Concrete slump prediction method and device based on image and flow velocity estimation
Through image and flow rate estimation technology, combined with machine learning and mixer data, the concrete slump is predicted in real time, which solves the accuracy and real-time problems of traditional detection methods, improves the accuracy and efficiency of the construction process, and ensures the quality and performance of the concrete .
Patent Information
- Application Number
- CN202411458352.1
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-10-18
- Publication Date
- 2024-11-22
- Estimated Expiration
- 2044-10-18
AI Technical Summary
Traditional concrete slump detection methods have problems such as low accuracy and difficulty in real-time monitoring. They cannot effectively predict changes in concrete fluidity and stability, which affects the real-time regulation of the construction process.
Using concrete slump prediction method based on image and flow velocity estimation, through machine learning and image processing technology, combined with mixer current data and video information, the slump of concrete is predicted in real time, and the formula and flowability are adjusted to ensure the stability of concrete and Liquidity meets the requirements.
Real-time monitoring and prediction of concrete slump is achieved, the accuracy and efficiency of the construction process are improved, the quality and construction performance of concrete are ensured, and production waste is reduced.
Smart Images

Figure CN119006444A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of concrete slump prediction, and in particular to a concrete slump prediction method and device based on image and flow velocity estimation. Background Art
[0002] The traditional method of concrete slump detection is to conduct a slump experiment after sampling. This method is simple and fast to operate, does not require advanced technical equipment, and can be performed on site. However, the operator's movement speed, strength, and observation accuracy may affect the results, thereby reducing the accuracy of the measurement. The most important thing is that the traditional method has a lag in measurement, making it difficult to achieve real-time monitoring and adjustment, which limits our ability to grasp the changes in concrete fluidity and stability in a timely manner, thereby affecting the immediate regulation of the construction process.
[0003] Early studies have shown that during the mixing process, key parameters of the mixer, such as current and power, vary with concrete of different slump standards. However, these parameters are mostly used to judge the mixing state of concrete rather than directly reflect the production quality. In recent years, the progress of deep learning has brought more possibilities to slump prediction technology. Some researchers use machine learning models to learn the relationship between concrete parameters (water-cement ratio, aggregate ratio, mixing time, etc.) and slump. However, these methods are less robust due to possible deviations in concrete raw material parameters. On the other hand, some techniques use gray-level co-occurrence matrices to extract texture information from the concrete surface and then use machine learning algorithms (such as support vector machines) to classify the slump of concrete. However, these methods rely on manually extracted shallow features and have poor effects in actual working conditions. Recent studies use three-dimensional visual sensors to convert concrete slump into digital three-dimensional morphological information for automatic measurement. Although the three-dimensional measurement method is simpler and more accurate than the traditional method, it still requires small batch sampling and measurement after mixing, and cannot achieve real-time monitoring. Summary of the invention
[0004] The invention discloses a concrete slump prediction method and device based on image and flow velocity estimation, which mainly solve the problem of poor robustness of traditional concrete slump prediction methods.
