An in-situ detection device and method for the inherent content of multiple components

Abstract

This invention discloses an in-situ detection device and method for the intrinsic content of multiple components, belonging to the field of metal material detection and analysis technology. The implementation method of this invention is as follows: 1. Setting sample measurement indicators; 2. Measuring sample measurement indicators using standard chemical methods; 3. Using a robotic arm to pick up the sample to be tested, and acquiring the infrared spectrum of the sample using an infrared signal acquisition spectrometer to form raw spectral data; 4. Adaptively preprocessing the raw spectral data using a data processing module that removes abnormal spectra and eliminates baseline drift and scattering to form preprocessed spectral data; 5. Constructing a multi-indicator joint quantitative correction model using partial least squares and variable selection methods, and training the model using a correction set and sample measurement indicators; inputting the validation set into the trained model to obtain the detection accuracy of the intrinsic content of multiple components; 6. Outputting the detection results of the intrinsic content of multiple components; Compared with the prior art, this invention achieves simultaneous detection of the intrinsic content of multiple components.

Description

Technical Field

[0001] This invention relates to an in-situ detection device and method for the inherent content of multiple components, belonging to the field of metal material detection and analysis technology. It is used to detect 12 indicators of alumina (Al2O3) and SiO2, Fe2O3, Na2O, CaO, loss on ignition, Li2O, angle of repose, loose density, particle size, tonnage, and K2O components. Background Technology

[0002] Alumina (Al2O3) is an important industrial raw material. Due to its excellent physical and chemical properties (high hardness, high melting point, corrosion resistance, good insulation, etc.), it is widely used in metallurgy, refractory materials, chemical engineering, and other fields. In the production process, rapid and accurate detection of the alumina and other key component content in raw materials, intermediate products, and final products is a core requirement for controlling product quality, optimizing production processes, and reducing energy costs.

[0003] Currently, industrial detection of target indicators in the aforementioned materials largely relies on traditional chemical analysis methods, such as X-ray fluorescence spectrometry (XRF), inductively coupled plasma atomic emission spectrometry (ICP), or EDTA complexometric titration. These methods suffer from the following significant drawbacks, all of which are problems that this invention can specifically address:

[0004] 1. Complex sample pretreatment: It requires grinding, digestion and other steps, which takes up to several hours and cannot meet the needs of rapid detection;

[0005] 2. Long analysis cycle: A single test requires waiting for the chemical reaction to complete or for the instrument to preheat and be calibrated, which makes it difficult to meet the real-time monitoring needs of the production site;

[0006] 3. Destructive testing: Samples cannot be recycled and reused after pretreatment, resulting in waste of raw materials;

[0007] 4. High operating threshold: It requires professional personnel to operate in a laboratory environment, the operation is cumbersome, and it cannot be directly applied to the production line;

[0008] 5. Limitations of single-component detection: Traditional methods usually require setting up separate detection procedures for a single indicator. For example, EDTA titration is used to measure Al2O3, while gravimetric method is used to measure SiO2. When detecting multiple components, repeated operations are required, resulting in extremely low efficiency.

[0009] Near-infrared (NIR) or mid-infrared (MIR) spectroscopy offers advantages such as speed, non-destructive testing, environmental friendliness, simultaneous multi-component determination, and ease of online detection, leading to its widespread application in various fields. However, its application in the rapid detection of multi-component content in alumina and complex industrial materials still faces challenges, including weak signals, numerous interferences, and difficulties in modeling. Specifically, these challenges include: 1. Difficulty in capturing weak signals: The Al-O bond vibrational transition energy of components such as Al2O3 is low, resulting in weak characteristic absorption signals in the NIR / MIR region. Furthermore, the characteristic peaks in the MIR region overlap with strong absorption peaks of SiO2 and Fe2O3, making them difficult to distinguish using conventional spectrometers; 2. Complex matrix interference: Differences in material particle size lead to light scattering fluctuations, strong moisture absorption masks the target signal, and changes in packing density alter the optical path, all of which interfere with spectral accuracy; 3. Unstable quantitative models: The matrix differences between materials from different sources are significant, resulting in insufficient coverage of traditional models; high-content components easily enter the spectral nonlinear region, and changes in temperature and humidity can easily lead to model prediction deviations.

