Self-adaptive control method for intelligent mixing of tunnel open cut tunnel section roadbed solidified soil

Abstract

The invention discloses a self-adaptive control method for intelligent mixing of roadbed solidified soil at an open cut tunnel section of a tunnel, relates to the technical field of road roadbed engineering construction, and solves the technical problems that an existing roadbed solidified soil mixing effect evaluation method lags behind time and highly depends on manpower. According to the method, the mixing state of the road mixer can be quickly and effectively judged by utilizing the characteristic that the gray difference between the curing agent and the soil matrix is large, calculating the global gray value of the image of the mixing area and preliminarily evaluating the mixing effect, and meanwhile, the gray-scale map of the mixing area is divided into the curing agent area and the soil matrix area by adopting an OSTU binarization segmentation method, so that the mixing effect of the road mixer is improved. Based on the distribution characteristic of the image gray scale, the curing agent can be effectively separated from the soil matrix. The image is divided into a plurality of windows, the proportion of the curing agent in each window is counted, and the variance of the proportion of the curing agent in all the windows is calculated, so that the distribution uniformity of the curing agent among the windows can be effectively judged, and the mixing effect can be effectively evaluated.

Description

Technical Field

[0001] This invention relates to the field of roadbed engineering construction technology, specifically to an adaptive control method for intelligent mixing of solidified soil in tunnel sections. Background Technology

[0002] With the rapid development of infrastructure construction, especially in transportation projects such as highways and railways, the quality of subgrade construction directly affects the safety and long-term stability of the project. Currently, the main subgrade treatment methods rely on the addition of chemical curing agents, improving the subgrade strength and stability through thorough mixing and reaction between the soil and the curing agent. Especially when treating poor soils (such as expansive soils and soft soils), if the curing agent is not fully mixed with the soil, the strength and anti-expansion performance of the poor subgrade will be severely affected. To ensure the construction quality of subgrade projects, engineers need to effectively evaluate the mixing effect of the subgrade solidification soil and adjust the mixing parameters of the road mixing machine in real time.

[0003] Existing methods for evaluating the mixing effect of roadbed solidification soil are mostly based on manual measurement. These methods involve manually sampling the mixed soil and measuring the solidifying agent content and unconfined compressive strength in the laboratory to determine the mixing effect of the roadbed solidification soil. Although this evaluation method accurately measures the mixing parameters of the samples, it is limited by the randomness of soil sampling and the limitations of manual testing, making it difficult to comprehensively evaluate the construction mixing effect of roadbed solidification soil.

[0004] Furthermore, the time lag between manual assessment of the mixing effect and the actual mixing process is significant, making real-time monitoring and adjustment of mixing parameters impossible. This leads to large fluctuations in the quality of the solidified soil, necessitating repeated construction. This not only impacts project progress but also results in unnecessary resource waste and increased costs.

[0005] Image processing algorithms have significant advantages in evaluating the mixing effect of roadbed solidification soil. They can not only accurately extract the actual mixing area, but also comprehensively and effectively calculate the solidifier ratio through automated algorithms, thereby judging the mixing effect and adjusting the mixing parameters of the road mixer. Although there is a lot of research on intelligent roadbed construction, the uneven lighting inside the tunnel opening can affect the overall mixing recognition during tunnel construction, leading to unstable solidification effect and affecting the bearing capacity and durability of the roadbed. Therefore, it is necessary to make corresponding adjustments based on the actual lighting conditions inside the tunnel opening to minimize the impact of uneven lighting. At the same time, most research focuses on the intelligent paving and compaction of the roadbed, with very little attention paid to the mixing evaluation of roadbed solidification soil, and there is a lack of rapid automated evaluation and adjustment methods applicable to roadbed solidification soil mixing. Summary of the Invention

[0006] To address the shortcomings of existing technologies, this invention provides an adaptive control method for intelligent mixing of roadbed solidification soil in tunnel open-cut sections, solving the problems of time lag and high dependence on manual labor in current methods for evaluating the mixing effect of roadbed solidification soil.