[0005] To achieve the above object, the technical solution of the present invention is implemented as follows: In a first aspect, the present invention provides a method for predicting concrete slump based on image and flow velocity estimation, comprising the following steps: Predict the slump loss value of concrete during transportation, obtain the transportation parameters of previous concrete, fit the regression relationship between slump loss value and transportation distance, humidity and temperature on the day of transportation through machine learning methods, input the actual transportation parameters into the relationship, obtain the slump loss value of concrete, and adjust the concrete formula according to the slump loss value of concrete; Acquire mixer current data to determine the mixing state of concrete, acquire current curve data collected during the concrete mixing process, analyze the relationship between the current curve data and the mixing state, and judge the actual mixing state of concrete based on the relationship between the current curve data and the mixing state; The mixing video of concrete in the mixer is obtained and preprocessed, and the mixing video collected when the concrete mixing state is a uniform mixing state is obtained, and the mixing video is output as a continuous mixing image at the original frame rate, and the continuous mixing image is subjected to fuzzy judgment and defuzzification processing to obtain a preprocessed mixing image sequence number set; Predict the slump of concrete in the mixer, obtain the image information of the preprocessed mixing image sequence data set, establish a three-dimensional convolutional network model based on ResNet according to the image information, obtain the relationship between the mixing video and the slump of the concrete when it is mixed evenly, and input the actual mixing video of the concrete when it is mixed evenly into the three-dimensional convolutional network model to obtain the predicted slump of the concrete when it is mixed evenly; Obtaining and preprocessing the concrete discharge video of the mixer discharge port, obtaining the discharge port discharge video collected when the mixer discharges, outputting it as a continuous discharge port discharge image at the original frame rate, and obtaining a preprocessed discharge port discharge image sequence number set; Predict the slump of concrete at the mixer discharge port, obtain the preprocessed discharge port discharge image sequence data set, calculate the velocity field distribution map between continuous images in the image sequence data set through the particle image velocimetry algorithm, calculate the average flow velocity of concrete according to the velocity field distribution map, and regress and fit the average flow velocity of concrete with the slump to obtain the predicted slump during discharge; Predict the slump of concrete when it leaves the factory, obtain previous concrete order data, and obtain a fully connected neural network for judging the weight of slump values through machine learning training. Input the predicted slump 1 and the predicted slump 2 into the fully connected neural network for weighted decision analysis, and add the weighted values of the two to obtain the predicted slump 3 of concrete when it leaves the factory. Obtaining and preprocessing the concrete unloading video when the mixer truck is unloading, obtaining the mixer truck unloading video collected when the mixer truck is unloading, outputting it as a continuous mixer truck unloading image at the original frame rate, and obtaining a preprocessed mixer truck unloading image sequence number set; Predict the slump of concrete in the mixer truck unloading hopper, obtain the preprocessed mixer truck unloading image sequence data set, calculate the velocity field distribution map between continuous images in the image sequence data set through the particle image velocimetry algorithm, calculate the average flow velocity of the concrete according to the velocity field distribution map, and regress the average flow velocity of the concrete with the slump to calculate the predicted slump four when the mixer truck is unloading, and judge whether it meets the use requirements according to the slump four. If it does not meet the requirements, adjust the fluidity of the concrete until it meets the use requirements.
[0006] In one embodiment, the transportation parameters of the concrete include transportation distance, humidity and temperature on the day of transportation.
[0007] In one embodiment, the time period before the current curve reaches the peak value in the current curve data is the feeding stage, the time period from the peak value to the gradual decrease of the current curve is the mixing stage, and the time period when the current curve remains unchanged or changes slightly within a certain period of time is the mixing uniformity stage; When the current curve is in the feeding stage, the mixing state of the concrete is loose; When the current curve is in the mixing stage, the mixing state of the concrete is the mixed state; When the current curve is in the uniform mixing stage, the mixing state of the concrete is the uniform mixing state.
[0008] In one embodiment, the method for judging whether the current curve remains unchanged or changes slightly is that the difference between the maximum value and the minimum value of the current is less than 2A for at least 5 consecutive seconds.
[0009] In one embodiment, the blur judgment method of the stirring image is to use the stirring center position in the stirring image as the judgment area, cut the stirring image to the corresponding position, and judge whether there is blur in the stirring image due to diffused material according to the grayscale mean of the stirring image after detection; the deblurring processing method of the stirring image is to perform histogram equalization processing on the stirring image after detection to reduce the image difference caused by lighting.
[0010] In one embodiment, one cycle of the mixing video of the concrete in a uniform mixing state is 80 frames of images.
[0011] In one embodiment, 40 frames of images are selected from the stirring video every other frame among the 80 frames of images, and the selected 40-frame image sequence is input into the three-dimensional convolutional network model of ResNet. The three-dimensional convolutional network in the three-dimensional convolutional network model of ResNet extracts the time and space dimension information in the image sequence to obtain a one-dimensional feature sequence, and a fully connected layer is used to map the one-dimensional feature sequence to a certain value, which is the predicted slump when the stirring is uniformly state.
