Engineering material moisture content detection method and system based on artificial intelligence

Abstract

The invention belongs to the technical field of engineering material detection, and discloses an artificial intelligence-based engineering material moisture content detection method and system. Comprising the following modules: a data acquisition module used for acquiring water content data of engineering materials in a storehouse; according to the method, intelligent path planning is set, a material area is divided into different priority levels according to the water content score and the change rate, differentiated drying strategies are adopted for different levels, to-be-dried cells are sequentially connected from an inlet to an outlet of a warehouse, and an initial drying path is generated; the patrol type drying robot can efficiently cover all key areas, mobile type drying equipment is intelligently deployed on a key common side based on a placement coefficient prediction model, a cooperative work network of fixed type, mobile type and patrol type drying equipment is formed, path conflicts are detected in real time, and the detection efficiency is improved. Working overlapping of the patrol type drying robot and the movable type drying equipment is avoided, and an optimized drying path is generated through dynamic re-planning.

Description

Technical Field

[0001] This invention relates to the field of engineering material testing technology, and more specifically, to an artificial intelligence-based method and system for detecting the moisture content of engineering materials. Background Technology

[0002] In the fields of construction engineering, civil engineering, and material storage, the moisture content of engineering materials is a key parameter affecting material performance, construction quality, and engineering safety. With the development of Internet of Things (IoT) technology, some automated moisture content monitoring systems have emerged in existing technologies. By deploying sensor networks in storage areas, a certain degree of real-time monitoring has been achieved.

[0003] However, these systems generally suffer from technical defects. Most systems rely on a single type of sensor and lack multimodal data fusion capabilities, failing to comprehensively reflect the true moisture content of materials. Existing systems have relatively simple response strategies, typically employing fixed threshold triggering mechanisms, which cannot intelligently predict and grade responses based on moisture content changes. In terms of drying equipment scheduling, traditional systems often use fixed drying equipment for full coverage or simple manual scheduling, lacking intelligent drying mechanisms based on spatial analysis, resulting in low energy efficiency and uneven drying effects. Especially in large engineering material warehouses, the uneven material packing density, complex spatial distribution, and dynamic changes in environmental factors make it difficult for single detection methods and fixed response strategies to meet actual needs. Abnormal moisture content in local areas of materials is often difficult to detect and handle in a timely manner, easily leading to problems such as material deterioration, strength reduction, and even mold growth. Summary of the Invention

[0004] To address the problems in the background art, this invention proposes an artificial intelligence-based method and system for detecting the moisture content of engineering materials.

[0005] To achieve the above objectives, the present invention provides the following technical solution: an artificial intelligence-based moisture content detection system for engineering materials, comprising the following modules: The data acquisition module is used to collect moisture content data of engineering materials in the warehouse; The moisture content assessment module is used to generate real-time moisture content scores for engineering materials based on the collected moisture content data, and to derive the rate of change of moisture content scores and the average moisture content score of engineering materials based on multiple moisture content scores. The drying path planning module is used to classify engineering materials according to the rate of change of moisture content score and moisture content score to obtain cells to be dried, and to plan the initial drying path of the patrol drying robot based on all cells to be dried. The placement coefficient prediction module is used to acquire historical placement coefficient data and build a placement coefficient prediction model. The placement coefficient prediction model is used to predict the placement coefficient of the common side of the cells to be dried, and the decision is made on whether to place the mobile drying equipment on the common side based on the placement coefficient. The drying path optimization module is used to optimize the initial drying path based on the mobile drying equipment placed on the shared side to obtain the optimized drying path.

[0006] Furthermore, the engineering materials are stored in a warehouse equipped with a dehumidifier. The warehouse area is divided into multiple cells, and the engineering materials are stacked in the cells. The edges of the cells form walkways. Each cell is equipped with drying equipment and testing equipment. The drying equipment is a fixed drying equipment used to reduce the moisture content of the engineering materials, and the testing equipment is used to collect moisture content data, including ambient humidity, material type coefficient, ambient temperature, air velocity, and material bulk density. Ambient humidity is obtained through a temperature and humidity sensor; Material type coefficients are obtained from a materials database; Ambient temperature is obtained through a temperature sensor; Air velocity is obtained through a wind speed sensor; The weight of the engineering material is obtained by using a weight sensor, and the bulk density of the material is obtained by dividing the weight by the volume of the engineering material.

[0007] Furthermore, the process of generating real-time moisture content scores for engineering materials based on the collected moisture content data includes: Moisture content score S: ;

[0008] In the formula, , , and These are weighting coefficients, derived from historical data training. The measured ambient humidity At maximum humidity, For material type coefficient, The actual ambient temperature. This is the upper limit of temperature. This is the measured airflow velocity. This is the upper limit of the flow rate. This represents the measured bulk density of the material. This represents the maximum density.

