Machine-made sand moisture content on-line detection system

Abstract

The invention relates to the technical field of machine-made sand processing, in particular to a machine-made sand moisture content on-line detection system, which comprises an intelligent sensor acquisition module used for performing real-time acquisition on the moisture content of machine-made sand through an intelligent sensor; the data calibration module is used for performing data calibration on the real-time acquisition value; the data preprocessing module is used for preprocessing the calibrated real-time acquisition value; the detection module is used for detecting the moisture content of the machine-made sand in real time according to the moisture content of the reference machine-made sand and the collected value after pretreatment; the detection optimization module is used for determining a detection optimization coefficient according to the initial real-time detection result and optimizing the initial real-time detection result according to the detection optimization coefficient; and the real-time regulation and control module is used for regulating and controlling the machine-made sand equipment in real time according to the final real-time detection result. According to the invention, the deviation of detection data can be reduced, and the machine-made sand equipment can be regulated and controlled in real time according to an accurate detection result.

Description

Technical Field

[0001] The present invention relates to the technical field of machine-made sand processing, and in particular to an online detection system for the moisture content of machine-made sand. Background Art

[0002] In the production and application of manufactured sand, accurate moisture content testing is crucial. Excessively high moisture content in manufactured sand will increase the concrete's water demand, reduce its strength, and easily cause segregation and bleeding; too low a moisture content will result in poor concrete fluidity and workability. Currently, the testing of manufactured sand moisture content has formed a process from sampling to data analysis, but the problems of error and fluctuation in test data are prominent. During sampling, the moisture content of manufactured sand varies greatly between batches and locations during production and stacking due to factors such as changes in ambient humidity and collisions during material handling. If the sampling points are not distributed properly and the sampling volume does not meet the standards, the data will be poorly representative. During testing, the drying method is affected by the equipment's heating efficiency and the accuracy of the temperature sensor, and the rapid detector is easily affected by the color and surface roughness of the manufactured sand, resulting in significant data deviations. In the data processing stage, due to large errors in the original data, a reliable quantitative basis cannot be formed, resulting in the inability to adjust the production equipment parameters in real time.

[0003] Chinese Patent Publication No. CN113457837A discloses a method for flushing and online monitoring of coarse aggregate median diameter in a manufactured sand and gravel production plant. The method includes the following steps: presetting the weight percentage and allowable deviation range for each coarse aggregate particle size in an intelligent control center; conveying the coarse aggregate via a feed metering conveyor to a flushing test screen for particle size screening and grading; conveying each particle size aggregate via separate metering conveyors to a finished product conveyor for mixing and storage; and collecting real-time weighing data from the feed metering conveyor and the screened aggregate conveyors. The intelligent control center calculates the weight percentage of each particle size coarse aggregate and compares it with the preset value. If the weight percentage exceeds the allowable deviation range, the intelligent control center issues an alert and notifies the operator to take corrective action. This solution, when testing the moisture content of manufactured sand, suffers from errors and fluctuations in the acquired test data, making it difficult to generate reliable and accurate quantitative data during data processing. Consequently, accurate and effective test results cannot be used to control the manufactured sand production equipment in real time. Summary of the Invention

[0004] To this end, the present invention provides an online detection system for the moisture content of manufactured sand, which is used to overcome the problem in the prior art that in the process of implementing detection operations on the moisture content of manufactured sand, the obtained detection data have errors and fluctuation deviations, making it difficult to form a reliable and accurate quantitative basis in the data processing link, and thus making it impossible to rely on accurate and effective detection results to perform real-time regulation of the manufactured sand production equipment.

[0005] To achieve the above object, the present invention provides an online detection system for the moisture content of manufactured sand, the system comprising:

[0006] The intelligent sensor acquisition module is used to collect the moisture content of the manufactured sand in real time through the intelligent sensor to obtain the real-time collection value;

[0007] A data calibration module, configured to perform data calibration on the real-time collected values to obtain calibrated real-time collected values;

[0008] A data preprocessing module, configured to preprocess the calibrated real-time collected values to obtain preprocessed collected values;

[0009] a detection module, configured to determine a benchmark water content of manufactured sand that matches the value collected after pretreatment, and to perform real-time detection of the water content of manufactured sand based on the benchmark water content of manufactured sand and the value collected after pretreatment, to obtain an initial real-time detection result;

[0010] A detection optimization module is used to determine a detection optimization coefficient based on the initial real-time detection result, and optimize the initial real-time detection result based on the detection optimization coefficient to obtain a final real-time detection result;

[0011] The real-time control module is used to control the machine-made sand equipment in real time according to the final real-time detection results.

[0012] Furthermore, the data calibration module performs real-time fitting on the real-time collected values to obtain real-time fitting results, determines calibration coefficients based on the real-time fitting results, and calibrates the real-time collected values based on the calibration coefficients to obtain calibrated real-time collected values.

[0013] Further, the data preprocessing module calculates a first preset threshold R1 and a second preset threshold R2 according to the mean μ and standard deviation σ of the calibrated real-time collected value. The mathematical expression of the first preset threshold R1 is: R1=μ-3σ, and the mathematical expression of the second preset threshold R2 is: R2=μ+3σ. R1<R2 is set, and the calibrated real-time collected value R0 is compared with the first preset threshold R1 and the second preset threshold R2. The data state type of the calibrated real-time collected value is judged according to the comparison result, and the calibrated real-time collected value is processed according to the judgment result to obtain an abnormal value processing result, wherein:

[0014] When R1≤R0≤R2, the data preprocessing module determines that the data state type of the calibrated real-time collected value is normal, and the data preprocessing module does not perform abnormal value processing on the calibrated real-time collected value;

[0015] When R0<R1, the data preprocessing module determines that the data state type of the calibrated real-time collected value is an abnormal situation, and eliminates the calibrated real-time collected value, and records the appearance time and position of the calibrated real-time collected value, obtains a first abnormal value processing result, and outputs the first abnormal value processing result as the abnormal value processing result;

[0016] When R0>R2, the data preprocessing module determines that the data state type of the calibrated real-time collected value is an abnormal situation, and removes the calibrated real-time collected value, and records the appearance time and position of the calibrated real-time collected value, obtains a second abnormal value processing result, and outputs the second abnormal value processing result as the abnormal value processing result;

[0017] The data preprocessing module performs smoothing processing on the abnormal value processing result to obtain a preprocessed collection value.