[0012] The advantages or beneficial effects of the above technical solution include at least: by using the concrete slump prediction method of the present invention to predict the slump of concrete, the slump loss can be prevented in advance according to factors such as transportation distance and weather, and the working performance of the concrete mixing process can be effectively predicted in real time inside the mixer and at the discharge port. It has strong anti-interference ability, and through data analysis and feedback, it ensures the consistency, fluidity and plasticity of concrete, improves product quality, and reduces production waste; at the same time, it combines the monitoring of flow rate and slump at the construction site, which can provide construction personnel with important data on the fluidity of concrete, so as to adjust the fluidity of concrete in time and improve the work efficiency of the construction party. At the same time, the slump index is also directly related to the working performance of concrete during construction, so that the construction unit can ensure that the design performance requirements are met.
[0013] In a second aspect, the present invention provides a concrete slump prediction device based on image and flow velocity estimation, comprising light source one, camera one, a mixer, camera two, light source two, a control box, a current clamp, an acquisition card, a video recorder, a computer, a discharge chute, camera three, and a filling funnel of a mixer truck, wherein the light source one and camera one are arranged above the mixer, the camera two and light source two are arranged below the discharge port of the mixer, the control box is used to control the mixer, the control box is connected to the computer through the current clamp and the acquisition card, the discharge chute is arranged at the discharge end of the filling funnel of the mixer truck, the camera three is arranged on one side of the discharge chute, the camera one, camera two and camera three are respectively connected to the computer through a video recorder, and the computer is used to execute the above-mentioned concrete slump prediction method based on image and flow velocity estimation.
[0014] The advantages or beneficial effects of the above technical solution include at least: by setting light source one and light source two, a stable lighting environment can be provided for camera one and camera two to improve the ability of camera one and camera two to capture information on the concrete surface; camera one is used to shoot the mixing video of concrete in the mixer; camera two is used to shoot the unloading video of the mixer discharge port; camera three is used to shoot the unloading video of the mixer truck unloaded from the unloading chute; the acquisition card is used to obtain the mixing current data in the control box, so that the computer can execute the above concrete slump prediction method. BRIEF DESCRIPTION OF THE DRAWINGS
[0015] The accompanying drawings illustrate exemplary embodiments of the present invention and together with the description serve to explain the principles of the present invention. These drawings are included to provide a further understanding of the present invention and are incorporated in and constitute a part of this specification.
[0016] Figure 1 A schematic diagram of a concrete slump prediction device according to an exemplary embodiment of the present invention is shown; Figure 2 A schematic diagram of a current curve during a concrete mixing process according to an exemplary embodiment of the present invention is shown; Figure 3 A schematic diagram of a three-dimensional convolutional network model according to an exemplary embodiment of the present invention is shown; Figure 4 A schematic diagram of concrete flow velocity measurement based on a particle image velocimetry algorithm according to an exemplary embodiment of the present invention is shown; Figure 5 A schematic diagram of measuring the concrete flow rate when a mixer truck is unloading according to an exemplary embodiment of the present invention is shown.
[0017] Description of reference numerals: 1. Light source one; 2. Camera one; 3. Mixer; 4. Camera two; 5. Light source two; 6. Control box; 7. Current clamp; 8. Capture card; 9. Video recorder; 10. Computer; 11. Discharge chute; 12. Camera three; 13. Mixer truck filling funnel. DETAILED DESCRIPTION
[0018] Embodiments of the present invention will be described in more detail below with reference to the accompanying drawings. Although certain embodiments of the present invention are shown in the accompanying drawings, it should be understood that the present invention can be implemented in various forms and should not be construed as being limited to the embodiments described herein, which are instead provided for a more thorough and complete understanding of the present invention. It should be understood that the drawings and embodiments of the present invention are only for exemplary purposes and are not intended to limit the scope of protection of the present invention.
[0019] It should be noted that, in the absence of conflict, the embodiments and features of the embodiments of the present invention can be combined with each other. The present invention will be described in detail below with reference to the accompanying drawings and in combination with the embodiments.
[0020] It should be understood that the various steps described in the method embodiments of the present invention may be performed in different orders and / or in parallel. In addition, the method embodiments may include additional steps and / or omit the steps shown. The scope of the present invention is not limited in this respect.