[0009] Furthermore, the process of deriving the rate of change of moisture content score and the average moisture content score of engineering materials based on multiple moisture content scores includes: Set an evaluation cycle. At the end of each evaluation cycle, the real-time moisture content score of the engineering materials in each cell needs to be obtained. The average moisture content score of the engineering materials in the cell is obtained based on the m moisture content scores. For each evaluation cycle, the moisture content scores of the previous 2n evaluation cycles are taken. The 2n moisture content scores are arranged in chronological order. The average of the first n moisture content scores is calculated to obtain the first average moisture content score. The average of the last n moisture content scores is calculated to obtain the second average moisture content score. The average of the second average moisture content score is subtracted from the average of the first average moisture content score to obtain the score difference. The score difference is divided by the average of the first average moisture content score to obtain the rate of change of moisture content score.

[0010] Furthermore, the process of grading engineering materials based on the rate of change of moisture content score and moisture content score to obtain cells to be dried includes: Based on historical moisture content score change rate data, set appropriate first change rate threshold and second change rate threshold. The historical moisture content score change rate data refers to the data set of previous engineering material moisture content score change rates. Compare the engineering material moisture content score change rate with the two change rate thresholds. Based on historical moisture content scoring data, set appropriate first and second scoring thresholds. The historical moisture content scoring data refers to the data set of previous engineering material moisture content scores. Compare the engineering material moisture content scores with the two scoring thresholds. If the rate of change of the moisture content score of an engineering material is less than the first rate of change threshold and the moisture content score is less than the first score threshold, the engineering material is judged to be a low moisture content engineering material. If the first rate of change threshold is less than the rate of change of the moisture content score of the engineering material, which is less than the second rate of change threshold, or if the first score threshold is less than the moisture content score of the engineering material, which is less than the second score threshold, then the engineering material is determined to be a medium moisture content engineering material, and the drying equipment in the corresponding cell needs to be activated. If the rate of change of the moisture content score of the engineering material is greater than the second rate of change threshold or the moisture content score of the engineering material is greater than the second score threshold, the engineering material is determined to be a high moisture content engineering material, and the drying equipment in the corresponding cell needs to be activated, and the corresponding cell is marked as a cell to be dried.

[0011] Furthermore, the process of planning the initial drying path for the patrol-type drying robot based on all cells to be dried includes: The patrol-type drying robot refers to a robot that patrols within a warehouse and assists in the drying of engineering materials; the initial drying path refers to the walking path of the patrol-type drying robot within the warehouse. Starting from the warehouse entrance, which serves as the path starting point, the nearest cells to be dried are iteratively selected and added to the path until all cells to be dried are covered. Each cell can have only one edge added to the path. Finally, the path is connected to the warehouse exit to form a complete path, thus obtaining the initial drying path of the patrol drying robot.

[0012] Furthermore, the process of acquiring historical placement coefficient data and constructing a placement coefficient prediction model, and then using this model to predict the placement coefficient of the shared edge of the cells to be dried, includes: The shared edge refers to the edge shared by two adjacent cells. The placement coefficient refers to the probability coefficient of placing a mobile drying device on the shared edge. The mobile drying device is used to move back and forth within the shared edge to assist in drying the engineering materials in the cells on both sides. It is necessary to predict the placement coefficient of the shared edge of each cell to be dried. Factors affecting the placement coefficient of shared sides of cells to be dried include: distance between cells, first average moisture content score, second average moisture content score, rate of change of first moisture content score, rate of change of second moisture content score, length of shared side, total area, and material hygroscopicity coefficient. The distance between them refers to the distance from the nearest fixed drying equipment to the shared edge; The first average moisture content score and the rate of change of the first moisture content score refer to the average moisture content score and the rate of change of the moisture content score of one of the cells corresponding to the shared edge; The second average moisture content score and the second moisture content score change rate refer to the average moisture content score and the moisture content score change rate of the other cell corresponding to the shared edge; The length of a shared side refers to the length of the side that shares a common edge. The total area refers to the sum of the areas of the two cells that share a common edge; The moisture absorption coefficient of the engineering materials is obtained by querying the database for the cells on both sides of the shared edge and taking the average of the two moisture absorption coefficients. Set a testing cycle and obtain historical placement coefficient data of engineering materials in each cell to be dried within the testing cycle. The historical placement coefficient data includes the distance between engineering materials in each cell to be dried within the testing cycle, the first average moisture content score, the second average moisture content score, the change rate of the first moisture content score, the change rate of the second moisture content score, the length of the common side, the total area, the material hygroscopicity coefficient, and the historical placement coefficient of engineering materials in each cell to be dried within the testing cycle. Based on the distance between engineering materials in each cell to be dried in different historical placement coefficient data during the detection cycle, the first average moisture content score, the second average moisture content score, the rate of change of the first moisture content score, the rate of change of the second moisture content score, the common side length, the total area, the material hygroscopic coefficient, and the corresponding historical placement coefficient, a placement coefficient prediction set is generated and divided into a training set and a test set. A convolutional neural network is constructed, and the distance between different historical placement coefficients in the training set, the first average moisture content score, the second average moisture content score, the rate of change of the first moisture content score, the rate of change of the second moisture content score, the length of the common side, the total area, and the material hygroscopic coefficient are used as the input data of the convolutional neural network. The corresponding historical placement coefficients in the training set are used as the output data of the convolutional neural network. The convolutional neural network is trained to obtain an initial convolutional neural network. The initial convolutional neural network is validated using a test set. The initial convolutional neural network whose output is less than or equal to the preset test error threshold is used as the placement coefficient prediction model. The distance between the engineering materials in each cell to be dried during the testing period, the first average moisture content score, the second average moisture content score, the rate of change of the first moisture content score, the rate of change of the second moisture content score, the length of the common side, the total area, and the material hygroscopic coefficient are input into the placement coefficient prediction model to obtain the predicted placement coefficient of the engineering materials in each cell to be dried.