[0018] Furthermore, the detection module includes:

[0019] A reference machine-made sand moisture content determination unit is used to identify the pre-processed collected value according to the reference machine-made sand moisture content identification model to obtain a reference machine-made sand moisture content that matches the pre-processed collected value;

[0020] The real-time detection unit is used to determine the water content deviation value of the manufactured sand according to the benchmark water content of the manufactured sand and the pre-processed collection value, and to perform real-time detection on the water content of the manufactured sand according to the water content deviation value of the manufactured sand to obtain an initial real-time detection result.

[0021] Furthermore, the benchmark machine-made sand moisture content determination unit obtains historical benchmark machine-made sand moisture content data, and performs feature extraction on the historical benchmark machine-made sand moisture content data to obtain a key benchmark machine-made sand moisture content data set, and trains a convolutional neural network model based on the key benchmark machine-made sand moisture content data set, and outputs the convolutional neural network model that meets the preset accuracy as a benchmark machine-made sand moisture content identification model, and uses the benchmark machine-made sand moisture content identification model to identify the preprocessed collected value to obtain a benchmark machine-made sand moisture content that matches the preprocessed collected value.

[0022] Furthermore, the real-time detection unit calculates a deviation value W0 of the water content of the manufactured sand according to the reference water content of the manufactured sand JZ and the pre-processed collected value YC. The mathematical expression of the deviation value W0 of the water content of the manufactured sand is: W0=|YC-JZ|. The deviation value W0 of the water content of the manufactured sand is compared with the third preset threshold value W1 and the fourth preset threshold value W2. The water content of the manufactured sand is detected in real time according to the comparison result to obtain an initial real-time detection result, wherein:

[0023] When W0<W1, the real-time detection unit determines that the water content of the manufactured sand is in an excessively low water content state, converts the excessively low water content state into a first control signal, and outputs the first control signal as an initial real-time detection result;

[0024] When W1≤W0≤W2, the real-time detection unit determines that the water content of the manufactured sand is in a normal state, converts the normal state of the water content of the manufactured sand into a second control signal, and outputs the second control signal as the initial real-time detection result;

[0025] When W0>W2, the real-time detection unit determines that the water content of the manufactured sand is in an excessively high water content state, converts the excessively high water content state into a third control signal, and outputs the third control signal as the initial real-time detection result.

[0026] Furthermore, the detection optimization module includes:

[0027] A detection optimization coefficient determination unit, configured to determine a detection optimization coefficient based on an initial real-time detection result;

[0028] The detection optimization unit is used to optimize the initial real-time detection result according to the detection optimization coefficient to obtain the final real-time detection result.

[0029] Furthermore, the detection optimization coefficient determination unit is used to obtain the moisture content M of each sample according to the initial real-time detection result. i , ambient temperature T and air humidity H, and according to the moisture content M of each sample i The mass fluctuation rate V of manufactured sand is calculated. The mathematical expression of the mass fluctuation rate V of manufactured sand is: The environmental interference coefficient E is calculated according to the ambient temperature T and the air humidity H. The mathematical expression of the environmental interference coefficient E is: E=1+0.01×(T-20)+0.05×(H-50). The detection optimization coefficient K is calculated according to the quality fluctuation rate V of the manufactured sand and the environmental interference coefficient E. The mathematical expression of the detection optimization coefficient K is: K=α×V-β×E, wherein n represents the total number of samples, JZ represents the moisture content of the benchmark manufactured sand, α represents the weight coefficient for adjusting the quality fluctuation rate V of the manufactured sand, β represents the weight coefficient for adjusting the environmental interference coefficient E, and α+β=1.

[0030] Furthermore, the detection optimization unit compares the detection optimization coefficient K with a preset detection optimization coefficient K0, performs optimization judgment on the initial real-time detection result according to the comparison result, and optimizes the initial real-time detection result according to the judgment result to obtain a final real-time detection result, wherein:

[0031] When K>K0, the detection optimization unit determines to optimize the initial real-time detection result, and optimizes the initial real-time detection result according to the preset detection optimization strategy corresponding to the preset detection optimization coefficient to obtain a detection optimization result, and outputs the detection optimization result as the final real-time detection result;

[0032] When K≤K0, the detection optimization unit determines not to optimize the initial real-time detection result, and outputs the initial real-time detection result as the final real-time detection result.

[0033] Furthermore, when the final real-time detection result is a detection optimization result, the real-time control module performs real-time control on the machine-made sand equipment according to the first control means;

[0034] When the final real-time detection result is the first control signal, the real-time control module controls the machine-made sand equipment in real time according to the second control means;

[0035] When the final real-time detection result is the second control signal, the real-time control module performs real-time control on the machine-made sand equipment according to the third control means;

[0036] When the final real-time detection result is the third control signal, the real-time control module performs real-time control on the machine-made sand equipment according to the fourth control means.