[0021] The term "including" and its variations used in this document are open inclusions, that is, "including but not limited to". The term "based on" means "based at least in part on". The term "one embodiment" means "at least one embodiment"; the term "another embodiment" means "at least one other embodiment"; the term "some embodiments" means "at least some embodiments". Relevant definitions of other terms will be given in the following description. It should be noted that the concepts of "first", "second", etc. mentioned in the present invention are only used to distinguish different devices, modules or units, and are not used to limit the order or interdependence of the functions performed by these devices, modules or units.
[0022] It should be noted that the modifications of "one" and "plurality" mentioned in the present invention are illustrative rather than restrictive, and those skilled in the art should understand that, unless otherwise clearly indicated in the context, it should be understood as "one or more".
[0023] The names of the messages or information exchanged between multiple devices in the embodiments of the present invention are only used for illustrative purposes, and are not used to limit the scope of these messages or information.
[0024] See also Figures 2 to 5 The embodiment of the present invention provides a method for predicting concrete slump based on image and flow velocity estimation, comprising the following steps: S1. Predict the slump loss value of concrete during transportation, obtain the transportation parameters of previous concrete, fit the regression relationship between the slump loss value and the transportation distance, the humidity and temperature on the day of transportation through machine learning methods, input the actual transportation parameters into the relationship, obtain the slump loss value of concrete, and adjust the concrete formula according to the slump loss value of concrete; by adjusting the formula of concrete before production through the slump loss value, the produced concrete can meet the construction requirements when it is transported to the construction site; The transportation parameters of concrete include but are not limited to the transportation distance, humidity and temperature on the day of transportation; S2. Obtaining the mixer current data to determine the mixing state of the concrete, obtaining the current curve data collected during the concrete mixing process, analyzing the relationship between the current curve data and the mixing state, and judging the actual mixing state of the concrete based on the relationship between the current curve data and the mixing state; Among them, the time period before the current curve reaches the peak in the current curve data is the feeding stage. In this stage, the friction between the concrete particles is large, the load of the mixer gradually increases, and the current rises until it reaches the peak; the time period from the peak to the gradual decrease of the current curve is the mixing stage. In this stage, cement hydration wraps the aggregate, reducing the friction between the particles, and the current gradually decreases; the time period when the current curve remains unchanged or changes slightly for a certain period of time is the mixing uniformity stage, such as Figure 2 As shown; When the current curve is in the feeding stage, the mixing state of the concrete is loose; When the current curve is in the mixing stage, the mixing state of the concrete is the mixed state; When the current curve is in the uniform mixing stage, the mixing state of the concrete is the uniform mixing state.
[0025] The method for judging whether the current curve remains unchanged or changes slightly is that the difference between the maximum and minimum current values is less than 2A for at least 5 consecutive seconds.
[0026] S3. Obtaining and preprocessing the mixing video of the concrete in the mixer, obtaining the mixing video collected when the concrete mixing state is a uniform mixing state, outputting it as a continuous mixing image at the original frame rate, and performing fuzzy judgment and defuzzification processing on the continuous mixing image to obtain a preprocessed mixing image sequence number set; The method for determining blur of the stirring image is to use the stirring center position in the stirring image as the determination area, cut the stirring image to the corresponding position, and determine whether there is blur in the stirring image caused by diffused material according to the grayscale mean value of the sheared stirring image. The deblurring method of the stirred image is to perform histogram equalization on the detected stirred image to reduce the image difference caused by illumination; One cycle of the mixing video when the concrete is mixed evenly is 80 frames of images.
[0027] S4. Predicting the slump of concrete in the mixer, obtaining image information of a preprocessed mixing image sequence number set, establishing a three-dimensional convolutional network model based on ResNet according to the image information, obtaining the relationship between the mixing video and the slump of the concrete when it is mixed evenly, and inputting the mixing video of the actual concrete when it is mixed evenly into the three-dimensional convolutional network model to obtain a predicted slump when it is mixed evenly; Among them, one cycle of the mixing video when the concrete is mixed evenly is 80 frames of images; 40 frames of images in the mixing video are selected every other frame from the 80 frames, and the selected 40-frame image sequence is input into the three-dimensional convolutional network model of ResNet. The three-dimensional convolutional network in the three-dimensional convolutional network model of ResNet extracts the time and space dimension information in the image sequence to obtain a one-dimensional feature sequence, and the one-dimensional feature sequence is mapped to a certain value using a fully connected layer. The value is the predicted slump when the concrete is mixed evenly, as shown in Figure 3 shown.