[0013] Furthermore, the process of deciding whether to place the mobile drying equipment on the shared side based on the placement coefficient includes: Set an appropriate placement coefficient threshold based on historical placement coefficient data, where historical placement coefficient data refers to the data set of past shared edge placement coefficients, and compare the predicted placement coefficient of the shared edge with the placement coefficient threshold. When the predicted placement coefficient of the shared edge is greater than the placement coefficient threshold, a mobile drying device is placed on the shared edge. The mobile drying device reciprocates on the shared edge at a set speed to dry the engineering materials in the cells on both sides of the shared edge.

[0014] Furthermore, the process of optimizing the initial drying path based on the mobile drying equipment placed along the shared edge to obtain the optimized drying path includes: Check whether the initial drying path passes through the shared edge of the already placed mobile drying equipment. The shared edge of the already placed mobile drying equipment is called the conflict edge. The mobile drying equipment is already working on this edge, and the patrol drying robot passing through this edge will cause redundancy or interference. The initial drying path is divided into segments. Each segment is checked to see if it intersects with a conflict edge. If it intersects, an alternative path is searched in the vicinity of the segment. The shortest path from the current point to the target point is used as the alternative path to avoid passing through conflict edges. The alternative path must still be an edge of the cell to be dried. The path sequence is updated to ensure that all cells to be dried are still visited, and finally the optimized drying path is obtained. A patrol cycle is set, during which the patrol-type drying robot patrols along the optimized drying path and assists in drying the engineering materials in the cells along the path. After each patrol cycle, the drying path of the patrol-type drying robot is regenerated and optimized based on the real-time moisture content score and the predicted placement coefficient.

[0015] An artificial intelligence-based method for detecting the moisture content of engineering materials includes the following steps: S1: Collect the moisture content data of engineering materials in each cell of the warehouse in real time through the detection equipment, and perform normalization preprocessing on the moisture content data; S2: Based on the pre-processed moisture content data, evaluate and calculate the real-time moisture content score of the engineering materials in each cell, periodically evaluate the moisture content status of the engineering materials, and calculate the rate of change of moisture content score through time series analysis. S3: Based on the moisture content score and the rate of change of the moisture content score, a three-level classification mechanism is used to classify the drying needs of cells, and cells containing engineering materials with high moisture content are marked as cells to be dried, forming a set of areas to be processed, which provides input for subsequent path planning; S4: Based on the spatial distribution of the cells to be dried, generate the initial drying path of the patrol drying robot from the warehouse entrance to the exit. Evaluate the deployment requirements of mobile drying equipment on each common side through the placement coefficient prediction model, and predict the placement coefficient of mobile drying equipment on the common side. S5: Based on the predicted placement coefficient, place the mobile drying equipment on the corresponding shared edge, detect the conflict between the initial drying path and the shared edge where the mobile drying equipment is placed, identify the conflicting edges that need to be avoided, and obtain the optimized drying path; S6: The patrol-type drying robot performs tasks according to the optimized drying path, monitors the drying effect in real time, and dynamically adjusts the drying strategy and path planning based on feedback data.

[0016] The technical effects and advantages of the artificial intelligence-based method and system for detecting the moisture content of engineering materials in this invention are as follows: (1) By setting up a placement coefficient prediction model and path planning, the efficient scheduling and optimized allocation of drying resources are realized. The material area is divided into different priority levels according to the moisture content score and change rate. Different drying strategies are adopted for different levels. The initial drying path is generated by connecting the cells to be dried sequentially from the warehouse entrance to the exit, ensuring that the patrol drying robot can efficiently cover all key areas. Based on the placement coefficient prediction model, mobile drying equipment is intelligently deployed on key common edges to form a collaborative working network of fixed, mobile and patrol drying equipment. Path conflicts are detected in real time to avoid the overlap of work between the patrol drying robot and the mobile drying equipment. Optimized drying paths are generated through dynamic replanning.

[0017] (2) By setting up a moisture content scoring model, the moisture content score is calculated after normalizing each parameter. This effectively overcomes the limitations and instability of single sensor detection, regularly assesses the moisture content status, and accurately calculates the rate of change of the moisture content score by comparing the average score of different time windows. This helps to identify the dynamic trend of the material's moisture content status. The trend analysis based on historical data can provide early warning of abnormal increases in moisture content, realizing the transformation from passive response to active prediction. By constructing a dual evaluation system of moisture content score and rate of change, comprehensive monitoring of the material's moisture content status is achieved. It not only focuses on the current moisture content level but also closely monitors the changing trend, providing a more scientific basis for drying decisions. Attached Figure Description

[0018] Figure 1 This is a schematic diagram of the system of the present invention. Detailed Implementation

[0019] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of the present invention.