[0037] Compared with the prior art, the principles and advantages of the present application lie in that the intelligent sensor acquisition module uses an intelligent sensor to collect the moisture content of the manufactured sand in real time, obtains the real-time acquisition value, and provides original data for subsequent detection. Then the data calibration module calibrates the real-time acquisition value, corrects the possible acquisition error, obtains the calibrated real-time acquisition value, and improves the accuracy of the data. Then the data preprocessing module preprocesses the calibrated real-time acquisition value to obtain a more reliable preprocessed acquisition value. The detection module determines the benchmark manufactured sand moisture content that matches the preprocessed acquisition value based on the preprocessed acquisition value, and performs real-time detection of the moisture content of the manufactured sand by comparing the two to obtain the initial real-time detection result and complete the preliminary moisture content judgment. The detection optimization module determines the detection optimization coefficient based on the initial real-time detection result, optimizes the initial result, further improves the detection accuracy, and obtains the final real-time detection result. Finally, the real-time control module controls the manufactured sand equipment in real time based on the final real-time detection result so that the moisture content of the manufactured sand meets the requirements.

[0038] Compared with the existing technology, in this application, the intelligent sensor acquisition module uses intelligent sensors to collect the moisture content of machine-made sand in real time, realize high-frequency dynamic acquisition of moisture content data, avoid the intermittent lag and subjective errors of manual detection, and provide real-time and continuous raw data input for subsequent processing, thereby ensuring the timeliness of data collection from the source and preventing distortion of detection benchmarks caused by untimely or intermittent data collection.

[0039] The real-time collected values are calibrated through the data calibration module to eliminate data deviations caused by factors such as sensor errors and environmental interference, so that the real-time collected values after calibration are closer to the actual value of the moisture content of the manufactured sand.

[0040] The calibrated data is preprocessed through the data preprocessing module to filter out random noise and invalid anomalies mixed in during data transmission or storage. At the same time, the data is converted into a standardized format that can be directly recognized by the detection algorithm to avoid interference of unstructured data on the detection model and ensure that the input data received by the detection module is highly reliable.

[0041] The detection module determines the benchmark water content of manufactured sand that matches the value collected after pretreatment, providing a clear reference standard for real-time detection, making the detection process more targeted and accurate, and ensuring that the test results can truly reflect the difference between the water content of manufactured sand and the benchmark.

[0042] The detection optimization module determines the detection optimization coefficient based on the initial real-time detection results, and uses the detection optimization coefficient to optimize the initial results to eliminate possible systematic errors and random errors in the detection process, making the final real-time detection results more accurate and closer to the actual moisture content of the manufactured sand.

[0043] By using the real-time control module to control the machine-made sand equipment in real time according to the final real-time detection results, various parameters in the machine-made sand production process can be accurately adapted to the actual moisture content, ensuring that the production process is always in the best state and improving product quality stability. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] Figure 1 This is a schematic structural diagram of an online detection system for moisture content in manufactured sand according to an embodiment of the present invention;

[0045] Figure 2 Schematic diagram of the structure of the detection module according to an embodiment of the present invention;

[0046] Figure 3 Schematic diagram of the structure of the detection optimization module according to an embodiment of the present invention. DETAILED DESCRIPTION

[0047] The following is further described in detail through specific implementation methods:

[0048] See also Figure 1 As shown in FIG, which is a schematic structural diagram of an online detection system for moisture content of manufactured sand according to an embodiment of the present invention, the system includes:

[0049] The intelligent sensor acquisition module is used to collect the moisture content of the manufactured sand in real time through the intelligent sensor to obtain the real-time collection value;

[0050] A data calibration module, configured to perform data calibration on the real-time collected values to obtain calibrated real-time collected values;

[0051] A data preprocessing module, configured to preprocess the calibrated real-time collected values to obtain preprocessed collected values;

[0052] a detection module, configured to determine a benchmark water content of manufactured sand that matches the value collected after pretreatment, and to perform real-time detection of the water content of manufactured sand based on the benchmark water content of manufactured sand and the value collected after pretreatment, to obtain an initial real-time detection result;

[0053] A detection optimization module is used to determine a detection optimization coefficient based on the initial real-time detection result, and optimize the initial real-time detection result based on the detection optimization coefficient to obtain a final real-time detection result;

[0054] The real-time control module is used to control the machine-made sand equipment in real time according to the final real-time detection results.

[0055] Specifically, the system is applied to the machine-made sand preparation and detection terminal. The system uses the intelligent sensor acquisition module to collect the moisture content of the machine-made sand in real time using the intelligent sensor, obtains the real-time acquisition value, and provides the original data for subsequent detection. Then the data calibration module calibrates the real-time acquisition value, corrects the possible acquisition error, obtains the calibrated real-time acquisition value, and improves the accuracy of the data. Then the data preprocessing module preprocesses the calibrated real-time acquisition value to obtain a more reliable preprocessed acquisition value. The detection module determines the matching benchmark machine-made sand moisture content based on the preprocessed acquisition value, and performs real-time detection of the moisture content of the machine-made sand by comparing the two, obtains the initial real-time detection result, and completes the preliminary moisture content judgment. The detection optimization module determines the detection optimization coefficient according to the initial real-time detection result, optimizes the initial result, further improves the detection accuracy, and obtains the final real-time detection result. Finally, the real-time control module controls the machine-made sand equipment in real time according to the final real-time detection result, so that the moisture content of the machine-made sand meets the requirements.

[0056] Compared with the existing technology, in this application, the intelligent sensor acquisition module uses intelligent sensors to collect the moisture content of machine-made sand in real time, realize high-frequency dynamic acquisition of moisture content data, avoid the intermittent lag and subjective errors of manual detection, and provide real-time and continuous raw data input for subsequent processing, thereby ensuring the timeliness of data collection from the source and preventing distortion of detection benchmarks caused by untimely or intermittent data collection.