[0028] S5. Obtain and preprocess the concrete discharge video of the mixer discharge port, obtain the discharge port discharge video collected when the mixer discharge port is discharged, output it as a continuous discharge port discharge image at the original frame rate, and obtain a preprocessed discharge port discharge image sequence number set; S6. Predict the slump of concrete at the mixer discharge port, obtain the preprocessed discharge port discharge image sequence data set, and calculate the velocity field distribution map between continuous images in the image sequence data set through the particle image velocimetry algorithm, such as Figure 4 As shown, the average flow velocity of concrete is calculated according to the velocity field distribution diagram, and the average flow velocity of concrete and the slump are regressed and fitted to obtain the predicted slump 2 when unloading; S7. predict the slump of concrete before it leaves the factory, obtain previous concrete order data, obtain a fully connected neural network for determining the weight of slump values through machine learning training, input the predicted slump 1 and the predicted slump 2 into the fully connected neural network for weighted decision analysis, add the weighted values of the two, and obtain the predicted slump 3 of concrete before it leaves the factory; S8. Obtaining and preprocessing the concrete unloading video of the mixer truck when unloading, obtaining the mixer truck unloading video collected when the mixer truck is unloading, outputting it as a continuous mixer truck unloading image at the original frame rate, and obtaining a preprocessed mixer truck unloading image sequence number set; S9. Predict the slump of concrete in the mixer truck unloading hopper. Obtain a preprocessed mixer truck unloading image sequence data set. Use the particle image velocimetry algorithm to calculate the velocity field distribution map between consecutive images in the image sequence data set. Calculate the average flow velocity of concrete based on the velocity field distribution map. Perform regression fitting between the average flow velocity of concrete and the slump to calculate the predicted slump of the mixer truck unloading. Figure 4 and Figure 5 As shown, judge whether it meets the use requirements based on the slump four. If it does not meet the requirements, adjust the fluidity of the concrete until it meets the use requirements.
[0029] By adopting the above method, by using the concrete slump prediction method of the present invention to predict the slump of concrete, the slump loss can be prevented in advance according to factors such as transportation distance and weather, and the working performance of the concrete mixing process can be effectively predicted in real time inside the mixer and at the discharge port at the same time, with strong anti-interference ability. Through data analysis and feedback, the consistency, fluidity and plasticity of concrete can be ensured, product quality can be improved, and production waste can be reduced; at the same time, the flow rate and slump monitoring are combined at the construction site, which can provide construction personnel with important data on the fluidity of concrete, so as to adjust the fluidity of concrete in time and improve the work efficiency of the construction party. At the same time, the slump index is also directly related to the working performance of concrete in construction, so that the construction unit can ensure that the design performance requirements are met.
[0030] See also Figures 1 to 5 The embodiment of the present invention provides a concrete slump prediction device based on image and flow velocity estimation, comprising a light source 1, a camera 2, a mixer 3, a camera 4, a light source 5, a control box 6, a current clamp 7, an acquisition card 8, a video recorder 9, a computer 10, a discharge chute 11, a camera 3 12, and a filling funnel 13 of a mixer truck. The light source 1 and the camera 2 are arranged above the mixer 1, the camera 4 and the light source 5 are arranged below the discharge port of the mixer 3, the control box 6 is used to control the mixer 3, the control box 6 is connected to the computer 10 through the current clamp 7 and the acquisition card 8, the discharge chute 11 is arranged at the discharge end of the filling funnel 13 of the mixer truck, the camera 3 12 is arranged on one side of the discharge chute 11, the camera 1 2, the camera 4 and the camera 3 12 are respectively connected to the computer 10 through the video recorder 9, and the computer 10 is used to execute the above-mentioned concrete slump prediction method based on image and flow velocity estimation. In actual application, through the setting of light source 1 and light source 2 5, a stable lighting environment can be provided for camera 1 2 and camera 2 4 to improve the ability of camera 1 2 and camera 2 4 to capture information on the concrete surface. Camera 1 2 is used to shoot the mixing video of concrete in the mixer 3, camera 2 4 is used to shoot the unloading video of the unloading port of the mixer 3, camera 3 12 is used to shoot the unloading video of the mixer truck unloaded by the unloading chute 11, and the acquisition card 8 is used to obtain the mixing current data in the control box 6.