[0007] To achieve the above objectives, the present invention provides the following technical solution: an adaptive control method for intelligent mixing of roadbed solidification soil in open-cut tunnel sections, the method specifically comprising the following steps: Step 1: Obtain images of the solidified soil in the roadbed after the road mixing machine, and perform perspective transformation and cropping on the images to obtain a grayscale image of the mixing area; Step 2: Calculate the global grayscale average value of the obtained grayscale image of the mixing area to preliminarily evaluate the overall mixing effect; Step 3: Using histograms Figure 2 Value-based methods were used to obtain a distribution characteristic map of the curing agent and subgrade within the mixing area; Step 4: Divide the distribution feature map into several windows, calculate the proportion of curing agent in each window, and calculate the variance of all windows. Then compare the variance with a set threshold. Step 5: Based on the preliminary and variance assessment results, adjust the road mixer parameters and record the areas with poor mixing.

[0008] As a further aspect of the present invention, the grayscale image of the mixing region is obtained as follows: Based on the shooting angle of the image acquisition system installed after the road mixer and the mixing area, four reference points for perspective transformation and image cropping are selected. Based on the reference points and the corner points of the transformed image, the homography matrix is ​​calculated. Using a reference point, the image information inside the reference point is preserved. A homography matrix is ​​applied to convert the cropped oblique image into a vertical viewpoint. The converted image is then converted to grayscale space to obtain a grayscale image of the mixing area.

[0009] As a further aspect of the present invention, the global grayscale average value is calculated as follows: Iterate through all pixels in the grayscale image and denote them as a, where a = 1, 2, ..., b, and b represents the number of pixels. Simultaneously, obtain the grayscale value Ga corresponding to pixel i, and calculate the initial sum S = and initial average value Next, calculate the standard deviation of the initial grayscale values. And remove grayscale values ​​exceeding [ -3 , +3 [Noise pixels], obtain the remaining valid pixels. And recalculate the average. = ,in The pixel grayscale value after removing outliers.

[0010] As a further aspect of the present invention, the method for initially evaluating the overall mixing effect is as follows: When the mixing is sufficient, i.e., the grayscale value is close to the global threshold, if the global average grayscale value is... If the value exceeds the global threshold, the mixing effect is considered poor. (This is followed by a seemingly unrelated statement about the global grayscale average value.) If the mixture is less than the global threshold, it is determined that the mixing is good, and a preliminary evaluation result is obtained.

[0011] As a further aspect of the present invention, the distribution feature map is obtained as follows: The grayscale histogram of the mixing area is statistically analyzed, and the proportion of pixels at each grayscale level is calculated. Then, all possible thresholds c are traversed, and the pixels are classified according to the threshold c. Pixels with grayscale values ​​≤ threshold c are classified as background pixels, and pixels with grayscale values ​​> threshold c are classified as foreground pixels. Obtain all background and foreground pixels, and calculate the inter-class variance between the two classes. = ,in The percentage of background pixels. Foreground pixel percentage and The average gray values ​​of the two classes are respectively, and the selection is... The largest c is used as the optimal binarization threshold T1. Regions with pixel values ​​> T1 in the grayscale image are set to white, and regions with pixel values ​​≤ T1 are set to black, thus generating a distribution feature map.

[0012] As a further aspect of the present invention, step four includes: Obtain all windows and label them as i, where i = 1, 2, ..., N, and N represents the number of windows. Calculate the proportion of the hardener to the total area of ​​each window, R. i =W i ÷T i , where R i W represents the proportion of the curing agent area to the i-window. i T represents the area of ​​the curing agent within window i. i Let T be the area of ​​window i. For windows with merged edges, adjust T according to the actual physical area ratio. i Next, the variance of the curing agent ratio for all windows is calculated using the following formula: = ,in The variance of the curing agent ratio for all windows is given by N, where N is the number of windows. The obtained variance Comparative analysis with a set threshold, when variance If the mixture is ≤ a set threshold, it is considered to be mixed evenly. When a threshold is set, the mixture is judged to be uneven, and the variance evaluation result is obtained.

[0013] As a further aspect of the present invention, step five includes: If the global grayscale average value is greater than the set threshold, or the local variance is greater than the set threshold, parameter adjustment will be triggered if either of the two conditions is met. If both the global and local assessments are qualified, the current road mixer parameters will be maintained and construction will continue. If both the global and local assessments are unqualified, basic adjustments will be made based on the global assessment results first, and then fine-tuning will be carried out in combination with the local variance magnitude. Analyze the trigger parameter adjustments to obtain the base speed v0, and adjust the speed based on different trigger conditions. If only a global evaluation is performed, it will be deemed unqualified. =v0x( -Global threshold) / Global threshold x 0.5, if only a local evaluation fails, =v0x( -Set threshold) / Set threshold x 0.8, if both are not up to standard, =0.6×[v0x( -Global threshold) / Global threshold x 0.5] + 0.4 × [v0x( -Set threshold) / Set threshold x 0.8], adjusted rotational speed v = v0 + v must be limited to the safe range of the equipment.