[0057] The real-time collected values are calibrated through the data calibration module to eliminate data deviations caused by factors such as sensor errors and environmental interference, so that the real-time collected values after calibration are closer to the actual value of the moisture content of the manufactured sand.

[0058] The calibrated data is preprocessed through the data preprocessing module to filter out random noise and invalid anomalies mixed in during data transmission or storage. At the same time, the data is converted into a standardized format that can be directly recognized by the detection algorithm to avoid interference of unstructured data on the detection model and ensure that the input data received by the detection module is highly reliable.

[0059] The detection module determines the benchmark water content of manufactured sand that matches the value collected after pretreatment, providing a clear reference standard for real-time detection, making the detection process more targeted and accurate, and ensuring that the test results can truly reflect the difference between the water content of manufactured sand and the benchmark.

[0060] The detection optimization module determines the detection optimization coefficient based on the initial real-time detection results, and uses the detection optimization coefficient to optimize the initial results to eliminate possible systematic errors and random errors in the detection process, making the final real-time detection results more accurate and closer to the actual moisture content of the manufactured sand.

[0061] By using the real-time control module to control the machine-made sand equipment in real time according to the final real-time detection results, various parameters in the machine-made sand production process can be accurately adapted to the actual moisture content, ensuring that the production process is always in the best state and improving product quality stability.

[0062] Specifically, the intelligent sensor acquisition module determines the collection points of the intelligent sensor according to the actual collection requirements of the moisture content of the machine-made sand, obtains a collection point set, and uses the intelligent sensor to collect the moisture content of the machine-made sand at each collection point in the collection point set in real time to obtain a real-time collection value.

[0063] Specifically, the actual machine-made sand moisture content collection requirement refers to the preset requirements in the process of machine-made sand production, processing, storage and transportation, in order to ensure product quality and control the production process. For example, in the machine-made sand production line, in order to control the feed humidity of the crusher, it is necessary to set a moisture content collection point in front of the crusher. The collection point set refers to the set of specific location points for collecting the moisture content of machine-made sand determined according to the actual machine-made sand moisture content collection requirement in the process of machine-made sand production, processing, storage or transportation. The moisture content of the machine-made sand at each collection point refers to the ratio of the water mass in the machine-made sand to the total mass of the machine-made sand obtained by real-time collection of each collection point in the collection set by the intelligent sensor. The real-time collection value refers to the value obtained by real-time collection of the collection point by the intelligent sensor. The moisture content data of manufactured sand, for example, at the moisture content collection point in front of the crusher, the real-time collection value is 8.5%, indicating that the moisture content of the currently fed manufactured sand is 8.5%. In the manufactured sand storage warehouse, the real-time collection value is 12.0%, indicating that the moisture content of the currently stored manufactured sand is 12.0%. This embodiment does not limit the specific deployment scheme of the intelligent sensor. Those skilled in the art can freely set it according to actual conditions, and only need to meet the needs of intelligent data collection. For example, it can be set to determine the collection point position in key links such as the finished product warehouse, conveyor belt, and vibrating feeder according to the actual requirements of the collection of moisture content of manufactured sand. At the selected collection point position, use a fixed bracket to firmly install the sensor to ensure that the sensor collection surface (ceramic collection surface) is facing the direction of material flow and maintains an appropriate distance from the material being measured.

[0064] Specifically, the intelligent sensor acquisition module determines the collection points according to the actual machine-made sand moisture content collection requirements, and can collect the moisture content of the machine-made sand at each collection point in real time to ensure the accuracy of the collection.

[0065] Specifically, the data calibration module performs real-time fitting on the real-time collected values to obtain real-time fitting results, determines calibration coefficients based on the real-time fitting results, and calibrates the real-time collected values based on the calibration coefficients to obtain calibrated real-time collected values.

[0066] Specifically, this embodiment does not limit the specific implementation method of real-time fitting of the real-time collected values. For example, it can be set to construct a historical linear regression fitting model based on historical machine-made sand moisture content data, and input the real-time collected values into the historical linear regression fitting model to calculate the actual mean square error, and output the historical linear regression fitting model whose actual mean square error meets the preset mean square error as the actual linear regression fitting model, and perform real-time fitting of the real-time collected values according to the actual linear regression fitting model. The preset mean square error can be set to 5%. The real-time fitting result refers to the result obtained by real-time fitting of the real-time collected values according to the data fitting model. The calibration coefficient refers to the parameter used to adjust the real-time collected values. This embodiment does not limit the specific calculation method of the calibration coefficient. For example, it can be set to the calibration coefficient Where n represents the total number of samples of the real-time collected value, x i represents the i-th real-time acquisition value, y i represents the i-th real-time fitting result, represents the mean of the real-time collected values, Indicates the mean of the real-time fitting results.

[0067] Specifically, the data calibration module calculates the calibration coefficient by fitting the real-time collected values in real time, and calibrates the real-time collected values to improve the accuracy of the data.

