Device and method for detecting moisture content of sand mold

Abstract

The invention discloses a device and a method for detecting the moisture content of a sand mold. The device comprises a conveying device, a laser sensor, a PLC (Programmable Logic Controller) and a detection device, wherein the conveying device is used for conveying a sand mold sample to be detected, the laser sensor is used for detecting whether the sand mold sample reaches a detection area or not and transmitting detected sample in-place information to the PLC, and the PLC is used for receiving the sample in-place information and sending a sample detection instruction to the detection device; and the detection device is used for detecting the sand mold sample reaching the detection area by utilizing a near infrared spectrum technology. By means of the device and the detection method based on the device, the precision and efficiency of sand mold moisture detection are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of moisture detection, and more particularly to a device and method for detecting the moisture content of a sand mold. Background Art

[0002] Sand casting is a common process technology in modern manufacturing and is widely used in the production of various metal castings. The quality of the sand mold directly affects the precision and performance of the casting, and moisture content is one of the important parameters of sand mold quality. Traditional moisture detection methods such as drying method and chemical analysis method, although with high accuracy, are destructive detection, and the operation is complicated and time-consuming, and are not suitable for real-time detection on the production line.

[0003] As a rapid and non-destructive detection technology, near infrared spectroscopy has been widely used in the fields of agricultural products, food, and medicine in recent years. Its basic principle is based on the characteristic absorption bands of different substances in the near infrared light region. By analyzing the spectral signal, the composition and content of the substance can be inferred. However, the application of near infrared spectroscopy in the detection of moisture content in sand molds is not yet popular, and there are few related studies.

[0004] On the other hand, due to the complexity of sand materials and the non-uniformity of moisture distribution, there are still many challenges in applying existing technologies to sand moisture detection. For example, the spectral signal is greatly affected by factors such as ambient temperature, sample thickness and surface roughness, resulting in unstable detection results. In addition, the spectral characteristics of different types of sand materials vary significantly, requiring targeted model establishment and optimization.

[0005] Therefore, how to improve the accuracy of sand mold moisture detection is a problem that technical personnel in this field need to solve urgently. Summary of the invention

[0006] In view of this, the present invention provides a device and method for detecting the moisture content of a sand mold, which are used to at least solve some of the technical problems in the background technology.

[0007] In order to achieve the above object, the present invention adopts the following technical solution:

[0008] On the one hand, the present invention discloses a device for detecting the moisture content of a sand mold, comprising: a conveying device, a laser sensor, a PLC controller and a detection mechanism;

[0009] The conveying device is used to transport the sand sample to be tested, the laser sensor is used to detect whether the sand sample has reached the test area and transmit the detected sample arrival information to the PLC controller, and the PLC controller is used to receive the sample arrival information and send a sample detection instruction to the detection mechanism;

[0010] The detection mechanism uses near infrared spectroscopy technology to detect the water content of the sand sample reaching the detection area.

[0011] Preferably, the conveying device includes a conveying belt made of polyurethane high wear-resistant material

[0012] Preferably, the surface of the conveyor belt is designed as an anti-slip structure with a hexagonal honeycomb arrangement.

[0013] Preferably, the detection mechanism includes a black box with no bottom plate at the bottom. When the detection structure receives the information that the sample is in place, the black box is turned upside down in the detection area under the drive of the robotic arm and covers the sand mold sample; the interior of the black box is evenly arranged with halogen lamps that can provide stable and uniform light, cooling fans for maintaining a stable detection environment temperature, and optical fibers connected to a fiber optic spectrometer and used to receive reflected spectral data of the sand mold sample; the spectral data received by the optical fiber is initially processed by the fiber optic spectrometer and then transmitted to a computing terminal for storage and final processing, and the detection result of the moisture content of the sand mold sample is obtained using a trained support vector regression model.

[0014] Preferably, the black box adopts a double-layer structure, wherein the outer layer is a high-strength aluminum alloy and the inner layer is an aluminum oxide nanoparticle coating, and the aluminum oxide nanoparticle coating is evenly covered on the inner surface of the black box through an electrophoretic coating process.