[0010] Therefore, how to achieve simultaneous online detection of the inherent content of multiple components has become an urgent problem to be solved. Summary of the Invention

[0011] The purpose of this invention is to address the technical problem of simultaneous online detection of the inherent content of multiple components by proposing an in-situ detection device and method for the inherent content of multiple components. This invention enables rapid online sampling and infrared synchronous detection of materials on a vehicle by a robotic arm, thereby achieving simultaneous detection of alumina and SiO2, Fe2O3, Na2O, CaO, loss on ignition, Li2O, angle of repose, bulk density, particle size, tonnage, and K2O.

[0012] The working principle of this invention is as follows: A robotic arm picks up the material to be tested and places it on a conveyor belt. An infrared spectroscopy emission module emits near-infrared or mid-infrared light onto the material. When the light shines on the material, each component in the material selectively absorbs light of a specific wavelength. The spectral receiving module collects the transmitted or reflected spectral signals. Subsequently, the signal processing module performs adaptive preprocessing on the collected spectra to eliminate the adverse effects of interference factors such as particle size, moisture, and temperature. The preprocessed spectral data is then input into a pre-constructed multi-component quantitative correction model. This model is established based on the correlation between the spectral data and chemical values ​​of standard samples with different contents. It can adapt to the nonlinear relationship of complex matrices and predict the contents of multiple components. Finally, the content of each component in the material to be tested is calculated by the model, and the results are output through the display module.

[0013] This invention discloses an in-situ detection method for the inherent content of multiple components, comprising the following steps:

[0014] Step 1: Alumina and a total of 12 indicators, including SiO2, Fe2O3, Na2O, CaO, loss on ignition, Li2O, angle of repose, loose density, particle size, tonnage, and K2O composition, were used as the sample measurement indicators;

[0015] Step 2: Prepare a standard sample set, and use standard chemical methods to determine the sample indicators to cover the expected content range and matrix changes of the standard samples;

[0016] Step 3: Use an infrared signal acquisition spectrometer to acquire the infrared spectrum of the sample to form raw spectral data;

[0017] Step 4: Adaptive preprocessing is performed on the raw spectral data using a data processing module that removes outliers and eliminates baseline drift and scattering, resulting in preprocessed spectral data;

[0018] Step 5: Using partial least squares and variable selection methods, construct a multi-index joint quantitative calibration model, and train the model using the calibration set and sample measurement indicators; input the validation set into the trained model to obtain the detection accuracy of the inherent content of multiple components;

[0019] Step 5.1: The preprocessed spectral data is randomly divided into a calibration set and a validation set;

[0020] Step 5.2: Construct a multi-index joint quantitative correction model using partial least squares method and variable selection method;

[0021] Step 5.3: Input the calibration set and sample measurement indicators into the multi-index joint quantitative calibration model for training to obtain the trained multi-index joint quantitative calibration model; input the validation set into the trained multi-index joint quantitative calibration model to obtain the detection accuracy of the inherent content of multiple components;

[0022] Step 6: The sample to be tested is picked up by the robotic arm and placed on the conveyor belt. It is transported to the material inlet of the test material via the conveyor belt. The spectrum of the sample to be tested is collected by the infrared signal acquisition spectrometer. After the data processing module processes the spectral data, the sample to be tested is output through the material outlet of the test material. At the same time, the multi-component inherent content detection results are visualized and output through the human-computer interaction module via the output module.

[0023] This invention discloses an in-situ detection device for the inherent content of multiple components, used to implement the above-mentioned method. The in-situ detection device for the inherent content of multiple components disclosed in this invention includes an infrared signal acquisition spectrometer, a robotic arm, a conveyor belt, a sample, a data processing module, a human-machine interaction module, an output module, a test material inlet, and a test material outlet.

[0024] The infrared signal acquisition spectrometer consists of an infrared spectrum emission module and an infrared spectrum receiving module, which are used to acquire infrared signals and will serve as the input to the data processing module.

[0025] Furthermore, the infrared spectral emission module, as the core of the light source, is used to generate and output detection light in the near-infrared or mid-infrared bands, and switches the bands according to the characteristics of the material matrix to emit infrared light.

[0026] Furthermore, the infrared spectrum receiving module employs an infrared detector to receive optical signals and convert them into electrical signals, and has signal amplification capabilities, enabling it to capture weak characteristic absorption signals.