[0020] Reference Figure 1 An AI-based moisture content detection system for engineering materials includes the following modules: The data acquisition module is used to collect moisture content data of engineering materials in the warehouse; The moisture content assessment module is used to generate real-time moisture content scores for engineering materials based on the collected moisture content data, and to derive the rate of change of moisture content scores and the average moisture content score of engineering materials based on multiple moisture content scores. The drying path planning module is used to classify engineering materials according to the rate of change of moisture content score and moisture content score to obtain cells to be dried, and to plan the initial drying path of the patrol drying robot based on all cells to be dried. The placement coefficient prediction module is used to acquire historical placement coefficient data and build a placement coefficient prediction model. The placement coefficient prediction model is used to predict the placement coefficient of the common side of the cells to be dried, and the decision is made on whether to place the mobile drying equipment on the common side based on the placement coefficient. The drying path optimization module is used to optimize the initial drying path based on the mobile drying equipment placed on the shared side to obtain the optimized drying path.

[0021] It should be further explained that, in the specific implementation process, the engineering materials are stored in a warehouse equipped with a dehumidifier. The warehouse area is divided into multiple cells, and the engineering materials are stacked in the cells. The edges of the cells form walkways. Each cell is equipped with drying equipment and testing equipment. The drying equipment is a fixed drying equipment used to reduce the moisture content of the engineering materials, and the testing equipment is used to collect moisture content data. The moisture content data includes ambient humidity, material type coefficient, ambient temperature, air velocity, and material bulk density. Ambient humidity is obtained through a temperature and humidity sensor; The material type coefficient is obtained from the material database. Different materials have different coefficients, such as 1.0 for wood and 0.7 for concrete. The higher the material type coefficient, the more easily the material absorbs moisture. Ambient temperature is obtained through a temperature sensor; Air velocity is obtained through a wind speed sensor; The weight of the engineering material is obtained by using a weight sensor, and the bulk density of the material is obtained by dividing the weight by the volume of the engineering material.

[0022] It should be further explained that, in the specific implementation process, the process of generating real-time moisture content scores for engineering materials based on the collected moisture content data includes: Moisture content score S: ;

[0023] In the formula, , , and These are weighting coefficients, derived from historical data training, specifically 0.25, 0.25, 0.2, 0.15, and 0.15. The measured ambient humidity This represents the maximum humidity, specifically 100%. For material type coefficient, The actual ambient temperature. This is the upper limit of the temperature range, specifically 50℃. This is the measured airflow velocity. This is the upper limit for flow velocity, specifically 5 m / s. This represents the measured bulk density of the material. For maximum density, specifically ; If the moisture content data of the engineering material in a certain cell at a certain moment is: 80% of the material is concrete. 20℃ It is 2m / s. for If the moisture content score S of the engineering material in that cell at that moment is 0.59.

[0024] It should be further explained that, in the specific implementation process, the process of deriving the rate of change of moisture content score and the average moisture content score of the engineering materials based on multiple moisture content scores includes: Set an evaluation cycle. At the end of each evaluation cycle, the real-time moisture content score of the engineering materials in each cell needs to be obtained. The average moisture content score of the engineering materials in the cell is obtained based on the m moisture content scores. For each evaluation cycle, the moisture content scores of the previous 2n evaluation cycles are taken. The 2n moisture content scores are arranged in chronological order. The average of the first n moisture content scores is calculated to obtain the first average moisture content score. The average of the last n moisture content scores is calculated to obtain the second average moisture content score. The average of the second average moisture content score is subtracted from the average of the first average moisture content score to obtain the score difference. The score difference is divided by the average of the first average moisture content score to obtain the rate of change of moisture content score.

[0025] It should be further explained that, in the specific implementation process, the process of classifying the engineering materials according to the rate of change of moisture content score and moisture content score to obtain the cells to be dried includes: Based on historical moisture content score change rate data, set appropriate first change rate threshold and second change rate threshold. The historical moisture content score change rate data refers to the data set of previous engineering material moisture content score change rates. Compare the engineering material moisture content score change rate with the two change rate thresholds. Based on historical moisture content scoring data, set appropriate first and second scoring thresholds. The historical moisture content scoring data refers to the data set of previous engineering material moisture content scores. Compare the engineering material moisture content scores with the two scoring thresholds. If the rate of change of the moisture content score of an engineering material is less than the first rate of change threshold and the moisture content score is less than the first score threshold, the engineering material is judged to be a low moisture content engineering material. If the first rate of change threshold is less than the rate of change of the moisture content score of the engineering material, which is less than the second rate of change threshold, or if the first score threshold is less than the moisture content score of the engineering material, which is less than the second score threshold, then the engineering material is determined to be a medium moisture content engineering material, and the drying equipment in the corresponding cell needs to be activated. If the rate of change of the moisture content score of the engineering material is greater than the second rate of change threshold or the moisture content score of the engineering material is greater than the second score threshold, the engineering material is determined to be a high moisture content engineering material, and the drying equipment in the corresponding cell needs to be activated, and the corresponding cell is marked as a cell to be dried.

