborative work of multiple modules. The material identification module can accurately divide the leveling area and the area to be processed, enabling the leveling device to carry out operations targeted, avoiding the low efficiency problem caused by traditional blind operations, and reducing unnecessary energy consumption and time waste. The cooperation between the parameter processing module and the state monitoring module realizes the dynamic monitoring and data collection of the leveling operation. The state monitoring module can not only collect basic morphological data, but also obtain information such as real-time movement trajectories, providing rich data support for the precise control of the system.

[0018] The grading module realizes the scientific determination of the feature level and flatness level of the material stack through the analysis of historical feature data and real-time collected data. This grading mechanism enables the system to formulate differentiated leveling strategies according to the complexity of the material and the finishing requirements. For example, for a material stack with a third-level feature and a third-level flatness, the system can adopt higher-intensity operation parameters to ensure the leveling effect; while for materials with a first-level feature and a first-level flatness, the operation intensity can be appropriately reduced to improve the operation efficiency while ensuring quality.

[0019] The strategy generation module generates a dynamic adjustment strategy with the flatness level, historical feature level, and the distance between the starting points of the operation as the priority factors, enabling the leveling device to reasonably plan the operation sequence and path in a complex material stacking scenario. When there are multiple material stacks, the system preferentially processes areas with a high flatness level, a high historical feature level, or a short distance, ensuring the scientific and orderly nature of the leveling operation. At the same time, the setting of the boundary point processing rules effectively solves the operation decision-making problem of traditional devices when connecting multiple regions, avoiding operation conflicts and omissions, and further improving the leveling efficiency.

[0020] The settings of the difference determination module and the distribution detection module enable the real-time monitoring and correction of the movement trajectory of the leveling device. By calculating the displacement difference and constructing the adjustment area, the system can accurately determine whether the leveling device enters the area that needs to be adjusted, and dynamically correct the movement trajectory according to the distribution difference characteristics. When the feedback value of the pressure sensor or the traveling speed is abnormal, the distribution detection module sends a correction instruction in a timely manner to ensure that the leveling device always operates along the reference path, avoiding the problem of uneven leveling caused by trajectory deviation, and improving the accuracy and stability of the leveling operation.

[0021] The combination of the surface analysis module and the operation terminal provides strong support for the surface state monitoring of the material stack. Through the comprehensive analysis of three-dimensional point cloud data and infrared thermal imaging data, the system can timely detect the defects on the material surface and the internal structure abnormalities, and perform visual processing through the operation terminal, enabling the operator to grasp the material state in real time, facilitating the timely adoption of targeted measures to prevent the occurrence of quality problems. This real-time monitoring and feedback mechanism further improves the reliability and safety of the leveling operation. Description of the Drawings

[0022] Figure 1 It is the working principle diagram of the intelligent control system of the leveling device described in the present invention; Figure 2 It is the design diagram of the expansion of the surface analysis system; Figure 3 It is the process design diagram determined by the grading module; Figure 4 It is the process design diagram of difference determination and distribution detection; Figure 5 It is the process design diagram of difference determination and distribution detection. Detailed Embodiments

[0023] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

[0024] Please refer to Figures 1 - 5 , an intelligent control system of a leveling device involved in the present invention, including a material identification module, a parameter processing module, a grading module, a storage unit, a strategy generation module, a state monitoring module, a distribution detection module, and a difference determination module. Each module works together to achieve the intelligent operation control of the leveling device, and the specific implementation is as follows: During system operation, the material identification module first divides the material stack into regions, accurately identifying the flat region and the region to be processed, providing a basic basis for subsequent material leveling operations. After the parameter processing module controls the leveling device to receive the instruction and start the leveling operation, in the initial leveling operation stage (step 1), the state monitoring module begins to collect the basic form data of the material stack, which includes but is not limited to physical characteristic parameters such as the thickness deviation value and humidity fluctuation value of the material stack, and transmits the collected basic form data to the grade classification module in real time. At the same time, the historical characteristic data of the material stack (such as the historical thickness fluctuation times and the adjustment duration corresponding to each fluctuation) pre-stored in the storage unit is also transmitted to the grade classification module.

[0025] Based on the received basic form data and historical characteristic data, the grade classification module determines the characteristic grade and flatness grade of the material stack (step 2). Specifically, first obtain the historical thickness fluctuation times of each material stack, then obtain the adjustment duration corresponding to each fluctuation, and calculate the average adjustment duration by cumulative calculation of the mean value; then calculate the characteristic coefficient based on the historical thickness fluctuation times and the average adjustment duration, compare the characteristic coefficient with the preset characteristic threshold, and determine that the historical characteristic grade is first, second, or third level, where the complexity corresponding to the third level characteristic is higher than the second level, and the second level is higher than the first level. Synchronously, calculate the flatness coefficient according to the thickness deviation value and humidity fluctuation value, and after comparing it with the preset flatness threshold, determine that the flatness grade is first, second, or third level, where the finishing requirement corresponding to the third level flatness is higher than the second level, and the second level is higher than the first level. The generated historical characteristic grade and flatness grade are transmitted to the strategy generation module.

[0026] The strategy generation module generates a dynamic adjustment strategy for the leveling device during operation based on the received flatness level and historical feature level (step 3). This module takes the flatness level as the first-priority factor, the historical feature level as the second-priority factor, and the distance between the current position of the leveling device and the starting point of the operation as the third-priority factor. The specific rules are as follows: Starting from the current operation starting point, if there are multiple material stacks with the same flatness level near the starting point, the one with a higher historical feature level is preferentially selected for processing; if the flatness level and the historical feature level are the same, the nearest material stack is selected. When the leveling device reaches the demarcation point, if the demarcation point is only connected to a single material stack, the stack is directly processed; if it is connected to two or more, the one with a higher flatness level is preferentially processed, if the levels are the same, the one with a higher historical feature level is preferentially processed, and if they are all the same, it is randomly selected; if the demarcation point is not connected to other stacks and there is an unprocessed area, the nearest unprocessed area is preferentially processed; if all areas have been processed, the device returns to the initial position and stops. In addition, in areas corresponding to different flatness levels or three-level features, the leveling device performs operations with different working intensities. After the dynamic adjustment strategy is generated, it is transmitted to the parameter processing module, which controls the leveling device to perform corresponding operations.