[0014] An adaptive control system for intelligent mixing of roadbed solidification soil in open-cut tunnel sections includes: The subgrade solidified soil image acquisition unit is used to acquire image data of the solidified soil after mixing; The subgrade solidified soil image preprocessing unit is used to perform perspective correction, cropping and grayscale conversion on the image of the solidified soil after mixing to obtain a grayscale image of the mixing area. The global mixing effect evaluation unit is used to preliminarily evaluate the global mixing effect of the roadbed solidification soil by comparing the global grayscale average value of the grayscale map of the mixing area with a set threshold. The local evaluation unit for mixing effect is used to obtain the distribution area of ​​curing agent by binarizing the grayscale image of the mixing area, calculating the variance of curing agent distribution within the image segmentation window, and evaluating the mixing effect. The mixing parameter adjustment module is used to adjust the road mixer parameters in real time based on the comprehensive evaluation results of the mixing effect, and to record areas with poor mixing.

[0015] A computer device includes a memory, a processor, a controller, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it implements an adaptive control method for intelligent mixing of solidified soil in the roadbed of a tunnel section.

[0016] A computer-readable storage medium storing a computer program that, when executed by a processor, implements an adaptive control method for intelligent mixing of solidified soil in the subgrade of a tunnel section.

[0017] This invention provides an adaptive control method for intelligent mixing of roadbed solidification soil in open-cut tunnel sections. Compared with existing technologies, it has the following advantages: This invention employs a reference point selection method to perform perspective transformation and cropping on images of solidified subgrade soil to obtain images of the mixing area. This method utilizes the constant angle between the image acquisition device and the subgrade, embedding the reference point into the algorithm. This allows for standardized processing of all acquired images of solidified subgrade soil, resulting in high computational efficiency. Simultaneously, converting oblique images to a vertical perspective effectively ensures the accuracy of subsequent mixing effect assessment, avoiding errors caused by perspective.

[0018] This invention utilizes the significant difference in grayscale between the curing agent and the subgrade. By calculating the global grayscale value of the mixing area image, it preliminarily assesses the mixing effect. This method is extremely fast and can effectively determine the mixing status of the road mixer. It also employs the OTU binarization segmentation method to divide the grayscale image of the mixing area into two regions: the curing agent region and the subgrade region. This method, based on the distribution characteristics of image grayscale, effectively separates the curing agent from the subgrade. Furthermore, the image is divided into several windows, and the proportion of the curing agent in each window is statistically analyzed, and the variance of the curing agent proportion across all windows is calculated. This method effectively determines the uniformity of the curing agent distribution between windows, thereby effectively evaluating the mixing effect.

[0019] This invention integrates global and local mixing effect evaluation results, thus avoiding false detections caused by a single indicator. Automated adjustment of the road mixer rotor speed based on the mixing evaluation results can improve the mixing efficiency of subgrade solidified soil and reduce repeated mixing. Simultaneously, accurately recording poorly mixed areas can further improve the quality of subgrade construction and achieve refined construction. Attached Figure Description

[0020] Figure 1 This is a flowchart of the intelligent method for mixing roadbed solidified soil according to the present invention; Figure 2 This is a system block diagram of the present invention; Figure 3 This invention provides reference points for marking images of solidified soil in roadbeds. Figure 4 This is a grayscale image of the mixing area after perspective transformation and cropping according to the present invention. Figure 5 This is a distribution map of the curing agent and the subgrade obtained by OSU binarization in this invention; Figure 6This is a graph showing the curing agent ratio calculated based on window segmentation according to the present invention. Detailed Implementation

[0021] 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.