[0027] The robotic arm is used to pick up the sample to be tested and place it on the conveyor belt;

[0028] The conveyor belt is used to place and transport the material to be tested, ensuring that the infrared detection light acts uniformly on the surface of the material.

[0029] The sample is used for detection by an infrared signal acquisition spectrometer;

[0030] The data processing module is used for adaptive preprocessing of spectral signals, including noise reduction, baseline correction, and scattering correction, and for correlating sample measurement indicators with chemical values. It automatically selects processing strategies based on the spectral interference types of different matrix materials; and serves as the input to the output module.

[0031] The human-computer interaction module is used to visually output the detection results of the inherent content of multiple components;

[0032] The output module is used to call the model to calculate the preprocessed spectral data, obtain the detection results of the inherent content of multiple components, and output them through the human-computer interaction module;

[0033] The material inlet is used to place the sample to be tested into the conveyor belt;

[0034] The test material outlet is used to recover the tested sample;

[0035] The workflow of this system is as follows:

[0036] The robotic arm places the sample to be tested on the conveyor belt, turns on the infrared signal acquisition spectrometer, and opens the conveyor belt to transport the sample to be tested from the material inlet to the detection area. During the detection process, the infrared signal acquisition spectrometer collects infrared light from the sample, and at the same time, the data processing module performs data analysis and displays the analysis results on the human-machine interaction module through the output module. The sample is then recovered through the material outlet.

[0037] Compared with existing technologies, it has the following beneficial effects:

[0038] 1. Simultaneous detection of multiple components: Traditional methods require setting up separate detection procedures for each single indicator, which is cumbersome and time-consuming. This invention combines a single spectral scan with a multi-component joint quantitative model to simultaneously detect 12 indicators, including Al2O3, SiO2, Fe2O3, Na2O, CaO, loss on ignition, Li2O, angle of repose, loose density, particle size, tonnage, and K2O, and outputs the results simultaneously. This significantly reduces the detection cycle and operation steps, and greatly improves detection efficiency.

[0039] 2. Fast and efficient: A test can be completed in tens of seconds, which is much faster than traditional chemical methods, greatly improving the detection efficiency and making it suitable for online real-time monitoring;

[0040] 3. Non-destructive and environmentally friendly: No complicated sample pretreatment is required, no chemical reagents are consumed, and samples can be recycled and reused, making it green and environmentally friendly;

[0041] 4. Strong online capability: The device can be integrated into the production line to achieve real-time quality feedback and control, thereby improving product quality stability;

[0042] 5. Easy to operate: It has a high degree of automation, requires less professional skills from operators, and reduces labor costs.

[0043] 6. Adaptive preprocessing: Built-in algorithms automatically identify and eliminate interference, improving model robustness and applicability.

[0044] 7. Potential for expanding multiple indicators: The range of detection indicators can be further expanded by updating the standard sample set and model, making it suitable for more industrial scenarios. Attached Figure Description

[0045] Figure 1 This is a flowchart of the detection method of the present invention;

[0046] Figure 2 This is a schematic diagram of the device of the present invention;

[0047] Figure 3 A correlation diagram showing the relationship between predicted and actual measured values ​​of alumina and its components, established using partial least squares (PLS) method.

[0048] Among them, 1-infrared signal acquisition spectrometer, 2-conveyor belt, 3-sample, 4-data processing module, 5-human-machine interaction module, 6-output module, 7-inlet of the material to be tested, 8-outlet of the material to be tested, and 9-robotic arm. Detailed Implementation

[0049] To better illustrate the purpose and advantages of the present invention, the invention will be further described below with reference to the accompanying drawings and examples. It should be noted that the implementation of the present invention is not limited to the following embodiments, and any modifications or alterations made to the present invention will fall within the protection scope of the present invention.

