Real-time muck weighing device and method

Abstract

The invention discloses a muck real-time weighing device and method, and relates to the technical field of engineering detection. The method comprises the steps that an RGB camera and a depth camera are installed in a slag outlet area of a shield tunneling machine, and an RGB image and a depth image of muck are collected in real time; generating a three-dimensional point cloud; non-muck noise points are identified and removed, and a muck point cloud model is constructed through filtering, point cloud registration and fusion processing; the muck volume and the muck density are obtained, the muck volume is calibrated in combination with the RGB image, the actual muck volume is obtained, and then the theoretical mass of muck is obtained; the actual weight of the muck is obtained through an array type pressure sensor; setting a threshold value, and continuously comparing the actual weight obtained in real time with the theoretical mass; and if the deviation exceeds a set threshold value, the system automatically records the timestamp, the deviation data and the corresponding image. According to the invention, real-time, accurate and non-contact monitoring of muck quality is realized, and core technical support is provided for intelligent management and control of engineering construction.

Description

Technical Field

[0001] This invention relates to the field of engineering testing technology, and in particular to a real-time weighing device and method for construction waste. Background Technology

[0002] In engineering fields such as tunnel boring, mining, and urban demolition, the transportation and disposal of construction waste consistently faces challenges such as insufficient measurement accuracy, difficulty in supervision, and significant safety hazards. Traditional methods of weighing construction waste largely rely on fixed weighbridges, which suffer from low weighing efficiency, inability to achieve real-time monitoring throughout the process, and susceptibility to human error, seriously affecting project progress control, cost accounting, and environmental compliance.

[0003] Therefore, providing a real-time weighing device and generation method for construction waste to solve the difficulties existing in the prior art is a problem that urgently needs to be solved by those skilled in the art. Summary of the Invention

[0004] In view of this, the present invention provides a real-time weighing device and method for construction waste, which realizes real-time, accurate and non-contact monitoring of construction waste quality, and provides core technical support for intelligent management and control of engineering construction.

[0005] To achieve the above objectives, the present invention adopts the following technical solution: A real-time weighing device for construction waste includes a hopper, an array of pressure sensors, an image acquisition module, and a data processing module. An array-type pressure sensor is installed at the bottom of the slag hopper to obtain the quality of the slag inside the hopper in real time; an image acquisition module is installed above the slag hopper to obtain the vehicle number in real time; and a data processing module is used to receive and process data from the array-type pressure sensor, laser module, and image acquisition module. The array-type pressure sensor consists of multiple pressure sensors, which are evenly spaced and distributed along the long side of the bottom of the slag hopper and are flush with the bottom panel of the slag hopper. The image acquisition module includes an automatic license plate recognition unit, which automatically identifies the transport group to which the dump truck belongs and its unique number within the group when the dump truck enters the weighing area.

[0006] Optionally, a 3D laser scanning instrument is also included, which is installed on the upper part of the trolley attached to the back of the tunnel boring machine to scan the volume of excavated soil in real time.

[0007] A method for real-time weighing of construction waste, applied to any of the above-mentioned real-time weighing devices for construction waste, includes the following steps: RGB cameras and depth cameras are installed in the muck discharge area of ​​the tunnel boring machine to collect RGB and depth images of the muck in real time. Input parameters into the system and combine them with the cross-sectional area of ​​the tunnel boring machine cutterhead to obtain the theoretical mass of the excavated soil based on the tunnel boring machine data. Obtain the actual weight of the excavated soil through an array of pressure sensors. Adjust the system input parameters according to the actual weight of the excavated soil to obtain the initial optimized input parameters. RGB images and depth image data are fused to generate a 3D point cloud; non-slag noise points are identified and removed, and then a slag point cloud model is constructed through filtering, point cloud registration and fusion processing. The volume and density of the construction waste are obtained based on the point cloud model of the construction waste, and the volume of the construction waste is calibrated by combining the RGB image to obtain the actual volume of the construction waste, and then the theoretical mass based on the volume of the construction waste is obtained. The input parameters after the initial optimization are optimized a second time using the theoretical mass based on the volume of the construction waste, and the mass of the construction waste is obtained in real time.

[0008] Optionally, the weight can be obtained based on the cross-sectional area of ​​the tunnel boring machine cutterhead as follows: Total weight of excavated soil per ring = Cross-sectional area of ​​cutterhead × Ring width × Density of excavated soil in the stratum, expressed as: T=πR 2 Δhρ; Among them, R, Δh, and ρ are all manually input through the tunnel boring machine system. Δh is obtained by subtracting the cylinder stroke from the previous time point from the current cylinder stroke.