[0026] It should be further explained that, in the specific implementation process, the process of planning the initial drying path of the patrol-type drying robot based on all cells to be dried includes: The patrol-type drying robot refers to a robot that patrols within a warehouse and assists in the drying of engineering materials; the initial drying path refers to the walking path of the patrol-type drying robot within the warehouse. Starting from the warehouse entrance, which serves as the path starting point, the nearest cells to be dried are iteratively selected and added to the path until all cells to be dried are covered. Each cell can have only one edge added to the path. Finally, the path is connected to the warehouse exit to form a complete path, thus obtaining the initial drying path of the patrol drying robot.

[0027] It should be further explained that, in the specific implementation process, the process of obtaining historical placement coefficient data and constructing a placement coefficient prediction model, and using the placement coefficient prediction model to predict the placement coefficient of the shared edge of the cells to be dried, includes: The shared edge refers to the edge shared by two adjacent cells. The placement coefficient refers to the probability coefficient of placing a mobile drying device on the shared edge. The mobile drying device is used to move back and forth within the shared edge to assist in drying the engineering materials in the cells on both sides. It is necessary to predict the placement coefficient of the shared edge of each cell to be dried. Factors affecting the placement coefficient of shared sides of cells to be dried include: distance between cells, first average moisture content score, second average moisture content score, rate of change of first moisture content score, rate of change of second moisture content score, length of shared side, total area, and material hygroscopicity coefficient. The distance between them refers to the distance from the nearest fixed drying equipment to the shared edge; The first average moisture content score and the rate of change of the first moisture content score refer to the average moisture content score and the rate of change of the moisture content score of one of the cells corresponding to the shared edge; The second average moisture content score and the second moisture content score change rate refer to the average moisture content score and the moisture content score change rate of the other cell corresponding to the shared edge; The length of a shared side refers to the length of the side that shares a common edge. The total area refers to the sum of the areas of the two cells that share a common edge; The moisture absorption coefficient of the engineering materials is obtained by querying the database for the cells on both sides of the shared edge and taking the average of the two moisture absorption coefficients. Set a testing cycle and obtain historical placement coefficient data of engineering materials in each cell to be dried within the testing cycle. The historical placement coefficient data includes the distance between engineering materials in each cell to be dried within the testing cycle, the first average moisture content score, the second average moisture content score, the change rate of the first moisture content score, the change rate of the second moisture content score, the length of the common side, the total area, the material hygroscopicity coefficient, and the historical placement coefficient of engineering materials in each cell to be dried within the testing cycle. Based on the distance between engineering materials in each cell to be dried in different historical placement coefficient data during the detection cycle, the first average moisture content score, the second average moisture content score, the rate of change of the first moisture content score, the rate of change of the second moisture content score, the common side length, the total area, the material hygroscopic coefficient, and the corresponding historical placement coefficient, a placement coefficient prediction set is generated and divided into a training set and a test set. A convolutional neural network is constructed, and the distance between different historical placement coefficients in the training set, the first average moisture content score, the second average moisture content score, the rate of change of the first moisture content score, the rate of change of the second moisture content score, the length of the common side, the total area, and the material hygroscopic coefficient are used as the input data of the convolutional neural network. The corresponding historical placement coefficients in the training set are used as the output data of the convolutional neural network. The convolutional neural network is trained to obtain an initial convolutional neural network. The initial convolutional neural network is validated using a test set. The initial convolutional neural network whose output is less than or equal to the preset test error threshold is used as the placement coefficient prediction model. The distance between the engineering materials in each cell to be dried during the testing period, the first average moisture content score, the second average moisture content score, the rate of change of the first moisture content score, the rate of change of the second moisture content score, the length of the common side, the total area, and the material hygroscopic coefficient are input into the placement coefficient prediction model to obtain the predicted placement coefficient of the engineering materials in each cell to be dried. In an embodiment of the present invention, the predicted placement coefficient of the engineering material in each cell to be dried is obtained by the placement coefficient prediction model. The predicted placement coefficient is related to the distance between the components, the first average moisture content score, the second average moisture content score, the rate of change of the first moisture content score, the rate of change of the second moisture content score, the length of the common side, the total area, and the hygroscopic coefficient of the material. The length of the distance between the two locations directly affects the magnitude of the predicted placement coefficient. The longer the distance, the weaker the coverage of the fixed drying equipment, and the larger the predicted placement coefficient. Therefore, the distance between the two locations is positively correlated with the predicted placement coefficient. The magnitude of the first average moisture content score and the second average moisture content score directly affects the magnitude of the predicted placement coefficient. The larger the first average moisture content score and the second average moisture content score, the greater the drying requirement of the engineering material and the greater the predicted placement coefficient. Therefore, the first average moisture content score and the second average moisture content score are positively correlated with the predicted placement coefficient. The magnitude of the change rate of the first moisture content score and the change rate of the second moisture content score directly affects the magnitude of the predicted placement coefficient. The larger the change rate of the first moisture content score and the change rate of the second moisture content score, the higher the urgency of the engineering material to be dried, and the larger the predicted placement coefficient. Therefore, the change rate of the first moisture content score and the change rate of the second moisture content score are positively correlated with the predicted placement coefficient. The length of the shared side directly affects the magnitude of the predicted placement coefficient. The longer the shared side, the larger the coverage area of ​​the mobile drying equipment, and the larger the predicted placement coefficient. Therefore, the length of the shared side is positively correlated with the predicted placement coefficient. The size of the total area directly affects the size of the predicted placement coefficient. The larger the total area, the higher the demand for drying, and the larger the predicted placement coefficient. Therefore, the total area and the predicted placement coefficient are positively correlated. The size of the material's hygroscopic coefficient directly affects the size of the predicted placement coefficient. The larger the material's hygroscopic coefficient, the easier the material is to absorb moisture, indicating a higher demand for drying and a larger predicted placement coefficient. Therefore, the material's hygroscopic coefficient and the predicted placement coefficient are positively correlated.