[0068] Specifically, the data preprocessing module calculates a first preset threshold R1 and a second preset threshold R2 according to the mean μ and standard deviation σ of the calibrated real-time acquisition value. The mathematical expression of the first preset threshold R1 is: R1=μ-3σ, and the mathematical expression of the second preset threshold R2 is: R2=μ+3σ. R1<R2 is set, and the calibrated real-time acquisition value R0 is compared with the first preset threshold R1 and the second preset threshold R2. The data state type of the calibrated real-time acquisition value is judged according to the comparison result, and the calibrated real-time acquisition value is processed according to the judgment result to obtain an abnormal value processing result, wherein:

[0069] When R1≤R0≤R2, the data preprocessing module determines that the data state type of the calibrated real-time collected value is normal, and the data preprocessing module does not perform abnormal value processing on the calibrated real-time collected value;

[0070] When R0<R1, the data preprocessing module determines that the data state type of the calibrated real-time collected value is an abnormal situation, and eliminates the calibrated real-time collected value, and records the appearance time and position of the calibrated real-time collected value, obtains a first abnormal value processing result, and outputs the first abnormal value processing result as the abnormal value processing result;

[0071] When R0>R2, the data preprocessing module determines that the data state type of the calibrated real-time collected value is an abnormal situation, and removes the calibrated real-time collected value, and records the appearance time and position of the calibrated real-time collected value, obtains a second abnormal value processing result, and outputs the second abnormal value processing result as the abnormal value processing result;

[0072] The data preprocessing module performs smoothing processing on the abnormal value processing result to obtain a preprocessed collection value.

[0073] Specifically, the first preset threshold R1 refers to the lowest limit value for determining whether the calibrated real-time collected value is within a normal range, for example, 5%. The second preset threshold R2 refers to the highest limit value for determining whether the calibrated real-time collected value is within a normal range, for example, 35%. This embodiment does not limit the implementation method of smoothing the abnormal value processing result. For example, the abnormal value processing result can be set to be smoothed by a moving average method. The abnormal value processing result refers to a result obtained by determining the data status type of the calibrated real-time collected value based on the comparison result of the calibrated real-time collected value R0 with the first preset threshold R1 and the second preset threshold R2, and performing abnormal value processing on the calibrated real-time collected value based on the judgment result. The first abnormal value processing result refers to the result obtained when R0<R1, the data preprocessing module determines that the data status type of the calibrated real-time collected value is abnormal and eliminates the calibrated real-time collected value. The second abnormal value processing result refers to the result obtained when R0>R2, the data preprocessing module determines that the data status type of the calibrated real-time collected value is abnormal and eliminates the calibrated real-time collected value.

[0074] Specifically, the data preprocessing module preprocesses the calibrated real-time collected values to improve data accuracy and reliability.

[0075] See also Figure 2 , which is a schematic diagram of the structure of a detection module according to an embodiment of the present invention, the detection module includes:

[0076] A reference machine-made sand moisture content determination unit is used to identify the pre-processed collected value according to the reference machine-made sand moisture content identification model to obtain a reference machine-made sand moisture content that matches the pre-processed collected value;

[0077] The real-time detection unit is used to determine the water content deviation value of the manufactured sand according to the benchmark water content of the manufactured sand and the pre-processed collection value, and to perform real-time detection on the water content of the manufactured sand according to the water content deviation value of the manufactured sand to obtain an initial real-time detection result.

[0078] Specifically, the benchmark machine-made sand moisture content determination unit obtains historical benchmark machine-made sand moisture content data, and performs feature extraction on the historical benchmark machine-made sand moisture content data to obtain a key benchmark machine-made sand moisture content data set. The convolutional neural network model is trained according to the key benchmark machine-made sand moisture content data set, and the convolutional neural network model that meets the preset accuracy is output as a benchmark machine-made sand moisture content identification model. The pre-processed collected value is identified by the benchmark machine-made sand moisture content identification model to obtain a benchmark machine-made sand moisture content that matches the pre-processed collected value.

[0079] Specifically, the historical benchmark machine-made sand moisture content data refers to the target benchmark value set by measuring the moisture content of machine-made sand within a preset time period in the past. This embodiment does not specifically limit the preset time period, such as setting the preset time period to 7 days. This embodiment does not limit the specific implementation method of feature extraction of the historical benchmark machine-made sand moisture content data, such as being set to feature extraction of the historical benchmark machine-made sand moisture content data through principal component analysis. The benchmark machine-made sand moisture content identification model refers to a model that meets the preset accuracy rate obtained by training the convolutional neural network model based on the key benchmark machine-made sand moisture content data set. The key benchmark machine-made sand moisture content data set refers to a data set preset for training the convolutional neural network model in the form of key benchmark machine-made sand moisture content data-benchmark machine-made sand moisture content. The convolutional neural network model refers to a machine learning model for extracting features from pre-processed collected values and predicting the benchmark machine-made sand moisture content. This embodiment does not limit the method of training the convolutional neural network model, and those skilled in the art can set it freely, as long as the pre-processed collected values are satisfied. For example, 75% of the key benchmark machine-made sand moisture content data set can be divided into a benchmark machine-made sand moisture content data training set, and 25% can be divided into a benchmark machine-made sand moisture content data test set. The benchmark machine-made sand moisture content data training set is input into the convolutional neural network model for training, and the benchmark machine-made sand moisture content data test set is input into the trained convolutional neural network model. The parameters in the convolutional neural network model are optimized and iteratively until the accuracy of the output result of the benchmark machine-made sand moisture content data test set of the convolutional neural network model reaches a preset accuracy. The convolutional neural network model is then output as the benchmark machine-made sand moisture content identification model. The preset accuracy refers to a preset value reflecting the accuracy of the training status of the convolutional neural network model. This embodiment does not limit the value of the preset accuracy. Relevant technicians in this field can freely set it as long as it meets the requirement of reflecting the training status of the convolutional neural network model. For example, the preset accuracy can be set to 95%. The benchmark machine-made sand moisture content refers to the result obtained by identifying the pre-processed collected value through the benchmark machine-made sand moisture content identification model, which matches the pre-processed collected value.

[0080] Specifically, the benchmark machine-made sand moisture content determination unit obtains historical benchmark machine-made sand moisture content data and performs feature extraction, and trains a convolutional neural network model to obtain a benchmark machine-made sand moisture content identification model, which can accurately identify the pre-processed collected values and match the corresponding benchmark machine-made sand moisture content, thereby improving the accuracy and efficiency of machine-made sand moisture content identification.