[0009] Optionally, a dual-modal feature fusion network is used to generate a 3D point cloud, and a deep learning-based semantic segmentation model is set to identify and remove noise points.

[0010] Optional, calibrating the volume of excavated soil includes: Simultaneously acquire NIR and Red images of the same waste soil area, and ensure that the two images are registered at the pixel level; Within the target area of ​​the construction waste, the ratio of its average gray value in the near-infrared band to its average gray value in the red band is calculated, and the ratio is defined as the moisture index (SMI) of the construction waste. Establish a calibration curve between SMI and moisture content ω. The higher the moisture content ω of the slag soil, the greater the decrease in the average gray value of the near-infrared band than the average gray value of the red band, which leads to a decrease in the SMI index. ω is determined by the SMI obtained through detection, and then the calibration factor is determined. The volume of slag and soil is calibrated based on the calibration factor.

[0011] Optionally, continuous comparison also includes: setting a sliding time window to detect the average deviation and trend analysis between the actual weight and the theoretical mass, and identifying hidden anomalies.

[0012] As can be seen from the above technical solution, compared with the prior art, the present invention provides a real-time weighing device and method for construction waste, which has the following beneficial effects: 1) The present invention greatly improves the detection accuracy by working in concert with an array-type pressure sensor and a laser rangefinder, combined with point cloud model construction and volume calibration technology based on deep learning; 2) The present invention has low system latency, requires no manual intervention throughout the process, and can realize real-time monitoring and feedback of construction waste quality, significantly improving weighing efficiency. Attached Figure Description

[0013] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on the provided drawings without creative effort.

[0014] Figure 1 This is a flowchart of a real-time weighing method for construction waste disclosed in this invention. Detailed Implementation

[0015] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0016] This invention discloses a real-time weighing device for construction waste, comprising a hopper, an array of pressure sensors, an image acquisition module, and a data processing module. An array-type pressure sensor is installed at the bottom of the slag hopper to obtain the quality of the slag inside the hopper in real time; an image acquisition module is installed above the slag hopper to obtain the vehicle number in real time; and a data processing module is used to receive and process data from the array-type pressure sensor, laser module, and image acquisition module. The array-type pressure sensor consists of multiple pressure sensors, which are evenly spaced and distributed along the long side of the bottom of the slag hopper and are flush with the bottom panel of the slag hopper. The image acquisition module includes an automatic license plate recognition unit, which automatically identifies the transport group to which the dump truck belongs and its unique number within the group when the dump truck enters the weighing area.

[0017] Furthermore, the slag hopper adopts an inverted trapezoidal structure, wider at the top and narrower at the bottom, facilitating rapid loading and unloading of slag. Anti-slip protrusions are also installed at the bottom of the hopper to ensure a tight fit between the array-type pressure sensors and the bottom of the hopper, preventing sensor displacement due to vibration during transportation.

[0018] Furthermore, the array-type pressure sensor consists of multiple pressure sensors, which are evenly spaced and distributed along the long side of the bottom of the slag hopper. This ensures that the sensor detection range can completely cover the bottom area of ​​the slag hopper without any blind spots. The sensor is installed with an embedded design, flush with the bottom panel of the slag hopper, to avoid slag accumulation or uneven pressure distribution caused by sensor protrusion. At the same time, each pressure sensor is equipped with an independent signal amplification module and temperature compensation unit to offset the impact of changes in the construction environment temperature on measurement accuracy. The image acquisition module includes an automatic license plate recognition unit. When a dump truck enters the weighing area, it automatically identifies its transport group and its unique number within that group. The automatic license plate recognition unit first preprocesses the image, using edge detection and morphological processing algorithms to accurately locate the license plate area. Next, it performs character segmentation on the located license plate image, extracting each character individually. Finally, it uses a deep learning-based character recognition algorithm to recognize the segmented characters and obtain the vehicle's license plate number. Simultaneously, the automatic license plate recognition unit also has a transport group recognition function. By recognizing the group markings painted on the vehicle body or reading the information from the vehicle's built-in RFID tag, it automatically identifies the transport group to which the truck belongs and its unique number within that group, achieving comprehensive vehicle information collection.

[0019] Furthermore, it also includes a 3D laser scanning instrument, installed on the upper part of the trolley attached to the back of the tunnel boring machine, to scan the volume of excavated soil in real time.