[0028] It should be further explained that, in the specific implementation process, the process of deciding whether to place the mobile drying equipment on the common side based on the placement coefficient includes: Set an appropriate placement coefficient threshold based on historical placement coefficient data, where historical placement coefficient data refers to the data set of past shared edge placement coefficients, and compare the predicted placement coefficient of the shared edge with the placement coefficient threshold. When the predicted placement coefficient of the shared edge is greater than the placement coefficient threshold, a mobile drying device is placed on the shared edge. The mobile drying device reciprocates on the shared edge at a set speed to dry the engineering materials in the cells on both sides of the shared edge.

[0029] It should be further explained that, in the specific implementation process, the process of optimizing the initial drying path based on the mobile drying equipment placed on the shared side to obtain the optimized drying path includes: Check whether the initial drying path passes through the shared edge of the already placed mobile drying equipment. The shared edge of the already placed mobile drying equipment is called the conflict edge. The mobile drying equipment is already working on this edge, and the patrol drying robot passing through this edge will cause redundancy or interference. The initial drying path is divided into segments. Each segment is checked to see if it intersects with a conflict edge. If it intersects, an alternative path is searched in the vicinity of the segment. The shortest path from the current point to the target point is used as the alternative path to avoid passing through conflict edges. The alternative path must still be an edge of the cell to be dried. The path sequence is updated to ensure that all cells to be dried are still visited, and finally the optimized drying path is obtained. A patrol cycle is set, during which the patrol-type drying robot patrols along the optimized drying path and assists in drying the engineering materials in the cells along the path. After each patrol cycle, the drying path of the patrol-type drying robot is regenerated and optimized based on the real-time moisture content score and the predicted placement coefficient.

[0030] An artificial intelligence-based method for detecting the moisture content of engineering materials includes the following steps: S1: Collect the moisture content data of engineering materials in each cell of the warehouse in real time through the detection equipment, and perform normalization preprocessing on the moisture content data; S2: Based on the pre-processed moisture content data, evaluate and calculate the real-time moisture content score of the engineering materials in each cell, periodically evaluate the moisture content status of the engineering materials, and calculate the rate of change of moisture content score through time series analysis. S3: Based on the moisture content score and the rate of change of the moisture content score, a three-level classification mechanism is used to classify the drying needs of cells, and cells containing engineering materials with high moisture content are marked as cells to be dried, forming a set of areas to be processed, which provides input for subsequent path planning; S4: Based on the spatial distribution of the cells to be dried, generate the initial drying path of the patrol drying robot from the warehouse entrance to the exit. Evaluate the deployment requirements of mobile drying equipment on each common side through the placement coefficient prediction model, and predict the placement coefficient of mobile drying equipment on the common side. S5: Based on the predicted placement coefficient, place the mobile drying equipment on the corresponding shared edge, detect the conflict between the initial drying path and the shared edge where the mobile drying equipment is placed, identify the conflicting edges that need to be avoided, and obtain the optimized drying path; S6: The patrol-type drying robot performs tasks according to the optimized drying path, monitors the drying effect in real time, and dynamically adjusts the drying strategy and path planning based on feedback data.

[0031] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

[0032] In conclusion, the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. An artificial intelligence-based moisture content detection system for engineering materials, characterized in that, Includes the following modules: The data acquisition module is used to collect moisture content data of engineering materials in the warehouse; The moisture content assessment module is used to generate real-time moisture content scores for engineering materials based on the collected moisture content data, and to derive the rate of change of moisture content scores and the average moisture content score of engineering materials based on multiple moisture content scores. The drying path planning module is used to classify engineering materials according to the rate of change of moisture content score and moisture content score to obtain cells to be dried, and to plan the initial drying path of the patrol drying robot based on all cells to be dried. The placement coefficient prediction module is used to acquire historical placement coefficient data and build a placement coefficient prediction model. The placement coefficient prediction model is used to predict the placement coefficient of the common side of the cells to be dried, and the decision is made on whether to place the mobile drying equipment on the common side based on the placement coefficient. The drying path optimization module is used to optimize the initial drying path based on the mobile drying equipment placed on the shared side to obtain the optimized drying path.

2. The artificial intelligence-based moisture content detection system for engineering materials according to claim 1, characterized in that, The engineering materials are stored in a warehouse equipped with dehumidifiers. The warehouse area is divided into multiple cells, and the engineering materials are stacked in the cells. The edges of the cells form walkways. Each cell is equipped with drying equipment and testing equipment. The drying equipment is a fixed drying equipment used to reduce the moisture content of the engineering materials, and the testing equipment is used to collect moisture content data, including ambient humidity, material type coefficient, ambient temperature, air velocity, and material bulk density. Ambient humidity is obtained through a temperature and humidity sensor; Material type coefficients are obtained from a materials database; Ambient temperature is obtained through a temperature sensor; Air velocity is obtained through a wind speed sensor; The weight of the engineering material is obtained by using a weight sensor, and the bulk density of the material is obtained by dividing the weight by the volume of the engineering material.