[0027] During the process of the leveling device performing the leveling operation according to the dynamic adjustment strategy (step 4), the state monitoring module is also responsible for collecting the real-time movement trajectory of the leveling device and transmitting the trajectory data to the difference determination module. The distribution difference features of the material stacks pre-stored in the storage unit (such as fluctuation data such as core coordinate points, reference fluctuation amplitude, attenuation change rate, etc.) are also transmitted to the difference determination module. The difference determination module calculates the displacement difference between the real-time movement trajectory and the core coordinate point of the distribution difference feature, and calculates the adjustment radius in combination with the reference fluctuation amplitude and attenuation change rate in the fluctuation data (adjustment radius = reference fluctuation amplitude / attenuation change rate), and constructs an adjustment area with the core coordinate point as the center and the adjustment radius as the coverage range. When the displacement difference is greater than the adjustment radius, the leveling device operates outside the adjustment area and maintains the current operating state; when the displacement difference is less than or equal to the adjustment radius, it is determined that the leveling device enters the adjustment area, generates a trajectory correction signal, and transmits the real-time movement trajectory to the distribution detection module. The distribution detection module sets the standard pressure range and the allowable offset threshold for the operation of the leveling device. If the value feedback by the pressure sensor is within the allowable offset threshold and the traveling speed is within the standard pressure range, the current operating parameters are maintained; if it exceeds the allowable offset threshold or the speed deviates from the standard pressure range, a movement trajectory correction instruction is sent to the parameter processing module, which controls the leveling device to adjust the movement trajectory.

[0028] The present invention will be further described below in conjunction with Embodiments 1 to 5: Embodiment 1: On the basis of the above overall solution, this embodiment further includes a surface analysis module and an operation terminal. While collecting the basic morphological data and motion trajectories, the state monitoring module is also responsible for collecting the real-time surface data of the material stack, which includes three-dimensional point cloud data and infrared thermal imaging data, and transmitting it to the surface analysis module. The surface analysis module analyzes the real-time surface data: First, it captures the three-dimensional point cloud data, converts it into a two-dimensional projection image, uses median filtering to eliminate noise points, sets a filtering threshold according to the height difference between the material surface and the background, and extracts the effective surface area; applies morphological closing operation to fill small holes and retain the complete surface contour features; detects straight edges through the Hough transform, matches geometric structure features, and filters out non-structural interference areas. At the same time, it obtains the infrared thermal imaging data, sets the temperature gradient range to identify abnormal heating areas, uses pseudo-color enhancement technology to enhance the temperature difference display, combines edge detection algorithms to extract the temperature mutation boundary, and identifies abnormal internal structures of the material. If both the three-dimensional point cloud data and the infrared thermal imaging data show abnormalities, a defect signal is generated and the corresponding coordinates are marked; if neither data shows abnormalities, the normal operation process is maintained; if any one of the data has an abnormality, a deformation signal is generated and the corresponding coordinates are marked. The surface analysis module transmits the generated coordinate marks and the corresponding real-time surface data to the operation terminal, and the operation terminal visualizes the real-time surface data of the material stack so that the operator can monitor the material state in real time.

[0029] The surface analysis module receives the real-time surface data of the material stack transmitted by the state monitoring module, and the three-dimensional point cloud data is obtained by a three-dimensional laser scanner installed on the leveling device. The scanner scans the surface of the material stack at a certain frequency to obtain the three-dimensional coordinate information of a large number of discrete points, forming point cloud data. The point cloud data contains geometric information such as the shape and contour of the material surface, but due to factors such as the interference of the scanning environment and the accuracy of the equipment, there will be noise points in the point cloud data. The surface analysis module first converts the three-dimensional point cloud data into a two-dimensional projection image for subsequent processing. During the conversion process, according to the distribution range and density of the point cloud data, a suitable projection plane and resolution are determined, and the points in the three-dimensional space are mapped onto the two-dimensional plane to form a grayscale image.

[0030] To improve the image quality, a median filtering algorithm is used to denoise the two-dimensional projection image. Median filtering is a non-linear filtering method that replaces the gray value of each pixel point in the image with the median value in its neighborhood. During the processing, it is crucial to select a suitable filtering window size. If the window is too small, it cannot effectively remove noise; if the window is too large, the details of the image will be lost. Through multiple experiments and experience summary, a window size that can both effectively remove noise and retain image details is selected. After median filtering, the noise points in the image are significantly suppressed, providing a good basis for subsequent area extraction.

[0031] According to the height difference between the material surface and the background, a reasonable filtering threshold is set to binarize the two-dimensional projection image and extract the effective surface area. In the material stack, there is an obvious height difference between the material surface and the surrounding background (such as the ground, equipment brackets, etc.), which is reflected as different depth values in the point cloud data. By analyzing the depth distribution of the point cloud data, a suitable threshold is determined, and the points with depth values within the threshold range are judged as material surface points, while other points are judged as background points. The material surface points in the two-dimensional projection image are marked as foreground (white), and the background points are marked as background (black) to obtain a binary image. In the binary image, there may be some small holes and discontinuous areas, which will affect subsequent analysis and processing.

[0032] To fill these small holes and retain the complete surface contour features, morphological closing operation is used to process the binary image. Morphological closing operation first performs dilation operation and then erosion operation. The dilation operation can expand the foreground area outward to fill the holes; the erosion operation can contract the expanded area back to its original size and remove some small noise points at the same time. By selecting appropriate structuring elements and the number of operations, small holes can be effectively filled, making the material surface area more continuous and complete. On the surface of the material stack, there may be some regular geometric structure features, such as edges, planes, etc. These features are very important for identifying the shape and structure of the material. To detect these geometric structure features, the Hough transform algorithm is used to perform edge detection on the processed binary image.

[0033] The Hough transform is a method for finding geometric shapes such as straight lines and circles in an image. During the processing, each point in the image is transformed into the parameter space, and the possible straight line parameters are found by voting. By setting a suitable threshold, the straight line edges that meet the conditions are screened out. The detected straight line edges are matched with a predefined geometric structure template to identify the geometric structure features on the material surface. For some non-structural interference areas, such as protrusions and depressions on the material surface, by analyzing their geometric features and context information, they are filtered out, and only the feature areas related to the material structure are retained.

[0034] During the synchronization of obtaining 3D point cloud data, the surface analysis module receives the infrared thermal imaging data transmitted by the status monitoring module. The infrared thermal imaging data is obtained by an infrared thermal imager installed on the leveling device. The thermal imager can detect the infrared radiation energy on the surface of an object and convert it into temperature values to form a thermal imaging image. Abnormalities in the internal structure of the material stack, such as uneven density and the presence of voids, may lead to abnormal surface temperature distribution. By analyzing the infrared thermal imaging data, these abnormal areas can be identified. First, a suitable temperature gradient range is set to identify abnormally heated areas. Under normal circumstances, the temperature distribution on the surface of the material stack should be relatively uniform, with a certain temperature gradient. If the temperature of a certain area is significantly higher or lower than that of the surrounding areas, exceeding the set temperature gradient range, it is considered that there is an abnormality in this area.