[0081] Specifically, the real-time detection unit calculates the machine-made sand moisture content deviation value W0 according to the reference machine-made sand moisture content JZ and the pre-processed collected value YC. The mathematical expression of the machine-made sand moisture content deviation value W0 is: W0=|YC-JZ|, and compares the machine-made sand moisture content deviation value W0 with the third preset threshold value W1 and the fourth preset threshold value W2. According to the comparison result, the moisture content of the machine-made sand is detected in real time to obtain an initial real-time detection result, wherein:

[0082] When W0<W1, the real-time detection unit determines that the water content of the manufactured sand is in an excessively low water content state, converts the excessively low water content state into a first control signal, and outputs the first control signal as an initial real-time detection result;

[0083] When W1≤W0≤W2, the real-time detection unit determines that the water content of the manufactured sand is in a normal state, converts the normal state of the water content of the manufactured sand into a second control signal, and outputs the second control signal as the initial real-time detection result;

[0084] When W0>W2, the real-time detection unit determines that the water content of the manufactured sand is in an excessively high water content state, converts the excessively high water content state into a third control signal, and outputs the third control signal as the initial real-time detection result.

[0085] Specifically, the third preset threshold W1 refers to the lowest limit value for judging whether the moisture content of the manufactured sand is within the normal range, for example, 3%. The fourth preset threshold W2 refers to the highest limit value for judging whether the moisture content of the manufactured sand is within the normal range, for example, 7%. The first control signal refers to the signal obtained by converting the too-low moisture content state of the manufactured sand when the real-time detection unit determines that the moisture content of the manufactured sand is in the too-low moisture content state. The second control signal refers to the signal obtained by converting the normal moisture content state of the manufactured sand when the real-time detection unit determines that the moisture content of the manufactured sand is in the normal moisture content state. The third control signal refers to the signal obtained by converting the too-high moisture content state of the manufactured sand when the real-time detection unit determines that the moisture content of the manufactured sand is in the too-high moisture content state. The initial real-time detection result refers to the result obtained by comparing the moisture content deviation value W0 of the manufactured sand with the third preset threshold W1 and the fourth preset threshold W2, and performing real-time detection of the moisture content of the manufactured sand according to the comparison result.

[0086] Specifically, the real-time detection unit can detect the moisture content of the manufactured sand in real time and accurately and convert it into a corresponding control signal output based on the comparison between the moisture content deviation value of the manufactured sand and the preset threshold.

[0087] See also Figure 3 , which is a schematic diagram of the structure of the detection optimization module according to an embodiment of the present invention, the detection optimization module includes:

[0088] A detection optimization coefficient determination unit, configured to determine a detection optimization coefficient based on an initial real-time detection result;

[0089] The detection optimization unit is used to optimize the initial real-time detection result according to the detection optimization coefficient to obtain the final real-time detection result.

[0090] Specifically, the detection optimization coefficient determination unit is used to obtain the moisture content M of each sample according to the initial real-time detection result. i , ambient temperature T and air humidity H, and according to the moisture content M of each sample i The mass fluctuation rate V of manufactured sand is calculated. The mathematical expression of the mass fluctuation rate V of manufactured sand is: The environmental interference coefficient E is calculated according to the ambient temperature T and the air humidity H. The mathematical expression of the environmental interference coefficient E is: E=1+0.01×(T-20)+0.05×(H-50). The detection optimization coefficient K is calculated according to the quality fluctuation rate V of the manufactured sand and the environmental interference coefficient E. The mathematical expression of the detection optimization coefficient K is: K=α×V-β×E, wherein n represents the total number of samples, JZ represents the moisture content of the benchmark manufactured sand, α represents the weight coefficient for adjusting the quality fluctuation rate V of the manufactured sand, β represents the weight coefficient for adjusting the environmental interference coefficient E, and α+β=1.

[0091] Specifically, the moisture content M of each sample is i It refers to the moisture content data of each sample collected from the machine-made sand production or storage process. The ambient temperature T refers to the real-time temperature of the machine-made sand production or storage environment. The air humidity H refers to the real-time air humidity of the machine-made sand production or storage environment.

[0092] Specifically, the detection optimization coefficient determination unit integrates the moisture content, ambient temperature and air humidity data of each sample, quantifies the quality fluctuation rate of machine-made sand and the environmental interference coefficient, and dynamically adjusts the weight coefficient to determine the detection optimization coefficient.

[0093] Specifically, the detection optimization unit compares the detection optimization coefficient K with the preset detection optimization coefficient K0, optimizes the initial real-time detection result according to the comparison result, and optimizes the initial real-time detection result according to the judgment result to obtain the final real-time detection result, wherein:

[0094] When K>K0, the detection optimization unit determines to optimize the initial real-time detection result, and optimizes the initial real-time detection result according to the preset detection optimization strategy corresponding to the preset detection optimization coefficient to obtain a detection optimization result, and outputs the detection optimization result as the final real-time detection result;

[0095] When K≤K0, the detection optimization unit determines not to optimize the initial real-time detection result, and outputs the initial real-time detection result as the final real-time detection result.

[0096] Specifically, the preset detection optimization coefficient K0 refers to a preset value compared with the detection optimization coefficient K, such as 0.5. The final real-time detection result refers to the result output after processing by the detection optimization unit. The preset detection optimization strategy refers to a pre-set strategy for optimizing the real-time detection result. This example does not limit the specific implementation method of the preset detection optimization strategy. For example, it can be set to establish a regression equation between the moisture content of machine-made sand and the ambient temperature and air humidity through regression analysis. When K>K0, the regression equation is used to optimize the initial real-time detection result. The detection optimization result refers to the result obtained by optimizing the real-time detection result according to the preset detection optimization strategy corresponding to the preset detection optimization coefficient.