[0015] On the other hand, the present invention also discloses a method for detecting the moisture content of a sand mold, comprising the following steps:

[0016] When the sand sample to be tested reaches the testing area by means of the transmission device, the spectrum data reflected by the sand sample is obtained;

[0017] Preprocess the spectral data to remove abnormal data;

[0018] Correct the spectral data with the abnormal data removed for environmental factors;

[0019] The corrected spectral data system uses a continuous projection algorithm to screen out several characteristic wavelengths related to the moisture content in the full spectrum;

[0020] The spectral data corresponding to the screened characteristic wavelengths are input into the trained support vector regression model to obtain the water content data of the sand sample to be tested.

[0021] Further, obtaining the spectral data reflected by the sand sample to be tested specifically includes the following steps:

[0022] The PLC controller is used to control the optical fiber to achieve multi-point scanning, and the spectral data of multiple discrete spectral data points are obtained on each sand mold sample.

[0023] Furthermore, the spectral data is preprocessed to remove abnormal data, which specifically includes the following steps:

[0024] Calculate the original spectrum S for each spectral data point i The Mahalanobis distance D between (λ) and the mean μ(S) of the data set i ;

[0025] Set the threshold D th , if the Mahalanobis distance D between the original spectral data of the i-th data point and the mean of the data set i >D th , it is considered as abnormal data and removed.

[0026] Furthermore, the Mahalanobis distance D i The calculation expression is as follows:

[0027]

[0028] Where C is the covariance matrix of spectral data of multiple spectral data points.

[0029] Furthermore, the spectral data from which the abnormal data is removed is corrected for environmental factors, specifically comprising the following steps:

[0030] The multivariate adaptive regression spline algorithm is used to establish a nonlinear relationship model between environmental factors and spectral deviations, and the spectral data with abnormal data removed is corrected for environmental factors. Specifically, the following nonlinear relationship model is used:

[0031] S'(λ)=S(λ)+f(T,H,D)

[0032] Among them, S'(λ) is the corrected spectral data, S(λ) is the preprocessed spectral data, f(T,H,D) is the multivariate adaptive regression spline model, T represents the temperature factor, H represents the humidity factor, and D represents the dust concentration.

[0033] It can be seen from the above technical solutions that, compared with the prior art, the present invention discloses a device and method for detecting the moisture content of a sand mold, which has the following beneficial effects:

[0034] Compared with the complex operation and time-consuming traditional methods, this device greatly shortens the detection time, can realize real-time and rapid detection on the transmission device, and improves production efficiency.

[0035] The device reduces errors caused by environmental changes and sample differences, ensuring the accuracy and reliability of the test data. BRIEF DESCRIPTION OF THE DRAWINGS

[0036] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on the provided drawings without paying creative work.

[0037] Figure 1 This is a schematic structural diagram of a sand mold moisture content detection device provided in an embodiment of the present invention.

[0038] Figure 2 A schematic diagram of the surface structure of a conveyor belt provided in an embodiment of the present invention.

[0039] Figure 3 This is a flow chart of the detection process of the sand mold moisture content detection device provided in an embodiment of the present invention.

[0040] Figure 4 A schematic flow chart of a method for detecting moisture content in sand molds provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0041] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0042] The embodiment of the present invention first proposes a sand mold moisture content detection device based on near infrared spectroscopy technology. The device can quickly and accurately detect the moisture content of the sand mold without destroying the sample. Figure 1 As shown, the detection device is mainly composed of the following parts: conveyor belt 1, sensor 2, PLC controller 3, cooling fan 4, black box 5, halogen lamp 6, optical fiber 7, spectrometer 8 and computer 9. The design and function of each component are as follows:

[0043] Conveyor belt 1: Made of highly wear-resistant polyurethane material, it ensures that the sand sample can be transported to the testing area at a constant and adjustable speed.

[0044] like Figure 2 As shown, in a specific embodiment, the surface of the conveyor belt is designed as a honeycomb + groove composite anti-skid structure, and the hexagonal honeycomb pits are used to increase friction and ensure that the sand sample 10 does not slide or deviate from the track during transmission.

[0045] Sensor 2: It uses a high-precision laser sensor, which can accurately detect the presence of the sand sample and immediately send a signal to the PLC controller.

[0046] PLC controller 3: mainly used to receive sample arrival information and send sample detection instructions to the detection mechanism. In the embodiment of the present invention, the detection mechanism is composed of a black box 5 and an optical fiber 7 inside the black box, an optical fiber spectrometer 8 connected to the optical fiber 7, and a computer 9.