[0035] To more clearly display the temperature difference, the pseudo-color enhancement technique is used to process the infrared thermal imaging data. Pseudo-color enhancement maps different gray values in a grayscale image to different colors, making the temperature difference more intuitive. By selecting a suitable color mapping table, the low-temperature area is displayed as blue, the medium-temperature area is displayed as green, the high-temperature area is displayed as red, etc. In this way, operators can quickly identify the temperature abnormal areas by observing the color distribution of the image. Combining with the edge detection algorithm, the temperature mutation boundary is extracted. The temperature mutation boundary usually corresponds to changes in the internal structure of the material, such as density mutation and the presence of interfaces. By detecting the temperature mutation boundary, the internal structure abnormality of the material can be identified more accurately. During the processing, the Canny edge detection algorithm is used, which has good edge detection performance and can detect weak edges in the image.

[0036] By setting a suitable threshold, the edges related to temperature mutation are screened out to form the temperature mutation boundary. According to the extracted temperature mutation boundary and the identified abnormally heated areas, the internal structure abnormality of the material is analyzed. For the detected abnormal areas, further analysis and classification are carried out to determine the type and severity of the abnormality. The analysis results are comprehensively compared with the analysis results of the 3D point cloud data to judge whether there are abnormalities in the surface and internal structure of the material stack. After independently analyzing the 3D point cloud data and the infrared thermal imaging data, the analysis results of the two need to be comprehensively compared to more accurately judge the state of the material stack. If the 3D point cloud data shows obvious unevenness and geometric structure abnormalities on the surface of the material, and at the same time the infrared thermal imaging data shows temperature abnormalities in the corresponding areas, it is considered that there is a defect in this area, a defect signal is generated, and the coordinates of this area are marked.

[0037] The defect signal contains information such as the type, location, and severity of the defect, providing a basis for subsequent processing. If neither the three-dimensional point cloud data nor the infrared thermal imaging data shows any abnormal conditions, it is considered that the surface and internal structure of the material stack are normal, and the normal operation process is maintained. The leveling device continues to operate according to the preset parameters and trajectories, and the status monitoring module continues to collect data in real time. If only one of the three-dimensional point cloud data or the infrared thermal imaging data shows an abnormal condition, it is considered that the material stack has deformation, a deformation signal is generated, and the coordinates of the corresponding area are marked. The deformation signal contains information such as the type, location, and degree of deformation. For the deformation situation, it is necessary to further analyze its impact on the leveling operation to determine whether it is necessary to adjust the operation parameters or take other measures.

[0038] The surface analysis module transmits the generated coordinate marks and the corresponding real-time surface data to the operation terminal. After receiving these data, the operation terminal performs visualization processing on the real-time surface data of the material stack. Using three-dimensional visualization technology, the three-dimensional point cloud data is reconstructed into a three-dimensional model to intuitively display the surface morphology of the material stack. On the three-dimensional model, the detected abnormal areas are marked, and different colors and symbols are used to represent the defect and deformation areas, so that the operator can quickly identify them. At the same time, the infrared thermal imaging data is fused with the three-dimensional model, and the temperature distribution is displayed on the three-dimensional model. By means of color gradient, different temperature regions are intuitively displayed, enabling the operator to understand the surface morphology and temperature distribution of the material stack at the same time.

[0039] The operation terminal provides an interactive interface that allows the operator to operate and analyze the visualization data. The operator can observe the three-dimensional model of the material stack from different angles through operations such as zooming, rotating, and translating; can query the detailed information of a specific area, such as temperature value, geometric size, etc.; can also play back and compare historical data to analyze the change trend of the state of the material stack. By monitoring the material state in real time, the operator can timely discover abnormal situations and take corresponding measures. For minor abnormal situations, the operation parameters of the leveling device can be adjusted for processing; for serious abnormal situations, the operation can be stopped in time for maintenance and processing to avoid the problem from further expanding and improve the efficiency and quality of the leveling operation.

[0040] Embodiment 2: In this embodiment, the basic form data is clearly defined as the thickness deviation value and humidity fluctuation value of the material stack, and the historical feature data is the historical thickness fluctuation times of the material stack and the adjustment duration corresponding to each fluctuation. The system realizes the accurate determination of the characteristic level and flatness level of the material stack through the collection, processing, and analysis of these data. The specific implementation process is as follows: The status monitoring module collects the basic morphological data of the material stack in real time through high-precision sensors installed on the leveling device. The collection of thickness deviation values relies on multiple groups of laser ranging sensors distributed at the front end of the leveling device. Each group of sensors contains multiple transmitting and receiving units, arranged in an array form, and can measure the heights of multiple points on the surface of the material stack simultaneously. For example, a group of sensors is set at a certain distance (such as 20 cm) in the traveling direction of the leveling device, and the lateral spacing of each group of sensors perpendicular to the traveling direction is 10 cm, forming a dense measurement network. Each laser ranging sensor emits a laser beam to the material surface, calculates the distance between the sensor and the material surface by measuring the time when the beam is reflected back, and then obtains the height values of each measurement point. By comparing the height values of each point in the same group of sensors, the difference between the maximum value and the minimum value is calculated, that is, the thickness deviation value of the corresponding area of this group of sensors is obtained. Through the comprehensive analysis of the data of all groups of sensors, the thickness deviation distribution of the entire material stack can be obtained.

[0041] The collection of humidity fluctuation values is achieved through humidity sensors integrated on the leveling device. The humidity sensors adopt an insertion design and can be automatically inserted into different depths (such as the surface layer, the middle layer, and the bottom layer) of the material stack during the operation of the leveling device for sampling. Each humidity sensor is equipped with a microprobe, and the surface of the probe is coated with an electrolyte material sensitive to moisture. The humidity value of the material is reflected by measuring the change in the conductivity of the electrolyte. To ensure the accuracy and representativeness of the data, multiple humidity sensors are set in different areas (such as the edge, the center, and the corner) of the material stack, and data is collected every certain time (such as 5 seconds), forming a fluctuation curve of humidity over time and space. The status monitoring module transmits the collected thickness deviation values and humidity fluctuation values to the grading module in real time to provide data support for the determination of the flatness grade.