[0022] First Embodiment Please see Figure 1 , Figures 3 to 6 This application provides an adaptive control method for intelligent mixing of roadbed solidification soil in open-cut tunnel sections, which specifically includes the following steps: Step 1: Based on the shooting angle of the image acquisition system installed behind the road mixer and the range of the mixing area, select four reference points for viewpoint transformation and image cropping. These four reference points are physical feature points located at the boundary of the mixing area behind the road mixer rotor, and must meet the following conditions: Spatial location: corresponding to the upper left, upper right, lower left, and lower right corners of the mixing area, that is, the four points of the solidified soil paving strip formed after the road mixer rotor operates, and the projection of the four points on the horizontal ground forms a rectangle, with the long side perpendicular to the road mixer's direction of travel and the short side parallel to the direction of travel. Identifiability: Set up reflective markers of preset size at the benchmark point, or automatically extract the natural boundary corners between the mixing area and the unconstructed area through image recognition algorithms, such as the boundary of color difference between the solidified soil and the original soil; Stability: The relative position of the benchmark point remains unchanged relative to the road mixer vehicle during construction, ensuring that the image acquisition system can stably identify the image every time it takes a picture; Based on the reference point and the corner points of the transformed image, the homography matrix is ​​calculated. The specific processing method is as follows: the source image coordinates (x1, y1), (x2, y2), (x3, y3), (x4, y4)) and the preset corner points of the transformed vertical view image (the target coordinates (X1, Y1), (X2, Y2), (X3, Y3), (X4, Y4)) are set. The target image size is preset to 500×300 pixels. For each reference point, [xi, yi, 1]^T=H·[Xi, Yi, 1]^T is satisfied. The linear equation system of 8 unknowns is solved by the least squares method to obtain the initial matrix. Then, the random sampling consensus algorithm is used to remove possible outliers. Based on the coordinates of the four identified reference points, a polygonal region enclosed by the four points is defined in the source image. Pixel masking technology is used to preserve image information within this region while removing irrelevant background outside the region. Specifically, this includes: Generate a binary mask of the same size as the source image, where the pixel value within the area enclosed by the reference point in the mask is 1, and the pixel value outside the area is 0; The source image is multiplied with the mask at the pixel level, and only pixels with a mask value of 1 are retained to achieve cropping; Applying the homography matrix H to perform perspective transformation on the cropped image converts an obliquely projected image into a view perpendicular to the ground. The oblique angle is typically 30° to 60° to the horizontal. The transformed image must satisfy the following conditions: Geometric consistency: The actual physical size of the solidified soil in the image is in a fixed ratio to the pixel size, ensuring the accuracy of area calculation in subsequent local assessments; No stretching distortion: The rectangular boundary of the mixing area remains right-angled in the image after conversion, avoiding misjudgment of the curing agent distribution due to viewing angle error; The converted RGB color image is converted into a grayscale image, and the grayscale value is calculated using the weighted average method according to the formula Gray=0.299×R+0.587×G+0.114×B to obtain the grayscale image of the mixing area.

[0023] Step 2: Based on the obtained grayscale image of the mixing region, traverse all pixels in the grayscale image and denote them as a, where a = 1, 2, ..., b, and b represents the number of pixels. Simultaneously, obtain the grayscale value Ga corresponding to pixel i, and calculate the initial sum S = and initial average value Next, calculate the standard deviation of the initial grayscale values. And remove grayscale values ​​exceeding [ -3 , +3 [Noise pixels], obtain the remaining valid pixels. And recalculate the average. = ,in The pixel grayscale value after removing outliers; Based on the difference in grayscale characteristics between the curing agent and the soil base, the threshold was determined through experimental calibration. In a laboratory environment, standard cured soil samples with different mixing ratios were prepared, and their vertical view grayscale images were collected. The global grayscale average value of each sample was calculated, and the maximum average value corresponding to the fully mixed state was taken as the initial benchmark threshold T0. At the same time, considering the changes in the construction environment, a light compensation coefficient k was introduced, and the value of k was 0.8-1.2. The actual application threshold T=T0·k, where k is dynamically calculated through real-time collected ambient light sensor data, k=1+0.0001x(L-5000), where L is the light intensity. When the mixture is fully mixed, the curing agent and the soil are uniformly mixed, and the grayscale value is in an intermediate state, that is, the grayscale value is close to the global threshold. If the global average grayscale value is... If the gray level is greater than the global threshold, it indicates that the hardener was not sufficiently dispersed or was added in excess, resulting in a high overall gray level and indicating poor mixing. If the global average gray level is higher... If the value is less than the global threshold, it indicates that the mixing ratio of the curing agent and the soil is reasonable and the overall distribution is balanced, indicating good mixing and obtaining preliminary evaluation results.