[0047] In specific applications, the PLC controller can adopt a multi-layer architecture control architecture, which includes three main subsystems: a top-level decision module, a middle-level coordination module, and a bottom-level execution module.

[0048] In terms of information interaction and task synchronization planning: the top-level decision-making module is responsible for overall process planning and resource allocation, while the middle-level coordination module handles information interaction and task synchronization between subsystems to ensure the coordination and consistency of the entire detection process.

[0049] In terms of instructions and status feedback: the control instructions issued by the PLC include: controlling the start and stop and speed of the conveyor belt, controlling the switch and brightness of the halogen lamp, controlling the switch and connection of the fan, and controlling the optical fiber to start collecting data. The corresponding status feedback includes: the sensor feedbacks the location information of the sand sample, the spectrometer feedbacks the data collection completion status, and the computer system feedbacks the progress and results of data processing.

[0050] Black box 5: In the embodiment of the present invention, black box 5 adopts a double-layer structure, the outer layer is a high-strength aluminum alloy, and the inner wall is coated with aluminum oxide nanoparticles. The coating is evenly covered on the inner wall surface through an electrophoretic coating process. It is used to isolate external light interference and provide a constant, light-free collection environment. The sensor detects the signal that the sample has reached the detection area, the PLC controls the conveyor belt to stop, and then the black box is covered with the sand sample manually or by a mechanical arm. After the detection is completed, the black box is lifted. The black box is placed in the gap between the staggered honeycombs.

[0051] The cooling fan 4 is installed inside the black box to remove the heat generated by the halogen lamp working for a long time and keep the temperature of the detection environment stable.

[0052] The halogen lamps 6 are symmetrically arranged inside the black box to provide light with uniform and stable intensity to illuminate the surface of the sand sample.

[0053] The optical fiber 7 is used to efficiently receive the spectral data reflected from the inside of the sample.

[0054] As a preferred embodiment, the optical fiber 7 can realize horizontal movement on the top of the black box 5 under the control of the PLC controller, thereby realizing multi-point scanning of the sand mold sample 10 to be tested. Specifically, an electric slide 11 capable of horizontal movement is installed on the top of the black box, and the optical fiber 7 is fixedly installed below the electric slide 11, and multi-point scanning of the sand mold sample 10 to be tested is realized with the help of the electric slide 11.

[0055] The fiber optic spectrometer 8 receives the spectrum data through the optical fiber 7 and transmits the data to the computer system.

[0056] The computer 9 stores and processes the received spectral data, and obtains and outputs the detection result of the moisture content of the sand mold using the trained support vector regression model.

[0057] The detection process and principle of the present invention are further explained below. The detection process is divided into six stages: sample delivery, sample detection, signal processing, spectrum acquisition, data processing and result output.

[0058] Sample transport: The sand sample is placed at the starting position of the conveyor belt. The conveyor belt runs at a stable and controllable speed, transporting the sample from the left side of the device to the spectrum collection area.

[0059] Sample detection: When the sample reaches the detection area in the middle of the device, the sensor detects the presence of the sample. The sensor adopts a non-contact design, which can accurately detect the position of the sample and immediately send a signal to the PLC input.

[0060] Signal processing: When the PLC controller receives the signal from the sensor, the internal program immediately starts the timing function. This process ensures the precise synchronization of subsequent spectrum acquisition and sample position.

[0061] Spectrum acquisition: After receiving the acquisition signal from the PLC controller, the computer starts the spectrum acquisition program. At this time, the sand sample has reached the bottom of the optical fiber and is ready to receive the near-infrared spectrum information.

[0062] In order to ensure that the spectrum collection process is not disturbed by external light, the sample passes through a special black box. The black box has a unique design and the inner wall has been specially treated to effectively isolate external light and provide a constant, light-pollution-free collection environment.

[0063] Halogen lamp light sources are arranged symmetrically inside the black box. The light intensity emitted by the halogen lamp is uniform and stable, irradiating the surface of the sand sample to ensure consistent lighting conditions at each collection point. The choice of halogen lamp light source is based on its spectral characteristics, which can provide sufficient light intensity in the near-infrared range to ensure the clarity of the spectral signal.

[0064] In order to prevent the heat generated by the halogen lamp from affecting the temperature of the detection environment when it is working for a long time, a cooling fan is installed inside the black box. The cooling fan removes excess heat through air convection, keeps the temperature inside the black box stable, and avoids the interference of temperature changes on the spectral data.