[0050] Example

[0051] like Figure 1 As shown in the figure, the specific implementation steps of the in-situ detection method for the inherent content of multiple components in this embodiment are as follows:

[0052] Step 1: Alumina and a total of 12 indicators, including SiO2, Fe2O3, Na2O, CaO, loss on ignition, Li2O, angle of repose, loose density, particle size, tonnage, and K2O composition, were used as the sample measurement indicators;

[0053] Step 2: Prepare a standard sample set, and use standard chemical methods to determine the sample indicators to cover the expected content range and matrix changes of the standard samples;

[0054] Step 3: Use an infrared signal acquisition spectrometer to acquire the infrared spectrum of the sample to form raw spectral data;

[0055] Step 4: Adaptive preprocessing is performed on the raw spectral data using a data processing module that removes outliers and eliminates baseline drift and scattering, resulting in preprocessed spectral data;

[0056] Step 5: Using partial least squares and variable selection methods, construct a multi-index joint quantitative calibration model, and train the model using the calibration set and sample measurement indicators; input the validation set into the trained model to obtain the detection accuracy of the inherent content of multiple components;

[0057] Step 5.1: The preprocessed spectral data is randomly divided into a calibration set and a validation set;

[0058] Step 5.2: Construct a multi-index joint quantitative correction model using partial least squares method and variable selection method;

[0059] Step 5.3: Input the calibration set and sample measurement indicators into the multi-index joint quantitative calibration model for training to obtain the trained multi-index joint quantitative calibration model; input the validation set into the trained multi-index joint quantitative calibration model to obtain the detection accuracy of the inherent content of multiple components;

[0060] Step 6: The sample to be tested is placed on the conveyor belt by the robotic arm and transported to the detection area. The spectrum of the sample to be tested is collected by the infrared signal acquisition spectrometer. After the data processing module processes the spectral data, the sample to be tested is output through the test material outlet. At the same time, the detection results of the inherent content of multiple components are visualized through the human-computer interaction module.

[0061] like Figure 2 As shown, this embodiment provides an in-situ detection device for the inherent content of multiple components, used to implement the above-mentioned method. This embodiment also provides an online detection device for the inherent content of multiple components using robotic arm sampling combined with infrared spectroscopy, comprising an infrared signal acquisition spectrometer, a robotic arm, a conveyor belt, a sample, a data processing module, a human-machine interaction module, an output module, a test material inlet, and a test material outlet.

[0062] The infrared signal acquisition spectrometer consists of an infrared spectrum emission module and an infrared spectrum receiving module, which are used to acquire infrared signals and will serve as the input to the data processing module.

[0063] Furthermore, the infrared spectral emission module, as the core of the light source, is used to generate and output detection light in the near-infrared or mid-infrared bands, and switches the bands according to the characteristics of the material matrix to emit infrared light.

[0064] Furthermore, the infrared spectrum receiving module employs an infrared detector to receive optical signals and convert them into electrical signals, and has signal amplification capabilities, enabling it to capture weak characteristic absorption signals.

[0065] The robotic arm is used to pick up the sample to be tested and place it on the conveyor belt;

[0066] The conveyor belt is used to place and transport the material to be tested, ensuring that the infrared detection light acts uniformly on the surface of the material.

[0067] The sample is used for detection by an infrared signal acquisition spectrometer;

[0068] The data processing module is used for adaptive preprocessing of spectral signals, including noise reduction, baseline correction, and scattering correction, and for correlating sample measurement indicators with chemical values. It automatically selects processing strategies based on the spectral interference types of different matrix materials; and serves as the input to the output module.

[0069] The human-computer interaction module is used to visually output the detection results of the inherent content of multiple components;

[0070] The output module is used to call the model to calculate the preprocessed spectral data, obtain the detection results of the inherent content of multiple components, and output them through the human-computer interaction module;

[0071] The material inlet is used to place the sample to be tested into the conveyor belt;

[0072] The test material outlet is used to recover the tested sample;

[0073] The workflow of this system is as follows:

[0074] The robotic arm places the sample to be tested on the conveyor belt, the infrared signal acquisition spectrometer is turned on, and the conveyor belt is opened to transport the sample to be tested from the material inlet to the testing area. During the testing process, the infrared signal acquisition spectrometer collects infrared light from the sample. At the same time, the data processing module performs data analysis and displays the analysis results on the human-machine interaction module through the output module. The test sample is then recovered through the material outlet.

[0075] To further illustrate the advantages of the present invention, an explanation will be given in conjunction with experimental data.

[0076] like Figure 3 The figure shows the correlation between the predicted and actual measured values ​​of the contents of 12 intrinsic components, established using partial least squares (PLS) based on the modeling results of the near-infrared absorbance curves. RMS E represents the root mean square error, indicating the average magnitude of the difference between the predicted and observed values. 2 It represents the proportion of the total variance in the target variable that the model's predictions can explain, and the magnitude of the value reflects how well the prediction set fits the model.