[0042] The historical feature data stored in the storage unit comes from the system's records of past leveling operations. During each leveling operation, the status monitoring module will monitor the thickness change of the material stack in real time. When it detects that the thickness value deviates from the preset normal range (such as exceeding ±5 mm), it is determined as a thickness fluctuation, and the time point and duration (i.e., the adjustment duration) of the fluctuation are recorded. The historical thickness fluctuation times are the total number of thickness fluctuations that a certain material stack has occurred in past operations, and the adjustment duration is the time experienced from the start to the recovery of normal state for each fluctuation. For example, a certain material stack has had 3 thickness fluctuations in the past 10 operations, and the adjustment durations of each fluctuation are 10 minutes, 8 minutes, and 12 minutes respectively. These data will be completely recorded by the storage unit to form the historical feature data of this material stack. When processing this material stack again, the storage unit will transmit these historical data to the grading module as the basis for feature grade determination.

[0043] The determination process of the grading module is divided into two parts: characteristic grading determination and flatness grading determination. In the characteristic grading determination, first, obtain the historical thickness fluctuation times and corresponding adjustment duration data of each material stack from the storage unit. For each material stack, accumulate the adjustment duration of each fluctuation and then divide it by the number of fluctuations to obtain the average adjustment duration. For example, the adjustment durations of a certain material stack for 3 times are 10 minutes, 8 minutes, and 12 minutes respectively. After accumulation, it is 30 minutes, and the average duration is 10 minutes. The average adjustment duration reflects the average recovery speed of the thickness fluctuation of this material stack during past operations. The longer the average duration, the greater the difficulty in processing this material.

[0044] Next, based on the historical thickness fluctuation times and the average adjustment duration, calculate the characteristic coefficient through a preset comprehensive evaluation logic. The evaluation logic combines the weights of the fluctuation times and the average duration. For example, it is considered that the influence of the fluctuation times on the characteristic complexity accounts for 60%, and the influence of the average duration accounts for 40%. Specifically, divide the historical thickness fluctuation times into different intervals, such as 0 - 2 times, 3 - 5 times, 6 times and above, and each interval corresponds to a different basic coefficient; at the same time, divide the average adjustment duration into different intervals, such as 0 - 5 minutes, 6 - 10 minutes, 11 minutes and above, and each interval corresponds to a different correction coefficient. Obtain the final characteristic coefficient through the product or weighted summation of the basic coefficient and the correction coefficient. For example, the historical thickness fluctuation times of a certain material stack is 4 times (belonging to the 3 - 5 times interval, with a basic coefficient of 0.8), and the average adjustment duration is 9 minutes (belonging to the 6 - 10 minutes interval, with a correction coefficient of 0.9), then the characteristic coefficient is 0.8×0.9 = 0.72.

[0045] Compare the calculated characteristic coefficient with the preset characteristic threshold. The preset threshold is set in advance according to factors such as material type and operation experience. For example, the first - level characteristic threshold is set to be below 0.5, the second - level characteristic threshold is 0.5 - 0.8, and the third - level characteristic threshold is above 0.8. If the characteristic coefficient is less than 0.5, determine the historical characteristic level as the first level, indicating that the thickness fluctuation characteristic complexity of this material stack is relatively low and the processing difficulty is relatively small; if the characteristic coefficient is between 0.5 and 0.8, determine it as the second level, indicating medium characteristic complexity; if the characteristic coefficient is greater than or equal to 0.8, determine it as the third level, indicating high characteristic complexity and requiring a more refined processing strategy.

[0046] In the determination of the flatness level, the level division module synchronously obtains the thickness deviation value and the humidity fluctuation value transmitted by the status monitoring module. The thickness deviation value reflects the degree of undulation on the surface of the material stack. The larger the deviation value, the worse the surface flatness. The humidity fluctuation value reflects the uniformity of the humidity distribution inside the material. The larger the fluctuation value, the worse the humidity uniformity, which may affect the effect of the leveling operation. Through the preset comprehensive evaluation logic, the thickness deviation value and the humidity fluctuation value are normalized, converted into numerical values with a unified dimension, and then the flatness coefficient is calculated by the method of weighted summation. For example, the weight of the thickness deviation value is set to 70%, and the weight of the humidity fluctuation value is set to 30%. The thickness deviation coefficient is obtained by dividing the thickness deviation value by the maximum allowable deviation value, and the humidity fluctuation coefficient is obtained by dividing the humidity fluctuation value by the maximum allowable humidity fluctuation value. The flatness coefficient = thickness deviation coefficient × 0.7 + humidity fluctuation coefficient × 0.3.

[0047] The calculated flatness coefficient is compared with the preset flatness threshold, which is also set according to the material characteristics and operation requirements. For example, the first-level flatness threshold is below 0.3, the second level is 0.3 - 0.6, and the third level is above 0.6. If the flatness coefficient is less than 0.3, the flatness level is determined to be the first level, indicating that the flatness of the material stack is relatively high and the finishing requirement is low. If it is between 0.3 and 0.6, it is determined to be the second level, indicating medium flatness and a certain degree of finishing is required. If it is greater than or equal to 0.6, it is determined to be the third level, indicating poor flatness and a high finishing requirement.

[0048] Through the above process, the level division module can generate accurate historical feature levels and flatness levels for each material stack. These level information are transmitted to the strategy generation module as the core basis for generating dynamic adjustment strategies. For example, for a material stack with a feature level of the third level and a flatness level of the third level, the strategy generation module will prioritize the operation and call high-intensity leveling parameters (such as increasing the pressure of the leveling device and reducing the travel speed). For a material stack with a feature level of the first level and a flatness level of the first level, it can be arranged for subsequent operations and conventional leveling parameters can be used, thereby realizing the intelligent and refined control of the leveling operation and improving the operation efficiency and quality.

[0049] Example 3: In the overall architecture of the intelligent control of the leveling operation, the strategy generation module dynamically generates the optimal operation strategy based on multi-dimensional parameters such as the flatness level, historical feature level of the material stack, and the distance between the current position of the leveling device and the starting point of the operation. This process involves complex priority decision-making logic and path planning algorithms.

[0050] After the system starts up, the policy generation module first receives the flatness level and historical feature level data of each material stack transmitted by the grading module. The flatness level is divided into three levels: first level, second level, and third level, corresponding to different finishing requirements; the historical feature level is also divided into three levels, reflecting the complexity of material processing. At the same time, through the positioning system installed on the leveling device, the current position coordinates of the device are obtained in real time, and combined with the starting point coordinates of each material stack pre-stored in the storage unit, the distance between the leveling device and each starting point is calculated.