[0097] Specifically, by comparing the detection optimization coefficient with the preset value, the detection optimization unit can accurately determine whether the initial real-time detection result needs to be optimized, improve the accuracy of the result, and reasonably choose whether to optimize based on the judgment result, taking into account both detection efficiency and resource utilization, to ensure that the result is reliable and adaptable to different scenarios.

[0098] Specifically, when the final real-time detection result is a detection optimization result, the real-time control module performs real-time control on the machine-made sand equipment according to the first control means;

[0099] When the final real-time detection result is the first control signal, the real-time control module controls the machine-made sand equipment in real time according to the second control means;

[0100] When the final real-time detection result is the second control signal, the real-time control module performs real-time control on the machine-made sand equipment according to the third control means;

[0101] When the final real-time detection result is the third control signal, the real-time control module performs real-time control on the machine-made sand equipment according to the fourth control means.

[0102] Specifically, the first control means refers to the optimization strategy adopted by the real-time control module when the final real-time detection result is the detection optimization result. This example does not limit the specific implementation method of the first control means. For example, it can be set to quickly adjust the moisture content of the machine-made sand to the target range by increasing the drying power or ventilation volume. The machine-made sand equipment refers to the mechanical equipment used to produce machine-made sand. The second control means refers to the optimization strategy adopted by the real-time control module when the final real-time detection result is the detection optimization result. This example does not limit the specific implementation method of the second control means. For example, it can be set to increase the diameter of the atomized water mist particles, increase the amount of water added per unit time, and start the standby The humidification pipe increases the water spray coverage range. The third control means refers to the optimization strategy adopted by the real-time control module when the final real-time detection result is the detection optimization result. This example does not limit the specific implementation method of the third control means. For example, it can be set to maintain the current nozzle design parameters and water spray frequency, and maintain the hot air temperature and wind speed setting values of the drying equipment. The fourth control means refers to the optimization strategy adopted by the real-time control module when the final real-time detection result is the detection optimization result. This example does not limit the specific implementation method of the fourth control means. For example, it can be set to switch to a microporous atomizing nozzle, reduce the amount of water added per unit time, increase the speed of the drying drum, and enhance the heat exchange efficiency.

[0103] Specifically, the real-time control module adopts a hierarchical control strategy to quickly respond to equipment anomalies and process fluctuations, reduce downtime risks, dynamically adjust the parameters of the machine-made sand equipment according to real-time data, balance output, quality and energy consumption, and reduce production costs.

[0104] The above are only embodiments of the present invention. Common knowledge such as the known specific structures and characteristics in the scheme are not described in detail here. Ordinary technicians in the field are aware of all common technical knowledge in the technical field of the invention before the application date or priority date, can obtain all existing technologies in the field, and have the ability to apply conventional experimental means before that date. Ordinary technicians in the field can improve and implement this scheme in combination with their own abilities under the inspiration given by this application. Some typical known structures or known methods should not become obstacles for ordinary technicians in the field to implement this application. It should be pointed out that for those skilled in the art, without departing from the structure of the present invention, several variations and improvements can be made, which should also be regarded as the scope of protection of the present invention. These will not affect the effect of the implementation of the present invention and the practicality of the patent. The scope of protection required by this application shall be based on the content of its claims, and the specific implementation methods and other records in the specification can be used to interpret the content of the claims.

Claims

1. The online detection system for moisture content of manufactured sand is characterized by: The system comprises: The intelligent sensor acquisition module is used to collect the moisture content of the manufactured sand in real time through the intelligent sensor to obtain the real-time collection value; A data calibration module, configured to perform data calibration on the real-time collected values to obtain calibrated real-time collected values; A data preprocessing module, configured to preprocess the calibrated real-time collected values to obtain preprocessed collected values; a detection module, configured to determine a benchmark water content of manufactured sand that matches the value collected after pretreatment, and to perform real-time detection of the water content of manufactured sand based on the benchmark water content of manufactured sand and the value collected after pretreatment, to obtain an initial real-time detection result; A detection optimization module is used to determine a detection optimization coefficient based on the initial real-time detection result, and optimize the initial real-time detection result based on the detection optimization coefficient to obtain a final real-time detection result; The real-time control module is used to control the machine-made sand equipment in real time according to the final real-time detection results.

2. The online detection system for moisture content of manufactured sand according to claim 1 is characterized by: The data calibration module performs real-time fitting on the real-time collected values to obtain real-time fitting results, determines calibration coefficients based on the real-time fitting results, and calibrates the real-time collected values based on the calibration coefficients to obtain calibrated real-time collected values.

3. The online detection system for moisture content of manufactured sand according to claim 1 is characterized in that: The data preprocessing module calculates a first preset threshold R1 and a second preset threshold R2 according to the mean μ and standard deviation σ of the calibrated real-time collected value. The mathematical expression of the first preset threshold R1 is: R1=μ-3σ, and the mathematical expression of the second preset threshold R2 is: R2=μ+3σ. R1<R2 is set, and the calibrated real-time collected value R0 is compared with the first preset threshold R1 and the second preset threshold R2. The data state type of the calibrated real-time collected value is judged according to the comparison result, and the calibrated real-time collected value is processed according to the judgment result to obtain an abnormal value processing result, wherein: When R1≤R0≤R2, the data preprocessing module determines that the data state type of the calibrated real-time collected value is normal, and the data preprocessing module does not perform abnormal value processing on the calibrated real-time collected value; When R0<R1, the data preprocessing module determines that the data state type of the calibrated real-time collected value is an abnormal situation, and eliminates the calibrated real-time collected value, and records the appearance time and position of the calibrated real-time collected value, obtains a first abnormal value processing result, and outputs the first abnormal value processing result as the abnormal value processing result; When R0>R2, the data preprocessing module determines that the data state type of the calibrated real-time collected value is an abnormal situation, and removes the calibrated real-time collected value, and records the appearance time and position of the calibrated real-time collected value, obtains a second abnormal value processing result, and outputs the second abnormal value processing result as the abnormal value processing result; The data preprocessing module performs smoothing processing on the abnormal value processing result to obtain a preprocessed collection value.