[0031] In the description of the present invention, it should be noted that the terms "center", "up", "down", "left", "right", "vertical", "horizontal", "inside", "outside", etc. indicate directions or positional relationships based on the directions or positional relationships shown in the accompanying drawings. They are only for the convenience of describing the present invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific direction, be constructed and operated in a specific direction. Therefore, they cannot be understood as limitations on the present invention.
[0032] It should be understood by those skilled in the art that the above embodiments are only for the purpose of clearly illustrating the present invention, and are not intended to limit the scope of the present invention. For those skilled in the art, other changes or modifications may be made based on the above invention, and these changes or modifications are still within the scope of the present invention.
Claims
1. A method for predicting concrete slump based on image and flow velocity estimation, characterized in that: The following steps are included: Predict the slump loss value of concrete during transportation, obtain the transportation parameters of previous concrete, fit the regression relationship between slump loss value and transportation distance, humidity and temperature on the day of transportation through machine learning methods, input the actual transportation parameters into the relationship, obtain the slump loss value of concrete, and adjust the concrete formula according to the slump loss value of concrete; Acquire mixer current data to determine the mixing state of concrete, acquire current curve data collected during the concrete mixing process, analyze the relationship between the current curve data and the mixing state, and judge the actual mixing state of concrete based on the relationship between the current curve data and the mixing state; The mixing video of concrete in the mixer is obtained and preprocessed, and the mixing video collected when the concrete mixing state is a uniform mixing state is obtained, and the mixing video is output as a continuous mixing image at the original frame rate, and the continuous mixing image is subjected to fuzzy judgment and defuzzification processing to obtain a preprocessed mixing image sequence number set; Predict the slump of concrete in the mixer, obtain the image information of the preprocessed mixing image sequence data set, establish a three-dimensional convolutional network model based on ResNet according to the image information, obtain the relationship between the mixing video and the slump of the concrete when it is mixed evenly, and input the actual mixing video of the concrete when it is mixed evenly into the three-dimensional convolutional network model to obtain the predicted slump of the concrete when it is mixed evenly; Obtaining and preprocessing the concrete discharge video of the mixer discharge port, obtaining the discharge port discharge video collected when the mixer discharges, outputting it as a continuous discharge port discharge image at the original frame rate, and obtaining a preprocessed discharge port discharge image sequence number set; Predict the slump of concrete at the mixer discharge port, obtain the preprocessed discharge port discharge image sequence data set, calculate the velocity field distribution map between continuous images in the image sequence data set through the particle image velocimetry algorithm, calculate the average flow velocity of concrete according to the velocity field distribution map, and regress and fit the average flow velocity of concrete with the slump to obtain the predicted slump during discharge; Predict the slump of concrete when it leaves the factory, obtain previous concrete order data, and obtain a fully connected neural network for judging the weight of slump values through machine learning training. Input the predicted slump 1 and the predicted slump 2 into the fully connected neural network for weighted decision analysis, and add the weighted values of the two to obtain the predicted slump 3 of concrete when it leaves the factory. Obtaining and preprocessing the concrete unloading video when the mixer truck is unloading, obtaining the mixer truck unloading video collected when the mixer truck is unloading, outputting it as a continuous mixer truck unloading image at the original frame rate, and obtaining a preprocessed mixer truck unloading image sequence number set; Predict the slump of concrete in the mixer truck unloading hopper, obtain the preprocessed mixer truck unloading image sequence data set, calculate the velocity field distribution map between continuous images in the image sequence data set through the particle image velocimetry algorithm, calculate the average flow velocity of the concrete according to the velocity field distribution map, and regress the average flow velocity of the concrete with the slump to calculate the predicted slump four when the mixer truck is unloading, and judge whether it meets the use requirements according to the slump four. If it does not meet the requirements, adjust the fluidity of the concrete until it meets the use requirements.