[0024] Step 3: Based on the acquired grayscale image of the mixing area, calculate the grayscale histogram of the mixing area, and determine the percentage of pixels at each grayscale level. Then, iterate through all possible thresholds c and classify the pixels according to the threshold c. Classify pixels with grayscale values ​​≤ threshold c as background pixels and pixels with grayscale values ​​> threshold c as foreground pixels. Obtain all background and foreground pixels and calculate the inter-class variance between the two classes. = ,in The percentage of background pixels. Foreground pixel percentage and The average gray values ​​of the two classes are respectively, and the selection is... The largest c is used as the optimal binarization threshold T1. Regions with pixel values ​​> T1 in the grayscale image are set to white, and regions with pixel values ​​≤ T1 are set to black, thus generating a distribution feature map. The binarized image is opened using a 3×3 pixel rectangular structuring element to remove isolated white noise points with an area ≤ 5 pixels. Similarly, the image is closed using a 3×3 structuring element (dilation followed by erosion) to fill black holes with an area ≤ 5 pixels, retain connected regions with an area ≥ 10 pixels, and remove small discrete regions to ensure the integrity of the curing agent and soil base regions in the feature map.

[0025] Step 4: Divide the obtained distribution feature map into windows of the same size that do not overlap. The division is based on the following criteria: the pixel size of the window is tied to construction parameters such as the operating width and mixing depth of the road mixer rotor; all windows have the same pixel size; the total number of windows N ≥ 25; the number of horizontal windows is proportional to the operating width of the road mixer; for every 1m increase in width, the number of horizontal windows increases by 2 to 3; and the actual physical size deviation is ≤ 5%. When the pixel size of the distribution feature map is not an integer multiple of the window size, perform additional edge processing steps. If the remaining pixel width of the image edge is less than 1 / 2 of the window size, the edge region is merged into the adjacent window, and the actual size of the merged window is recorded. If the remaining pixel width is greater than or equal to 1 / 2 of the window size, the region is divided into an independent window, and the deviation of its size from the standard window is corrected by interpolation. Obtain all windows and label them as i, where i = 1, 2, ..., N, and N represents the number of windows. Calculate the proportion of the hardener to the total area of ​​each window, R. i =Wi ÷T i , where R i W represents the proportion of the curing agent area to the i-window. i T represents the area of ​​the curing agent within window i. i Let T be the area of ​​window i. For windows with merged edges, adjust T according to the actual physical area ratio. i Next, the variance of the curing agent ratio for all windows is calculated using the following formula: = ,in The variance of the curing agent ratio for all windows is given by N, where N is the number of windows. Next, based on the analysis of setting thresholds using engineering standards and experimental data, in accordance with the requirement of ≥95% uniformity of solidified soil mixing in the "Technical Specification for Highway Subgrade Construction", 100 solidified soil samples that meet the standard were prepared in the laboratory, covering different soil types and solidifying agent dosages. Their variances were calculated, and the maximum value was taken as the benchmark threshold. At the same time, a correction coefficient was introduced, and the product of the benchmark threshold and the correction coefficient was further calculated to obtain the set threshold. The obtained variance Compared with the set threshold, when the mixture is uniform, the R values ​​in each window are analyzed. i The variance should fluctuate slightly around the design proportion. ≤ set threshold, when When setting a threshold, it indicates that at least 30% of the window area R... i If the deviation from the design ratio is more than ±20%, there are obvious areas of hardener aggregation or unmixed areas, which directly affect the uniformity of the roadbed strength. Therefore, it is judged as uneven mixing, and the variance evaluation result is obtained.

[0026] Step 5: Combine the preliminary and method evaluation results for comprehensive analysis. If the global grayscale average value is greater than the set threshold, or the local variance is greater than the set threshold, parameter adjustment will be triggered if either of the two conditions is met. If both the global and local evaluations are qualified, the current road mixer parameters will be maintained and construction will continue. If both the global and local evaluations are unqualified, basic adjustments will be made based on the global evaluation results first, and then fine-tuning will be carried out in combination with the local variance amplitude. Analyze the trigger parameter adjustments to obtain the base speed v0, and adjust the speed based on different trigger conditions. If only a global evaluation is performed, it will be deemed unqualified. =v0x( -Global threshold) / Global threshold x 0.5, if only a local evaluation fails, =v0x( -Set threshold) / Set threshold x 0.8, if both are not up to standard, =0.6×[v0x( -Global threshold) / Global threshold x 0.5] + 0.4 × [v0x( -Set threshold) / Set threshold x 0.8], adjusted rotational speed v = v0 + v must be limited to the safe range of the equipment.