[0051] In the process of job priority decision-making, the policy generation module constructs a three-dimensional decision space. Among them, the flatness level is used as the first priority factor, and the system will give priority to processing the material stacks with a higher flatness level. This is because areas with poor flatness (such as the third-level flatness) usually require more finishing operations, and processing these areas as early as possible can avoid interference with subsequent operations. The historical feature level is used as the second priority factor. When encountering material stacks with the same flatness level, the system will give priority to selecting the one with a higher historical feature level for processing. For example, if there are two areas with a second-level flatness, but one of them has a historical feature level of three (high processing complexity) and the other has a level of one, the system will give priority to processing the area with a third-level feature to concentrate resources on solving complex problems. The distance between the current position of the leveling device and the starting point of the job is used as the third priority factor. When the flatness and historical feature levels are the same, the system will select the material stack closest to it for operation to reduce the equipment movement time and improve the overall operation efficiency.

[0052] The specific operation process is as follows: The leveling device starts operating from the current starting point of the job, first scans the material stacks within a certain range around it, and obtains their flatness level and historical feature level information. If there are multiple stacks with the same flatness level, the system will further compare their historical feature levels and give priority to selecting the one with a higher level for processing. For example, in a certain operation scenario, there are three material stacks with a second-level flatness around the leveling device, one of which has a historical feature level of three, and the other two have a level of two. At this time, the system will give priority to dispatching the leveling device to process the stack with a third-level feature.

[0053] When the leveling device reaches the demarcation point of the operation area, the system will make a decision based on the situation of the material stacks connected to the demarcation point. If the demarcation point is only connected to a single material stack, the leveling device directly calls the corresponding policy for processing. If the demarcation point is connected to multiple material stacks, the system first compares their flatness levels and gives priority to processing the one with a higher level; if the flatness levels are the same, then it compares the historical feature levels and gives priority to processing the one with a higher level; if all levels are the same, it randomly selects a material stack for processing through a random number generation algorithm.

[0054] If the demarcation point is not connected to other stacked bodies and there are unprocessed areas, the system will determine the positions of the unprocessed areas through the positioning system, calculate the distances between the leveling device and these areas, and select the nearest one for processing. After all areas are processed, the leveling device returns to the initial position and stops through the built-in navigation system.

[0055] During the entire operation process, the policy generation module will continuously update the operation policy according to real-time data. For example, when the leveling device is processing a certain material stacked body, if the status monitoring module feedbacks that the flatness or feature level of this area has changed, the policy generation module will re-evaluate the priority of this area and adjust the operation sequence if necessary.

[0056] To quantify this priority decision-making process, the system introduces a priority evaluation function:

[0057] where represents the comprehensive priority score of the material stacked body; represents the flatness level (1 for first level, 2 for second level, 3 for third level); represents the historical feature level (1 for first level, 2 for second level, 3 for third level); represents the distance from the current position of the leveling device to the starting point of the operation of this material stacked body; represents the farthest distance from the leveling device among all the material stacked bodies to be processed; , , are the weight coefficients of each factor respectively ( , , ). By comprehensively considering the flatness level, historical feature level and distance factors, this function calculates the priority score of each material stacked body. The system will give priority to processing the material stacked bodies with higher scores, so as to optimize the operation sequence.

[0058] Through this dynamic adjustment strategy, the leveling device can automatically adjust the operation sequence and intensity according to the actual situation of the materials, improving the operation efficiency and quality. The whole process does not require manual intervention, realizing the intelligent control of the leveling operation.

[0059] Embodiment 4: During the leveling operation, the distribution difference characteristics of the material stacked bodies have an important impact on the leveling effect. Through the collection, analysis and application of the fluctuation data, the dynamic correction of the movement trajectory of the leveling device is realized to ensure the operation accuracy. The following details the specific implementation process of this implementation method.

[0060] During the operation of the leveling device, the status monitoring module continuously collects real-time data of the material stack. These data include multi-dimensional features such as three-dimensional point cloud information, temperature distribution, and humidity change on the material surface. The distribution detection module extracts fluctuation data from these real-time data, specifically including the core coordinate points, reference fluctuation amplitude, and attenuation change rate. The core coordinate points are determined through three-dimensional modeling of the material stack and represent the central position of the distribution difference characteristics; the reference fluctuation amplitude is the fluctuation range of this characteristic under ideal conditions; the attenuation change rate reflects the attenuation speed of the characteristic fluctuation with the spatial distance.

[0061] After receiving the real-time motion trajectory data transmitted by the status monitoring module and the fluctuation data in the storage unit, the difference determination module first calculates the displacement difference between the real-time position of the leveling device and the core coordinate points of the distribution difference characteristics. This process is realized through a positioning system that uses multi-sensor fusion technology, combines GPS positioning, inertial navigation, and visual SLAM algorithms to ensure high-precision acquisition of the position data of the leveling device. For example, in a certain material processing scenario, the core coordinate points are determined to be (50, 30, 20) (unit: meters) through three-dimensional modeling, and the real-time position coordinates of the leveling device are (52, 31, 19). The displacement difference is calculated to be approximately 2.45 meters through the Euclidean distance formula.

[0062] Subsequently, the difference determination module extracts the reference fluctuation amplitude and attenuation change rate corresponding to the distribution difference characteristics from the fluctuation data. The reference fluctuation amplitude is determined by the material characteristics and historical operation data. For example, for a certain granular material, its reference fluctuation amplitude is set to 3 meters; the attenuation change rate is obtained by analyzing the change law of the characteristic fluctuation with distance, such as set to 0.5 meters / meter, indicating that for every 1 meter away from the core coordinate point, the fluctuation amplitude attenuates by 0.5 meters. By dividing the reference fluctuation amplitude by the attenuation change rate, the adjustment radius is calculated to be 6 meters (3 meters ÷ 0.5 meters / meter).

[0063] An adjustment area is constructed with the core coordinate points as the center and a radius of 6 meters. When the displacement difference of the leveling device is greater than 6 meters, it is determined that it is operating outside the adjustment area. At this time, the material characteristics are relatively stable, and the current operating state is maintained. If the displacement difference is less than or equal to 6 meters, it is determined that the leveling device enters the adjustment area and a trajectory correction signal is generated. For example, when the leveling device moves to the coordinates (54, 32, 18), the displacement difference is calculated to be 5.1 meters, which is less than the adjustment radius, and the system immediately generates a trajectory correction signal.