3. The artificial intelligence-based moisture content detection system for engineering materials according to claim 2, characterized in that, The process of generating a real-time moisture content score for engineering materials based on the collected moisture content data includes: Moisture content score S: ； In the formula, , , and These are weighting coefficients, derived from historical data training. The measured ambient humidity At maximum humidity, For material type coefficient, The actual ambient temperature. This is the upper limit of temperature. This is the measured airflow velocity. This is the upper limit of the flow rate. This represents the measured bulk density of the material. This represents the maximum density.

4. The artificial intelligence-based moisture content detection system for engineering materials according to claim 3, characterized in that, The process of deriving the rate of change of moisture content score and the average moisture content score of engineering materials based on multiple moisture content scores includes: Set an evaluation cycle. At the end of each evaluation cycle, the real-time moisture content score of the engineering materials in each cell needs to be obtained. The average moisture content score of the engineering materials in the cell is obtained based on the m moisture content scores. For each evaluation cycle, the moisture content scores of the previous 2n evaluation cycles are taken. The 2n moisture content scores are arranged in chronological order. The average of the first n moisture content scores is calculated to obtain the first average moisture content score. The average of the last n moisture content scores is calculated to obtain the second average moisture content score. The average of the second average moisture content score is subtracted from the average of the first average moisture content score to obtain the score difference. The score difference is divided by the average of the first average moisture content score to obtain the rate of change of moisture content score.

5. The artificial intelligence-based moisture content detection system for engineering materials according to claim 4, characterized in that, The process of grading engineering materials based on the rate of change of moisture content score and moisture content score to obtain cells to be dried includes: Based on historical moisture content score change rate data, set appropriate first change rate threshold and second change rate threshold. The historical moisture content score change rate data refers to the data set of previous engineering material moisture content score change rates. Compare the engineering material moisture content score change rate with the two change rate thresholds. Based on historical moisture content scoring data, set appropriate first and second scoring thresholds. The historical moisture content scoring data refers to the data set of previous engineering material moisture content scores. Compare the engineering material moisture content scores with the two scoring thresholds. If the rate of change of the moisture content score of an engineering material is less than the first rate of change threshold and the moisture content score is less than the first score threshold, the engineering material is judged to be a low moisture content engineering material. If the first rate of change threshold is less than the rate of change of the moisture content score of the engineering material, which is less than the second rate of change threshold, or if the first score threshold is less than the moisture content score of the engineering material, which is less than the second score threshold, then the engineering material is determined to be a medium moisture content engineering material, and the drying equipment in the corresponding cell needs to be activated. If the rate of change of the moisture content score of the engineering material is greater than the second rate of change threshold or the moisture content score of the engineering material is greater than the second score threshold, the engineering material is determined to be a high moisture content engineering material, and the drying equipment in the corresponding cell needs to be activated, and the corresponding cell is marked as a cell to be dried.

6. The artificial intelligence-based moisture content detection system for engineering materials according to claim 5, characterized in that, The process of planning the initial drying path for the patrol-type drying robot based on all cells to be dried includes: The patrol-type drying robot refers to a robot that patrols within a warehouse and assists in the drying of engineering materials; the initial drying path refers to the walking path of the patrol-type drying robot within the warehouse. Starting from the warehouse entrance, which serves as the path starting point, the nearest cells to be dried are iteratively selected and added to the path until all cells to be dried are covered. Each cell can have only one edge added to the path. Finally, the path is connected to the warehouse exit to form a complete path, thus obtaining the initial drying path of the patrol drying robot.