[0077] The above detailed description further illustrates the purpose, technical solution, and beneficial effects of the invention. It should be understood that the above description is only a specific embodiment of the present invention and is not intended to limit the scope of protection of the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A method for in-situ detection of the inherent content of multiple components, characterized in that: Includes the following steps, Step 1: Alumina and a total of 12 indicators, including SiO2, Fe2O3, Na2O, CaO, loss on ignition, Li2O, angle of repose, loose density, particle size, tonnage, and K2O composition, were used as the sample measurement indicators; Step 2: Prepare a standard sample set, and use standard chemical methods to determine the sample indicators to cover the expected content range and matrix changes of the standard samples; Step 3: Use an infrared signal acquisition spectrometer to acquire the infrared spectrum of the sample to form raw spectral data; Step 4: Adaptive preprocessing is performed on the raw spectral data using a data processing module that removes outliers and eliminates baseline drift and scattering, resulting in preprocessed spectral data; Step 5: Using partial least squares and variable selection methods, construct a multi-index joint quantitative calibration model, and train the model using the calibration set and sample measurement indicators; input the validation set into the trained model to obtain the detection accuracy of the inherent content of multiple components; Step 5.1: The preprocessed spectral data is randomly divided into a calibration set and a validation set; Step 5.2: Construct a multi-index joint quantitative correction model using partial least squares method and variable selection method; Step 5.3: Input the calibration set and sample measurement indicators into the multi-index joint quantitative calibration model for training to obtain the trained multi-index joint quantitative calibration model; input the validation set into the trained multi-index joint quantitative calibration model to obtain the detection accuracy of the inherent content of multiple components; Step 6: The sample to be tested is picked up by the robotic arm and placed on the conveyor belt. It is transported to the material inlet of the test material via the conveyor belt. The spectrum of the sample to be tested is collected by the infrared signal acquisition spectrometer. After the data processing module processes the spectral data, the sample to be tested is output through the material outlet of the test material. At the same time, the detection results of the inherent content of multiple components are visualized through the human-computer interaction module via the output module.

2. A multi-component intrinsic content in-situ detection device for implementing the method described in claim 1, characterized in that: It includes an infrared signal acquisition spectrometer, a robotic arm, a conveyor belt, a sample, a data processing module, a human-machine interaction module, an output module, a test material inlet, and a test material outlet; The infrared signal acquisition spectrometer consists of an infrared spectrum emission module and an infrared spectrum receiving module, which are used to acquire infrared signals and will serve as the input to the data processing module. The robotic arm is used to pick up the sample to be tested and place it on the conveyor belt; The conveyor belt is used to place and transport the material to be tested, ensuring that the infrared detection light acts uniformly on the surface of the material. The sample is used for detection by an infrared signal acquisition spectrometer; The data processing module is used to perform noise reduction, baseline correction, and scattering correction on the spectral signal, perform adaptive preprocessing, and correlate the sample measurement indicators with chemical values. It also automatically selects processing strategies based on the spectral interference type of different matrix materials. It will be used as the input to the output module; The human-computer interaction module is used to visually output the detection results of the inherent content of multiple components; The output module is used to call the model to calculate the preprocessed spectral data, obtain the detection results of the inherent content of multiple components, and output them through the human-computer interaction module; The material inlet is used to place the sample to be tested into the conveyor belt; The test material outlet is used to recover the tested sample; The workflow of this system is as follows: The robotic arm places the sample to be tested on the conveyor belt, the infrared signal acquisition spectrometer is turned on, and the conveyor belt is opened to transport the sample from the material inlet to the detection area. During the detection process, the infrared signal acquisition spectrometer collects infrared light from the sample, and the data processing module performs data analysis. The analysis results are displayed on the human-machine interaction module through the output module, and the sample is recovered through the material outlet.

3. The in-situ detection device for the inherent content of multiple components as described in claim 2, characterized in that: The infrared spectral emission module, as the core of the light source, is used to generate and output detection light in the near-infrared or mid-infrared bands, and switches the band according to the characteristics of the material matrix to emit infrared light. The infrared spectral receiving module uses an infrared detector to receive optical signals and convert them into electrical signals. It has signal amplification function and can capture weak characteristic absorption signals.

Images