[0064] The trajectory correction signal contains correction direction and correction amount information. The correction direction is determined according to the type of distribution difference feature. For example, for the feature difference caused by uneven density, the correction direction points to the region with higher density; the correction amount is related to the ratio of the displacement difference to the adjustment radius. The larger the ratio, the larger the correction amount. In the above example, the ratio of the displacement difference to the adjustment radius is 5.1÷6≈0.85, and the system calculates the corresponding correction amount accordingly.

[0065] After receiving the trajectory correction signal, the distribution detection module analyzes the real-time motion trajectory of the leveling device. This module pre-sets the standard pressure range and the allowable offset threshold for the operation. For example, the standard pressure range is [80, 120] kPa, and the allowable offset threshold is ±10 kPa. The pressure sensor feeds back the operation pressure value in real time, and the speed sensor monitors the traveling speed. If the pressure value is within the range of 70 - 130 kPa and the traveling speed matches the pressure value (for example, when the pressure is 100 kPa, the speed should be between 0.5 - 0.7 m / s), it is considered that the operation state is normal and the current parameters are maintained.

[0066] If the value fed back by the pressure sensor exceeds the allowable offset threshold, or the traveling speed deviates from the reasonable range corresponding to the standard pressure range, the distribution detection module determines that the motion trajectory needs to be adjusted. For example, when the pressure value reaches 140 kPa and the traveling speed is 0.9 m / s, the system determines that the current operation parameters may cause excessive compaction of the material and generates a motion trajectory correction instruction.

[0067] After receiving the correction instruction, the parameter processing module controls the leveling device to perform trajectory adjustment. The adjustment methods include changing the traveling direction, adjusting the operation speed and pressure, etc. In the above example of abnormal pressure, the system first reduces the traveling speed to 0.4 m / s, and at the same time finely adjusts the angle of the leveling device to make it more conform to the material surface. After adjustment, the value fed back by the pressure sensor gradually returns to the standard range, and the system maintains the new operation parameters and continues the operation.

[0068] During the entire trajectory correction process, the state monitoring module continuously collects data, and the distribution detection module evaluates the adjustment effect in real time. If the adjusted parameters still do not meet the standards, the system will further optimize the correction strategy until the operation state returns to normal. Through this closed-loop feedback mechanism, the leveling device can adapt to the material distribution difference and ensure the stability of the operation quality.

[0069] In practical applications, different types of material stacks may have multiple distribution difference characteristics. The system calculates the adjustment area for each characteristic separately and dynamically determines the area where the leveling device is located based on its real-time position. For example, a certain material stack has two characteristics: density difference and humidity difference. The corresponding core coordinate points are (45, 28, 19) and (55, 32, 21) respectively, and the adjustment radii are 5 meters and 7 meters respectively. When the leveling device is located at (50, 30, 20), it is within both adjustment areas at the same time. The system will comprehensively consider the influences of the two characteristics and generate a composite correction signal.

[0070] For complex material scenarios, the system adopts a hierarchical processing strategy. First, the distribution difference characteristics are sorted according to their importance and influence range, and the characteristics with greater influence are processed preferentially. For example, when dealing with materials that have both density difference and humidity difference, if the density difference has a more significant impact on the leveling effect, the system will first correct the trajectory for the density difference characteristic. After the influence of this characteristic is reduced, the humidity difference characteristic will be processed.

[0071] During the trajectory correction process, the system also considers the mechanical performance limitations of the leveling device. For example, an excessive adjustment angle may cause the equipment to become unbalanced, and a sudden change in the traveling speed may cause vibration and other problems. Therefore, when generating correction instructions, the system combines the dynamic model of the equipment to ensure that the adjustment process is stable and controllable. For example, when a large change in the traveling direction is required, the system will adopt a progressive adjustment strategy, changing the direction in small increments multiple times to avoid the adverse effects caused by sudden turns of the equipment.

[0072] Example 5: During the leveling operation, the distribution detection module realizes precise control of the movement trajectory of the leveling device through real-time monitoring and intelligent analysis of the operation pressure and traveling speed. This module pre-sets the standard pressure range and allowable offset threshold for the operation. These parameters are determined through a large number of experiments and experience summaries based on factors such as material characteristics and leveling process requirements. For example, for a certain type of ore material with uniform particle size, the standard pressure range is set to 80 - 120 kPa, the allowable offset threshold is ±10 kPa, and the corresponding reasonable traveling speed range is 0.5 - 0.7 m / s.

[0073] When the leveling device starts operating, the pressure sensor installed at the bottom of the device collects operation pressure data at a frequency of 10 times per second in real time, and the speed sensor synchronously monitors the traveling speed. These data are transmitted to the distribution detection module for analysis in real time. In the leveling operation of a certain material stack, the initial pressure value feedback by the pressure sensor is 105 kPa, and the traveling speed is 0.6 m / s, both of which are within the standard range. The system determines that the operation state is normal, and the leveling device continues to operate according to the preset trajectory.

[0074] As the operation progresses, when the leveling device reaches the edge area of the material stack, the feedback value of the pressure sensor suddenly rises to 135 kPa, exceeding the upper limit of the allowable offset threshold (130 kPa). The distribution detection module immediately analyzes this abnormal data. Combining with the traveling speed of 0.65 m / s feedback by the speed sensor, it is judged that there may be a situation of uneven material stacking density. At this time, the system generates a first-level warning signal and triggers a preliminary trajectory correction strategy.

[0075] After receiving the warning signal, the parameter processing module first controls the leveling device to reduce the traveling speed to 0.4 m / s. At the same time, it increases the damping coefficient of the pressure regulation system to make the pressure change more smoothly. After the adjustment, it continuously monitors for 5 seconds. If the pressure value still does not return to the normal range, the in-depth correction strategy is started. In this example, after the speed is reduced, the pressure value briefly drops to 128 kPa, but then rises back to 132 kPa. The system determines that further adjustment is needed.

[0076] The distribution detection module calls the historical data in the storage unit to analyze the material characteristics of this area. Through comparison, it is found that there has been a density gradient change in this edge area during past operations. Based on this, the system generates a targeted correction instruction: The leveling device performs a spiral trajectory scan with a radius of 0.5 m centered on the current position, and at the same time, the pressure sensor encrypts and collects data at a frequency of 20 times per second. During the scan process, the system draws a three-dimensional map of the pressure change, clearly showing the trend of the material density increasing from the center to the edge.

[0077] Based on the analysis results of the map, the parameter processing module calculates the optimal operation path. The new path avoids the high-density area and gradually advances towards the edge in a roundabout way. During the advancement process, the system dynamically adjusts the pressure value and adaptively adjusts it within the range of 85 - 115 kPa according to the real-time detected density change. The traveling speed is also adjusted synchronously with the pressure change to ensure that the matching relationship between pressure and speed is always within the reasonable range corresponding to the standard pressure range.