[0020] Furthermore, the image acquisition module employs a combination of wireless (4G / 5G) and wired communication for data transmission, ensuring that vehicle information can be transmitted to the data processing module in real time and stably. The module also features local data storage capabilities; in areas with poor network signals at the construction site, it stores recent vehicle identification information and automatically uploads it once the network is restored, preventing data loss.

[0021] A method for real-time weighing of construction waste, applied to any of the aforementioned real-time weighing devices for construction waste, such as... Figure 1 As shown, it includes the following steps: RGB cameras and depth cameras are installed in the muck discharge area of ​​the tunnel boring machine to collect RGB and depth images of the muck in real time. Input parameters into the system and combine them with the cross-sectional area of ​​the tunnel boring machine cutterhead to obtain the theoretical mass of the excavated soil based on the tunnel boring machine data. Obtain the actual weight of the excavated soil through an array of pressure sensors. Adjust the system input parameters according to the actual weight of the excavated soil to obtain the initial optimized input parameters. RGB images and depth image data are fused to generate a 3D point cloud; non-slag noise points are identified and removed, and then a slag point cloud model is constructed through filtering, point cloud registration and fusion processing. The volume and density of the construction waste are obtained based on the point cloud model of the construction waste, and the volume of the construction waste is calibrated by combining the RGB image to obtain the actual volume of the construction waste, and then the theoretical mass based on the volume of the construction waste is obtained. The input parameters after the initial optimization are optimized a second time using the theoretical mass based on the volume of the construction waste, and the mass of the construction waste is obtained in real time.

[0022] Furthermore, based on the cross-sectional area of ​​the tunnel boring machine cutterhead, the weight of the excavated soil per ring is calculated as: Total weight of excavated soil per ring = Cross-sectional area of ​​cutterhead × Ring width × Density of excavated soil in the strata. The expression is: T=πR 2 Δhρ; Among them, R, Δh, and ρ are all manually input through the tunnel boring machine system. Δh is obtained by subtracting the cylinder stroke from the previous time point from the current cylinder stroke.

[0023] Furthermore, a dual-modal feature fusion network is used to generate a 3D point cloud, and a deep learning-based semantic segmentation model is set up to identify and remove noise points.

[0024] Specifically, the semantic segmentation model adopts the U-Net network architecture. It takes the generated 3D point cloud data as input, extracts and reduces the dimensions of the input data through the encoder, and then upsamples and restores the features through the decoder, finally outputting the semantic segmentation result of the point cloud.

[0025] Furthermore, the filtering process employs a Gaussian filtering algorithm, which eliminates random noise in the point cloud data by performing a weighted average of the neighborhood points around each point cloud, making the point cloud surface smoother. Point cloud registration uses the ICP algorithm to align point cloud data collected at different times to a unified world coordinate system, ensuring the consistency of the point cloud model. Point cloud fusion is the process of fusing registered multi-frame point cloud data to fill in the gaps and missing areas in a single frame of point cloud data, thereby constructing a complete and accurate point cloud model of construction waste.

[0026] Furthermore, calibrating the volume of excavated soil includes: NIR and Red images of the same slag area are acquired simultaneously. The NIR image (near-infrared image) can reflect the moisture content information of the slag, while the Red image (red band image) is used as a reference benchmark to eliminate the influence of changes in lighting conditions on the measurement results and ensure that the two images are registered at the pixel level. Within the target area of ​​the construction waste, the ratio of its average gray value in the near-infrared band to the average gray value in the red band is calculated, including: selecting multiple target areas of construction waste in the registered image, calculating the average gray value of the NIR image and the Red image in each area, calculating the SMI value of each area, and taking the average value of the SMI values ​​of all areas as the final SMI measurement result, and defining the ratio as the construction waste moisture index SMI; Establish a calibration curve between SMI and moisture content ω. The higher the moisture content ω of the slag soil, the greater the decrease in the average gray value of the near-infrared band than the average gray value of the red band, which leads to a decrease in the SMI index. ω is determined by the SMI obtained through detection, and then the calibration factor is determined. The initial volume calculated using the point cloud model The volume of slag is calibrated by multiplying it by the calibration factor k.

[0027] Furthermore, continuous comparison also includes: setting a sliding time window to detect the average deviation and trend analysis between the actual weight and the theoretical mass, and identifying hidden anomalies.