4. The online detection system for moisture content of manufactured sand according to claim 1 is characterized in that: The detection module includes: A reference machine-made sand moisture content determination unit is used to identify the pre-processed collected value according to the reference machine-made sand moisture content identification model to obtain a reference machine-made sand moisture content that matches the pre-processed collected value; The real-time detection unit is used to determine the water content deviation value of the manufactured sand according to the benchmark water content of the manufactured sand and the pre-processed collection value, and to perform real-time detection on the water content of the manufactured sand according to the water content deviation value of the manufactured sand to obtain an initial real-time detection result.

5. The online detection system for moisture content of manufactured sand according to claim 4 is characterized in that: The benchmark machine-made sand moisture content determination unit obtains historical benchmark machine-made sand moisture content data, and performs feature extraction on the historical benchmark machine-made sand moisture content data to obtain a key benchmark machine-made sand moisture content data set. A convolutional neural network model is trained based on the key benchmark machine-made sand moisture content data set, and the convolutional neural network model that meets a preset accuracy rate is output as a benchmark machine-made sand moisture content recognition model. The pre-processed collected value is identified by the benchmark machine-made sand moisture content recognition model to obtain a benchmark machine-made sand moisture content that matches the pre-processed collected value.

6. The online detection system for moisture content of manufactured sand according to claim 5 is characterized in that: The real-time detection unit calculates a deviation value W0 of the water content of the manufactured sand according to the reference water content of the manufactured sand JZ and the pre-processed collected value YC. The mathematical expression of the deviation value W0 of the water content of the manufactured sand is: W0=|YC-JZ|. The deviation value W0 of the water content of the manufactured sand is compared with the third preset threshold value W1 and the fourth preset threshold value W2. The water content of the manufactured sand is detected in real time according to the comparison result to obtain an initial real-time detection result, wherein: When W0<W1, the real-time detection unit determines that the water content of the manufactured sand is in an excessively low water content state, converts the excessively low water content state into a first control signal, and outputs the first control signal as an initial real-time detection result; When W1≤W0≤W2, the real-time detection unit determines that the water content of the manufactured sand is in a normal state, converts the normal state of the water content of the manufactured sand into a second control signal, and outputs the second control signal as the initial real-time detection result; When W0>W2, the real-time detection unit determines that the water content of the manufactured sand is in an excessively high water content state, converts the excessively high water content state into a third control signal, and outputs the third control signal as the initial real-time detection result.

7. The online detection system for moisture content of manufactured sand according to claim 6 is characterized in that: The detection optimization module includes: A detection optimization coefficient determination unit, configured to determine a detection optimization coefficient based on an initial real-time detection result; The detection optimization unit is used to optimize the initial real-time detection result according to the detection optimization coefficient to obtain the final real-time detection result.

8. The online detection system for moisture content of manufactured sand according to claim 7 is characterized in that: The detection optimization coefficient determination unit is used to obtain the moisture content M of each sample according to the initial real-time detection result. i , ambient temperature T and air humidity H, and according to the moisture content M of each sample i The mass fluctuation rate V of manufactured sand is calculated. The mathematical expression of the mass fluctuation rate V of manufactured sand is: The environmental interference coefficient E is calculated according to the ambient temperature T and the air humidity H. The mathematical expression of the environmental interference coefficient E is: E=1+0.01×(T-20)+0.05×(H-50). The detection optimization coefficient K is calculated according to the quality fluctuation rate V of the manufactured sand and the environmental interference coefficient E. The mathematical expression of the detection optimization coefficient K is: K=α×V-β×E, wherein n represents the total number of samples, JZ represents the moisture content of the benchmark manufactured sand, α represents the weight coefficient for adjusting the quality fluctuation rate V of the manufactured sand, β represents the weight coefficient for adjusting the environmental interference coefficient E, and α+β=1.

9. The online detection system for moisture content of manufactured sand according to claim 8, characterized in that: The detection optimization unit compares the detection optimization coefficient K with the preset detection optimization coefficient K0, performs optimization judgment on the initial real-time detection result according to the comparison result, and optimizes the initial real-time detection result according to the judgment result to obtain a final real-time detection result, wherein: When K>K0, the detection optimization unit determines to optimize the initial real-time detection result, and optimizes the initial real-time detection result according to the preset detection optimization strategy corresponding to the preset detection optimization coefficient to obtain a detection optimization result, and outputs the detection optimization result as the final real-time detection result; When K≤K0, the detection optimization unit determines not to optimize the initial real-time detection result, and outputs the initial real-time detection result as the final real-time detection result.

10. The online detection system for moisture content of manufactured sand according to claim 9, characterized in that: When the final real-time detection result is a detection optimization result, the real-time control module performs real-time control on the machine-made sand equipment according to the first control means; When the final real-time detection result is the first control signal, the real-time control module controls the machine-made sand equipment in real time according to the second control means; When the final real-time detection result is the second control signal, the real-time control module performs real-time control on the machine-made sand equipment according to the third control means; When the final real-time detection result is the third control signal, the real-time control module performs real-time control on the machine-made sand equipment according to the fourth control means.

Images