[0065] Data transmission: The optical fiber is installed at the top of the black box and connected to the spectrometer. The efficient transmission characteristics of the optical fiber ensure that the spectral data reflected from the inside of the sample can be fully transmitted to the spectrometer. After receiving the spectral data, the spectrometer performs a preliminary spectral analysis and transmits it to the computer system in real time through the data line.

[0066] Secondly, the embodiment of the present invention discloses a method for detecting the moisture content of a sand mold based on near infrared spectroscopy technology. The method can be implemented using a disclosed sand mold moisture content detection device. The detection process is as follows: Figure 3 , Figure 4 As shown, the sand mold sample to be tested is first transferred to the detection area for spectral data collection. Since the sand mold surface may be uneven, the system uses multi-point scanning technology to obtain multiple spectral data points on each sample to ensure the quality and reliability of the data. The computer system then further processes the spectral data of multiple spectral data points received from the same sample.

[0067] Specifically, the processing process of the spectral data of the sand mold sample is as follows: first, the collected raw data is preprocessed to remove abnormal spectral data. In the embodiment of the present invention, the Mahalanobis distance is used to identify and exclude abnormal spectral data. For each spectral data point, the original spectrum S i (λ), calculate its Mahalanobis distance D with the mean μ(S) of the data set i :

[0068] Where C is the covariance matrix of the spectral data of multiple spectral data points, and “′” represents the matrix transposition symbol.

[0069] Set the threshold D th , if the Mahalanobis distance D between the original spectral data of the i-th data point and the mean of the data set i >D th , they are considered as abnormal data and removed from subsequent analysis.

[0070] Secondly, considering the influence of the temperature and dusty environment of the foundry on the near-infrared spectral measurement, the environmental factor correction module was introduced to correct the spectral data of the pre-processed sand sample. This module uses the multivariate adaptive regression spline (MARS) algorithm to establish a nonlinear relationship model between environmental factors and spectral deviation:

[0071] S'1(λ)=S(λ)+f(T,H,D)

[0072] Among them, S'1(λ) is the corrected spectrum, S(λ) is the spectral data of the sand sample after preprocessing, f(T,H,D) is the MARS model, and T, H, and D represent temperature, humidity, and dust concentration, respectively.

[0073] Next, the calibrated sand sample spectral data is used to select characteristic wavelengths using the successive projection algorithm (SPA). The successive projection algorithm iteratively selects wavelengths to minimize the redundancy between each newly selected wavelength and the wavelength selected in the previous round.

[0074] Specifically, a wavelength is randomly selected from the corrected spectral data S'1(λ) as the starting wavelength, and projection selection is performed iteratively to ensure that each newly selected wavelength λ j The projection distance to the selected wavelength is maximized, thereby minimizing collinearity. The projection distance formula is:

[0075]

[0076] Among them, p j is the projection vector of the jth wavelength, s j is the spectral vector corresponding to the current wavelength. k It represents the spectral data vector corresponding to the kth selected characteristic wavelength.

[0077] After N rounds of iterations, a set of wavelength combinations λ with maximum linear independence from the original data is obtained. out , used as input for the support vector regression model.

[0078] A support vector regression model SVR for predicting the moisture content of sand samples is constructed and trained. The spectral data S'3(λ) corresponding to the characteristic wavelength composition selected by the continuous projection algorithm SPA is used as the input of the trained support vector regression model (SVR) to obtain the predicted moisture content of the sand samples.

[0079] In the SVR model, the selected characteristic wavelength is reflected by the data S'3(λ). Through the optimized characteristic wavelength, the SVR model can capture the complex nonlinear relationship between moisture content and spectrum. The basic expression of the SVR model is as follows:

[0080]

[0081] in, is the predicted moisture content, K(S' i ,S') is the kernel function, α i and is the Lagrange multiplier, representing the weight of the support vector, and b is the bias term.

[0082] After the data is analyzed, the computer system generates a test report and displays the results on the user interface. The user interface displays the moisture content of each sample in real time, allowing you to visually observe the test results.

[0083] In this specification, each embodiment is described in a progressive manner, and each embodiment focuses on the differences from other embodiments. The same or similar parts between the embodiments can be referred to each other. For the device disclosed in the embodiment, since it corresponds to the method disclosed in the embodiment, the description is relatively simple, and the relevant parts can be referred to the method part.