[0028] Specifically, latent anomalies refer to situations where the weight deviation does not exceed the set threshold in a single instance, but the deviation shows a significant upward or downward trend over a period of time. Such anomalies are usually caused by factors such as sensor performance degradation and gradual changes in the characteristics of the slag. If not detected in time, they may gradually develop into serious anomalies, affecting weighing accuracy.

[0029] The above description of the disclosed embodiments enables those skilled in the art to make or use the invention. Various modifications to these embodiments will be readily apparent to those 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 invention. Therefore, the invention is not to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A real-time weighing device for construction waste, characterized in that, It includes a slag hopper, an array of pressure sensors, an image acquisition module, and a data processing module. An array-type pressure sensor is installed at the bottom of the slag hopper to obtain the quality of the slag inside the hopper in real time; an image acquisition module is installed above the slag hopper to obtain the vehicle number in real time; and a data processing module is used to receive and process data from the array-type pressure sensor, laser module, and image acquisition module. The array-type pressure sensor consists of multiple pressure sensors, which are evenly spaced and distributed along the long side of the bottom of the slag hopper and are flush with the bottom panel of the slag hopper. The image acquisition module includes an automatic license plate recognition unit, which automatically identifies the transport group to which the dump truck belongs and its unique number within the group when the dump truck enters the weighing area.

2. The real-time weighing device for construction waste according to claim 1, characterized in that, It also includes a 3D laser scanning instrument, which is installed on the upper part of the trolley behind the tunnel boring machine to scan the volume of excavated soil in real time.

3. A method for real-time weighing of construction waste, applied to the real-time weighing device for construction waste described in any one of claims 1-2, characterized in that, Includes the following steps: RGB cameras and depth cameras are installed in the muck discharge area of ​​the tunnel boring machine to collect RGB and depth images of the muck in real time. Input parameters into the system and combine them with the cross-sectional area of ​​the tunnel boring machine cutterhead to obtain the theoretical mass of the excavated soil based on the tunnel boring machine data. Obtain the actual weight of the excavated soil through an array of pressure sensors. Adjust the system input parameters according to the actual weight of the excavated soil to obtain the initial optimized input parameters. RGB images and depth image data are fused to generate a 3D point cloud; Non-slag noise points are identified and eliminated, and then a slag point cloud model is constructed through filtering, point cloud registration and fusion processing. The volume and density of the construction waste are obtained based on the point cloud model of the construction waste, and the volume of the construction waste is calibrated by combining the RGB image to obtain the actual volume of the construction waste, and then the theoretical mass based on the volume of the construction waste is obtained. The input parameters after the initial optimization are optimized a second time using the theoretical mass based on the volume of the construction waste, and the mass of the construction waste is obtained in real time.

4. The method for real-time weighing of construction waste according to claim 3, characterized in that, The weight of the excavated soil per ring is calculated based on the cross-sectional area of ​​the tunnel boring machine cutterhead. The formula is: Total weight of excavated soil per ring = Cross-sectional area of ​​cutterhead × Ring width × Density of excavated soil in the strata. T=πR 2 Δhρ; Among them, R, Δh, and ρ are all manually input through the tunnel boring machine system. Δh is obtained by subtracting the cylinder stroke from the previous time point from the current cylinder stroke.

5. The method for real-time weighing of construction waste according to claim 3, characterized in that, A dual-modal feature fusion network is used to generate 3D point clouds, and a deep learning-based semantic segmentation model is set up to identify and remove noise points.

6. The method for real-time weighing of construction waste according to claim 3, characterized in that, The calibration of the slag volume includes: Simultaneously acquire NIR and Red images of the same waste soil area, and ensure that the two images are registered at the pixel level; Within the target area of ​​the construction waste, the ratio of its average gray value in the near-infrared band to its average gray value in the red band is calculated, and the ratio is defined as the moisture index (SMI) of the construction waste. Establish a calibration curve between SMI and moisture content ω. The higher the moisture content ω of the slag soil, the greater the decrease in the average gray value of the near-infrared band than the average gray value of the red band, which leads to a decrease in the SMI index. ω is determined by the SMI obtained through detection, and then the calibration factor is determined. The volume of slag and soil is calibrated based on the calibration factor.

7. The method for real-time weighing of construction waste according to claim 3, characterized in that, Continuous comparison also includes: setting a sliding time window to detect the average deviation and trend analysis between the actual weight and the theoretical mass, and identifying hidden anomalies.

Images