[0078] When the leveling device enters the central area of the material stack, the feedback value of the pressure sensor stabilizes at about 95 kPa, but the speed sensor shows that the traveling speed drops to 0.4 m / s, lower than the lower limit of the standard speed range. The distribution detection module analyzes that it may be due to the higher humidity of the material in this area, resulting in an increase in friction. The system immediately generates a humidity anomaly warning and instructs the humidity detection probe installed on the leveling device to perform multi-point sampling.

[0079] The sampling results show that the humidity in the central area is 20% higher than that in the edge area. The system quickly calls the pre-stored humidity-pressure correction model, adjusts the standard pressure range to 90 - 130 kPa, and at the same time allows the traveling speed to fluctuate within the range of 0.4 - 0.6 m / s. After the adjustment, the operating pressure of the leveling device in this area automatically increases to 110 kPa, the traveling speed is maintained at 0.5 m / s, and the operating state returns to normal.

[0080] During the entire operation process, the distribution detection module continuously conducts correlation analysis on the pressure and speed data. By establishing a dynamic data window (such as the data in the most recent 30 seconds), the pressure-speed change rate is calculated. When the change rate exceeds the preset threshold (such as the pressure change exceeding 15 kPa per minute and the speed change exceeding 0.2 m / s), the system determines that there is an abnormal working condition and immediately triggers an emergency correction procedure.

[0081] For example, during a certain operation process, the pressure suddenly drops from 100 kPa to 75 kPa, and at the same time the speed soars from 0.6 m / s to 0.9 m / s. The distribution detection module detects this abnormal change within 0.3 seconds and quickly generates an emergency braking instruction. After receiving the instruction, the parameter processing module first cuts off the power output and at the same time activates the electromagnetic braking system to safely stop the leveling device within 0.5 m.

[0082] The system then enters the fault troubleshooting mode. By comparing the three-dimensional point cloud data and the historical operation records, it is found that there is a hidden depression in this area. The distribution detection module automatically generates a filling plan according to the size and depth of the depression: the leveling device first pre-presses the edge of the depression with a lower pressure (60 kPa), and then fills the material towards the center in a spiral trajectory, and the pressure gradually increases to the standard range during the filling process.

[0083] During the trajectory correction process, the distribution detection module uses multi-sensor fusion technology to ensure the correction accuracy. In addition to the pressure and speed sensors, it also combines the data of the laser rangefinder, the inclination sensor, and the acceleration sensor for comprehensive judgment. For example, when the pressure abnormally increases, the system will simultaneously analyze the change in the surface height of the material feedback by the laser rangefinder and the equipment attitude data detected by the inclination sensor to exclude misjudgments caused by equipment tilt.

[0084] In a complex material scenario, the leveling device encounters a protrusion with an irregular shape. The pressure sensor feedback shows that the pressure value fluctuates greatly (80 - 140 kPa), and the laser rangefinder indicates that the height of this area is 15 cm higher than the surrounding area. After comprehensive analysis by the distribution detection module, it is determined to be a hard caking. The system immediately switches to the special treatment mode: the leveling device first performs a circular cutting around the edge of the caking to reduce the connection strength between the caking and the surrounding materials; then gradually increases the pressure to 150 kPa (short-term overlimit), while reducing the traveling speed to 0.3 m / s to crush the caking; finally, levels the crushed area with the standard pressure.

[0085] The distribution detection module also has the ability of adaptive learning. After each operation is completed, the system compares the actual operation parameters with the preset standards and establishes a deviation database. By analyzing these deviation data through machine learning algorithms, the standard pressure range and the allowable offset threshold are continuously optimized. For example, after processing a certain type of material multiple times, the system finds that the optimal pressure range in actual operation is 5 - 10 kPa lower than the preset value, so it automatically adjusts the standard parameters to improve the operation efficiency and quality.

[0086] During the long-term continuous operation process, the distribution detection module will continuously monitor the equipment status parameters (such as motor temperature, hydraulic system pressure, etc.) and conduct correlation analysis with the operation pressure and speed data. When it is found that the equipment status parameters are abnormal and accompanied by fluctuations in operation parameters, the system will give an early warning of potential failures. For example, when the motor temperature rises by 10℃ and the pressure value shows periodic fluctuations, the system anticipates that it may be due to the loose drive belt and promptly reminds the operator to perform maintenance.

[0087] Through this intelligent pressure - speed collaborative control mechanism, the leveling device can automatically adapt to the unevenness of material distribution, effectively avoiding problems such as excessive material compaction caused by overpressure or poor leveling effect caused by underpressure. The entire operation process requires no manual intervention, realizing the full-closed-loop automatic control from data acquisition, analysis to decision execution, significantly improving the accuracy and efficiency of the leveling operation, and reducing the labor cost and equipment wear.

[0088] It should be noted that in this article, relational terms such as first and second are only used 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 term "including", "comprising" or any other variant thereof is intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements not only includes those elements, but also includes other elements not expressly listed, or also includes elements inherent to such process, method, article or device.

[0089] Although embodiments of the present invention have been shown and described, it will be understood by those of ordinary skill in the art that various changes, modifications, substitutions and variations can be made to these embodiments without departing from the principles and spirit of the present invention, and the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. An intelligent control system for a flat material device, characterized in that, It includes a material identification module, a parameter processing module, a grading module, a storage unit, a strategy generation module, a status monitoring module, a distribution detection module, and a difference determination module; The material identification module is used to divide the flat area and the area to be processed of the material stack; the parameter processing module is used to control the flatting device to start the flatting operation after receiving an instruction; when performing the initial flatting operation, the status monitoring module is used to collect the basic form data of the material stack and transmit it to the grading module; the storage unit is used to store the historical feature data of the material stack and transmit it to the grading module; the grading module is used to determine the feature grade and flatness grade of the material stack, generate the historical feature grade and flatness grade, and transmit them to the strategy generation module; the strategy generation module is used to generate a dynamic adjustment strategy for the flatting device during operation according to the flatness grade and historical feature grade, and transmit the dynamic adjustment strategy to the parameter processing module; When the flatting device performs the flatting operation according to the dynamic adjustment strategy, the status monitoring module is also used to collect the real-time movement trajectory of the flatting device and transmit it to the difference determination module; the storage unit is also used to store the distribution difference features in the material stack and the fluctuation data corresponding to the distribution difference features and transmit them to the difference determination module; The difference determination module is used to determine whether the flatting device enters the adjustment area. If a trajectory correction signal is generated, it will transmit the real-time movement trajectory of the flatting device in the adjustment area to the distribution detection module; the distribution detection module is used to judge whether the flatting device deviates from the reference path during the operation in the adjustment area according to the real-time movement trajectory. If it deviates, it will send a movement trajectory correction instruction to the parameter processing module.