2. The method for predicting concrete slump based on image and flow velocity estimation according to claim 1, characterized in that: The transportation parameters of the concrete include transportation distance, humidity and temperature on the day of transportation.
3. The method for predicting concrete slump based on image and flow velocity estimation according to claim 1, characterized in that: The time period before the current curve reaches the peak value in the current curve data is the feeding stage, the time period from the peak value to the gradual decrease of the current curve is the mixing stage, and the time period when the current curve remains unchanged or changes slightly within a certain period of time is the mixing uniformity stage; When the current curve is in the feeding stage, the mixing state of the concrete is loose; When the current curve is in the mixing stage, the mixing state of the concrete is the mixed state; When the current curve is in the uniform mixing stage, the mixing state of the concrete is the uniform mixing state.
4. The method for predicting concrete slump based on image and flow velocity estimation according to claim 3, characterized in that: The method for judging whether the current curve remains unchanged or changes slightly is that the difference between the maximum value and the minimum value of the current is less than 2A for at least 5 consecutive seconds.
5. The method for predicting concrete slump based on image and flow velocity estimation according to claim 1, characterized in that: The blur judgment method of the stirring image is to use the stirring center position in the stirring image as the judgment area, cut the stirring image to the corresponding position, and judge whether there is blur in the stirring image due to diffused material distribution according to the grayscale mean value of the stirring image after detection; the deblurring processing method of the stirring image is to perform histogram equalization processing on the stirring image after detection to reduce the image difference caused by lighting.
6. The method for predicting concrete slump based on image and flow velocity estimation according to claim 1, characterized in that: One cycle of the mixing video when the concrete is mixed evenly is 80 frames of images.
7. The method for predicting concrete slump based on image and flow velocity estimation according to claim 6, characterized in that: 40 frames of images in the stirring video are selected every other frame from the 80 frames of images, and the selected 40-frame image sequence is input into the three-dimensional convolutional network model of ResNet. The three-dimensional convolutional network in the three-dimensional convolutional network model of ResNet extracts the time and space dimension information in the image sequence to obtain a one-dimensional feature sequence, and the one-dimensional feature sequence is mapped to a certain value using a fully connected layer. The value is the predicted slump of one when the stirring is in a uniform state.
8. A concrete slump prediction device based on image and flow velocity estimation, characterized in that: The method comprises a light source 1, a camera 1, a mixer, a camera 2, a light source 2, a control box, a current clamp, an acquisition card, a video recorder, a computer, a discharge chute, a camera 3, and a filling funnel of a mixer truck. The light source 1 and the camera 1 are arranged above the mixer, the camera 2 and the light source 2 are arranged below the discharge port of the mixer, the control box is used to control the mixer, the control box is connected to the computer through the current clamp and the acquisition card, the discharge chute is arranged at the discharge end of the filling funnel of the mixer truck, the camera 3 is arranged on one side of the discharge chute, the camera 1, the camera 2 and the camera 3 are respectively connected to the computer through the video recorder, and the computer is used to execute the concrete slump prediction method based on image and flow velocity estimation as described in any one of claims 1 to 7.
Citation Information
Patent Citations
Method and device for measuring workability of concrete mixture
CN114266989A
Automatic slump monitoring method and system in concrete production
CN115343284A
Slump detection method and device, engineering machinery and equipment
CN116030386A
Slump monitoring method and device based on image sequence and readable medium
CN117218118A
Slump monitoring method and device based on multi-modal data and readable medium
CN117390586A
Cited By
Embedded-free concrete workability detection method based on deep learning
CN119905164A
Online updating method and system for concrete slump model of concrete mixer
CN120038847A
Concrete slump control system and method combining AI monitoring and ERP
CN120560023A