[0084] The above description of the disclosed embodiments enables one skilled in the art to implement or use the present invention. Various modifications to these embodiments will be apparent to one skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention will not be limited to the embodiments shown herein, but rather to the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A device for detecting moisture content of sand molds, characterized in that: include: Conveying device, laser sensor, PLC controller and detection mechanism; The conveying device is used to transport the sand sample to be tested, the laser sensor is used to detect whether the sand sample has reached the test area and transmit the detected sample arrival information to the PLC controller, and the PLC controller is used to receive the sample arrival information and send a sample detection instruction to the detection mechanism; The detection mechanism uses near infrared spectroscopy technology to detect the water content of the sand sample reaching the detection area.

2. A sand mold moisture content detection device according to claim 1, characterized in that: The conveying device comprises a conveying belt made of polyurethane highly wear-resistant material.

3. A sand mold moisture content detection device according to claim 1, characterized in that: The surface of the conveyor belt is designed as an anti-slip structure with a hexagonal honeycomb arrangement.

4. A sand mold moisture content detection device according to claim 1, characterized in that: The detection mechanism comprises a black box, the bottom of which has no bottom plate. When the detection structure receives the information that the sample is in place, the black box is turned upside down in the detection area under the drive of the mechanical arm and covers the sand sample; the inside of the black box is evenly arranged with halogen lamps that can provide stable and uniform light, a cooling fan for maintaining a stable detection environment temperature, and an optical fiber connected to a fiber optic spectrometer and used to receive the reflection spectrum data of the sand sample; The spectral data received by the optical fiber is initially processed by the optical fiber spectrometer and then transmitted to the computing terminal for storage and final processing, and the detection result of the moisture content of the sand mold sample is obtained by using the trained support vector regression model.

5. A sand mold moisture content detection device according to claim 4, characterized in that: The black box adopts a double-layer structure, wherein the outer layer is a high-strength aluminum alloy and the inner layer is an aluminum oxide nanoparticle coating, and the aluminum oxide nanoparticle coating is evenly covered on the inner surface of the black box through an electrophoretic coating process.

6. A method for detecting the moisture content of a sand mold, characterized in that: The following steps are involved: When the sand sample to be tested reaches the testing area by means of the transmission device, the spectrum data reflected by the sand sample is obtained; Preprocess the spectral data to remove abnormal data; Correct the spectral data with the abnormal data removed for environmental factors; The corrected spectral data system uses a continuous projection algorithm to screen out several characteristic wavelengths related to the moisture content in the full spectrum; The spectral data corresponding to the screened characteristic wavelengths are input into the trained support vector regression model to obtain the water content data of the sand sample to be tested.

7. A method for detecting moisture content of sand molds according to claim 6, characterized in that: Obtaining the spectral data reflected by the sand sample to be tested specifically includes the following steps: The PLC controller is used to control the optical fiber to achieve multi-point scanning, and the spectral data of multiple discrete spectral data points are obtained on each sand mold sample.

8. A method for detecting moisture content of sand molds according to claim 6, characterized in that: Preprocess the spectral data to remove abnormal data, including the following steps: Calculate the original spectrum S for each spectral data point i The Mahalanobis distance D between (λ) and the mean μ(S) of the data set i ; Set the threshold D th , if the Mahalanobis distance D between the original spectral data of the i-th data point and the mean of the data set i >D th , it is considered as abnormal data and removed.

9. A method for detecting moisture content of sand molds according to claim 8, characterized in that: The Mahalanobis distance D i The calculation expression is as follows: Where C is the covariance matrix of spectral data of multiple spectral data points.

10. A method for detecting moisture content of sand molds according to claim 6, characterized in that: Correcting the spectral data from which the abnormal data is removed for environmental factors specifically includes the following steps: The multivariate adaptive regression spline algorithm is used to establish a nonlinear relationship model between environmental factors and spectral deviations, and the spectral data with abnormal data removed is corrected for environmental factors. Specifically, the following nonlinear relationship model is used: S'(λ)=S(λ)+f(T,H,D) Among them, S'(λ) is the corrected spectral data, S(λ) is the preprocessed spectral data, f(T,H,D) is the multivariate adaptive regression spline model, T represents the temperature factor, H represents the humidity factor, and D represents the dust concentration.

Images