2. The intelligent control system of a flat material device according to claim 1, characterized in that It also includes a surface analysis module and an operation terminal. The status monitoring module is also used to collect the real-time surface data of the material stack and transmit it to the surface analysis module. The surface analysis module is used to analyze the real-time surface data of the material stack in real time. If a defect signal or a deformation signal is generated by the analysis, it will transmit the coordinate mark and the corresponding real-time surface data to the operation terminal. The operation terminal is used for visual processing of the real-time surface data of the material stack.

3. The intelligent control system of a flat material device according to claim 1, characterized in that, The basic form data includes the thickness deviation value and humidity fluctuation value of the material stack; The historical feature data is the historical thickness fluctuation times of the material stack and the adjustment duration corresponding to each fluctuation.

4. The intelligent control system of a flat material device according to claim 3, characterized in that The determination process of the grading module is specifically as follows: Obtain the historical thickness fluctuation times of each material stack, and then obtain the adjustment duration corresponding to each fluctuation of each material stack. Accumulate and calculate the mean value of the adjustment duration corresponding to each fluctuation to obtain the adjustment average duration of each material stack; Calculate the feature coefficient of each material stack based on the historical thickness fluctuation times and the adjustment average duration; Compare the feature coefficient with the preset feature threshold to determine that the historical feature grade of the material stack is a first-level feature, a second-level feature, or a third-level feature; Synchronously obtain the thickness deviation value and humidity fluctuation value of each material stack, and calculate the flatness coefficient of each material stack; Compare the flatness coefficient with the preset flatness threshold to determine that the flatness grade of the material stack is a first-level flatness, a second-level flatness, or a third-level flatness.

5. The intelligent control system of a flat material device according to claim 4, characterized in that, The complexity corresponding to the third-level features is higher than that of the second-level features, and the complexity corresponding to the second-level features is higher than that of the first-level features; The finishing requirement corresponding to the third-level flatness is higher than that of the second-level flatness, and the finishing requirement corresponding to the second-level flatness is higher than that of the first-level flatness.

6. The intelligent control system of a flat material device according to claim 5, characterized in that, The working process of the strategy generation module is specifically as follows: The flattening device takes the flatness level of the material stack as the first priority factor, the historical feature level of the material stack as the second priority factor, and the distance between the current position of the flattening device and the starting point of the operation of the material stack as the third priority factor; The flattening device starts operating from the current starting point of operation. If there are multiple material stacks with the same flatness level near the starting point, it preferentially selects the material stack with a higher historical feature level for processing; If the flatness levels and historical feature levels of the material stacks are the same, it selects the material stack closest to the starting point of operation for processing; When the flattening device reaches the demarcation point: If the demarcation point is only connected to a single material stack, the flattening device directly processes the material stack; If the demarcation point is connected to two or more material stacks, it preferentially processes the material stack with a higher flatness level; when the flatness levels of the multiple material stacks connected by the demarcation point are the same, it preferentially processes the material stack with a higher historical feature level; when the flatness levels and historical feature levels of the material stacks connected by the demarcation point are the same, it randomly selects a material stack for processing; If the demarcation point is not connected to other material stacks and there is an unprocessed area, it preferentially processes the unprocessed area closest to it; when the flattening device reaches any demarcation point, if all other areas have been processed, it returns to the initial position and stops; Generate a dynamic adjustment strategy for the flattening device during operation according to the above rules; When the flattening device operates according to the dynamic adjustment strategy, it performs operations with different operation intensities in areas corresponding to different flatness levels or third-level features.

7. The intelligent control system of a flat material device according to claim 1, characterized in that, The fluctuation data are the core coordinate points, the reference fluctuation amplitude, and the attenuation change rate of the distribution difference features.

8. The intelligent control system of a flat material device according to claim 7, characterized in that, The determination process of the difference determination module is specifically as follows: Calculate the displacement difference between the real-time movement trajectory of the flattening device and the core coordinate points of the distribution difference features; Obtain the fluctuation data corresponding to the distribution difference features in the material stack, extract the core coordinate points, the reference fluctuation amplitude, and the attenuation change rate, and obtain the adjustment radius of the distribution difference features by dividing the reference fluctuation amplitude by the attenuation change rate; Taking the core coordinate points as the center point and the adjustment radius as the coverage range, construct the adjustment area of the distribution difference features; When the displacement difference is greater than the adjustment radius, the flattening device performs operations outside the adjustment area and maintains the current operating state; When the displacement difference is less than or equal to the adjustment radius, the flattening device enters the adjustment area to perform operations and generates a trajectory correction signal.

9. The intelligent control system of a flat material device according to claim 8, characterized in that, The judgment process of the distribution detection module is specifically as follows: Set the standard pressure range and the allowable offset threshold for the operation of the flattening device; If the value feedback by the pressure sensor is within the allowable offset threshold and the traveling speed of the flattening device is within the standard pressure range, maintain the current operation parameters; If the value feedback by the pressure sensor exceeds the allowable offset threshold or the traveling speed of the flattening device deviates from the standard pressure range, generate a motion trajectory correction instruction.

10. The intelligent control system of a flat material device according to claim 2, characterized in that, The parsing process of the surface analysis module is specifically as follows: Capture the three-dimensional point cloud data of the material stack; Convert the point cloud data into a two-dimensional projection image, use median filtering to eliminate noise, set a filtering threshold according to the height difference between the material surface morphology and the background, and extract the effective surface area of the material stack; apply morphological closing operation to fill small holes and retain the complete surface contour features; Detect the straight edges through the Hough transform, match the geometric structure features of the material stack, and filter out the non-structural interference areas; Synchronously obtain the infrared thermal imaging data of the material stack, set the temperature gradient range to identify the abnormally heated areas, use the pseudo-color enhancement technology to strengthen the temperature difference display, and combine with the edge detection algorithm to extract the temperature mutation boundary to identify the internal structure abnormality of the material; If both the three-dimensional point cloud data and the infrared thermal imaging data show abnormalities, generate a defect signal and mark the corresponding coordinates; If there is no abnormality in both types of data, maintain the normal operation process; If any one of the data is abnormal, generate a deformation signal and mark the corresponding coordinates.

Images