7. The artificial intelligence-based moisture content detection system for engineering materials according to claim 6, characterized in that, The process of acquiring historical placement coefficient data and constructing a placement coefficient prediction model, and then using this model to predict the placement coefficient of the shared edge of the cells to be dried, includes: The shared edge refers to the edge shared by two adjacent cells. The placement coefficient refers to the probability coefficient of placing a mobile drying device on the shared edge. The mobile drying device is used to move back and forth within the shared edge to assist in drying the engineering materials in the cells on both sides. It is necessary to predict the placement coefficient of the shared edge of each cell to be dried. Factors affecting the placement coefficient of shared sides of cells to be dried include: distance between cells, first average moisture content score, second average moisture content score, rate of change of first moisture content score, rate of change of second moisture content score, length of shared side, total area, and material hygroscopicity coefficient. The distance between them refers to the distance from the nearest fixed drying equipment to the shared edge; The first average moisture content score and the rate of change of the first moisture content score refer to the average moisture content score and the rate of change of the moisture content score of one of the cells corresponding to the shared edge; The second average moisture content score and the second moisture content score change rate refer to the average moisture content score and the moisture content score change rate of the other cell corresponding to the shared edge; The length of a shared side refers to the length of the side that shares a common edge. The total area refers to the sum of the areas of the two cells that share a common edge; The moisture absorption coefficient of the engineering materials is obtained by querying the database for the cells on both sides of the shared edge and taking the average of the two moisture absorption coefficients. Set a testing cycle and obtain historical placement coefficient data of engineering materials in each cell to be dried within the testing cycle. The historical placement coefficient data includes the distance between engineering materials in each cell to be dried within the testing cycle, the first average moisture content score, the second average moisture content score, the change rate of the first moisture content score, the change rate of the second moisture content score, the length of the common side, the total area, the material hygroscopicity coefficient, and the historical placement coefficient of engineering materials in each cell to be dried within the testing cycle. Based on the distance between engineering materials in each cell to be dried in different historical placement coefficient data during the detection cycle, the first average moisture content score, the second average moisture content score, the rate of change of the first moisture content score, the rate of change of the second moisture content score, the common side length, the total area, the material hygroscopic coefficient, and the corresponding historical placement coefficient, a placement coefficient prediction set is generated and divided into a training set and a test set. A convolutional neural network is constructed, and the distance between different historical placement coefficients in the training set, the first average moisture content score, the second average moisture content score, the rate of change of the first moisture content score, the rate of change of the second moisture content score, the length of the common side, the total area, and the material hygroscopic coefficient are used as the input data of the convolutional neural network. The corresponding historical placement coefficients in the training set are used as the output data of the convolutional neural network. The convolutional neural network is trained to obtain an initial convolutional neural network. The initial convolutional neural network is validated using a test set. The initial convolutional neural network whose output is less than or equal to the preset test error threshold is used as the placement coefficient prediction model. The distance between the engineering materials in each cell to be dried during the testing period, the first average moisture content score, the second average moisture content score, the rate of change of the first moisture content score, the rate of change of the second moisture content score, the length of the common side, the total area, and the material hygroscopic coefficient are input into the placement coefficient prediction model to obtain the predicted placement coefficient of the engineering materials in each cell to be dried.

8. The artificial intelligence-based moisture content detection system for engineering materials according to claim 7, characterized in that, The process of deciding whether to place a mobile drying device on the shared side based on the placement coefficient includes: Set an appropriate placement coefficient threshold based on historical placement coefficient data, where historical placement coefficient data refers to the data set of past shared edge placement coefficients, and compare the predicted placement coefficient of the shared edge with the placement coefficient threshold. When the predicted placement coefficient of the shared edge is greater than the placement coefficient threshold, a mobile drying device is placed on the shared edge. The mobile drying device reciprocates on the shared edge at a set speed to dry the engineering materials in the cells on both sides of the shared edge.

9. The artificial intelligence-based moisture content detection system for engineering materials according to claim 8, characterized in that, The process of optimizing the initial drying path based on the mobile drying equipment placed along the shared side to obtain the optimized drying path includes: Check whether the initial drying path passes through the shared edge of the already placed mobile drying equipment. The shared edge of the already placed mobile drying equipment is called the conflict edge. The mobile drying equipment is already working on this edge, and the patrol drying robot passing through this edge will cause redundancy or interference. The initial drying path is divided into segments. Each segment is checked to see if it intersects with a conflict edge. If it intersects, an alternative path is searched in the vicinity of the segment. The shortest path from the current point to the target point is used as the alternative path to avoid passing through conflict edges. The alternative path must still be an edge of the cell to be dried. The path sequence is updated to ensure that all cells to be dried are still visited, and finally the optimized drying path is obtained. A patrol cycle is set, during which the patrol-type drying robot patrols along the optimized drying path and assists in drying the engineering materials in the cells along the path. After each patrol cycle, the drying path of the patrol-type drying robot is regenerated and optimized based on the real-time moisture content score and the predicted placement coefficient.

10. A method for detecting the moisture content of engineering materials based on artificial intelligence, implemented based on the artificial intelligence-based moisture content detection system for engineering materials according to any one of claims 1-9, characterized in that, Includes the following steps: S1: Collect the moisture content data of engineering materials in each cell of the warehouse in real time through the detection equipment, and perform normalization preprocessing on the moisture content data; S2: Based on the pre-processed moisture content data, evaluate and calculate the real-time moisture content score of the engineering materials in each cell, periodically evaluate the moisture content status of the engineering materials, and calculate the rate of change of moisture content score through time series analysis. S3: Based on the moisture content score and the rate of change of the moisture content score, a three-level classification mechanism is used to classify the drying needs of cells, and cells containing engineering materials with high moisture content are marked as cells to be dried, forming a set of areas to be processed, which provides input for subsequent path planning; S4: Based on the spatial distribution of the cells to be dried, generate the initial drying path of the patrol drying robot from the warehouse entrance to the exit. Evaluate the deployment requirements of mobile drying equipment on each common side through the placement coefficient prediction model, and predict the placement coefficient of mobile drying equipment on the common side. S5: Based on the predicted placement coefficient, place the mobile drying equipment on the corresponding shared edge, detect the conflict between the initial drying path and the shared edge where the mobile drying equipment is placed, identify the conflicting edges that need to be avoided, and obtain the optimized drying path; S6: The patrol-type drying robot performs tasks according to the optimized drying path, monitors the drying effect in real time, and dynamically adjusts the drying strategy and path planning based on feedback data.

Images