[0027] Second Embodiment Please see Figure 2 This application provides an adaptive control system for intelligent mixing of roadbed solidification soil in open-cut tunnel sections, comprising: The subgrade solidified soil image acquisition unit is used to acquire image data of the solidified soil after mixing; The subgrade solidified soil image preprocessing unit is used to perform perspective correction, cropping and grayscale conversion on the image of the solidified soil after mixing to obtain a grayscale image of the mixing area. The global mixing effect evaluation unit is used to preliminarily evaluate the global mixing effect of the roadbed solidification soil by comparing the global grayscale average value of the grayscale map of the mixing area with a set threshold. The local evaluation unit for mixing effect is used to obtain the distribution area of ​​curing agent by binarizing the grayscale image of the mixing area, calculating the variance of curing agent distribution within the image segmentation window, and evaluating the mixing effect. The mixing parameter adjustment unit is used to adjust the road mixer parameters in real time based on the comprehensive evaluation results of the mixing effect, and to record areas with poor mixing.

[0028] Some of the data in the above formulas are numerical calculations with dimensions removed, and the contents not described in detail in this specification are all prior art known to those skilled in the art.

[0029] The above embodiments are only used to illustrate the technical methods of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical methods of the present invention without departing from the spirit and scope of the technical methods of the present invention.

Claims

1. An adaptive control method for intelligent mixing of roadbed solidification soil in open-cut tunnel sections, characterized in that, The method specifically includes the following steps: Step 1: Obtain images of the solidified soil in the roadbed after the road mixing machine, and perform perspective transformation and cropping on the images to obtain a grayscale image of the mixing area; Step 2: Calculate the global grayscale average value of the obtained grayscale image of the mixing area to preliminarily evaluate the overall mixing effect; Step 3: Use the histogram binarization method to obtain the distribution characteristics of the curing agent and the subgrade within the mixing area; Step 4: Divide the distribution feature map into several windows, calculate the proportion of curing agent in each window, and calculate the variance of all windows. Then compare the variance with a set threshold. Step 5: Based on the preliminary and variance assessment results, adjust the road mixer parameters and record the areas with poor mixing.

2. The adaptive control method for intelligent mixing of roadbed solidification soil in open-cut tunnel sections according to claim 1, characterized in that, The method for obtaining the grayscale image of the mixing area is as follows: Based on the shooting angle of the image acquisition system installed after the road mixer and the mixing area, four reference points for perspective transformation and image cropping are selected. Based on the reference points and the corner points of the transformed image, the homography matrix is ​​calculated. Using a reference point, the image information inside the reference point is preserved. A homography matrix is ​​applied to convert the cropped oblique image into a vertical viewpoint. The converted image is then converted to grayscale space to obtain a grayscale image of the mixing area.

3. The adaptive control method for intelligent mixing of roadbed solidification soil in open-cut tunnel sections according to claim 1, characterized in that, The global grayscale average value is calculated as follows: Iterate through all pixels in the grayscale image and denote them as a, where a = 1, 2, ..., b, and b represents the number of pixels. Simultaneously, obtain the grayscale value Ga corresponding to pixel i, and calculate the initial sum S = and initial average value Next, calculate the standard deviation of the initial grayscale values. And remove grayscale values ​​exceeding [ -3 , +3 [Noise pixels], obtain the remaining valid pixels. and recalculate the average. = ,in The pixel grayscale value after removing outliers.

4. The adaptive control method for intelligent mixing of roadbed solidification soil in open-cut tunnel sections according to claim 1, characterized in that, initially... The method for evaluating the overall mixing effect is as follows: When the mixing is sufficient, i.e., the grayscale value is close to the global threshold, if the global average grayscale value is... If the value exceeds the global threshold, the mixing effect is considered poor. (This is followed by a seemingly unrelated statement about the global grayscale average value.) If the mixture is less than the global threshold, it is determined that the mixing is good, and a preliminary evaluation result is obtained.

5. The adaptive control method for intelligent mixing of roadbed solidification soil in open-cut tunnel sections according to claim 1, characterized in that, The distribution feature map is obtained as follows: The grayscale histogram of the mixing area is statistically analyzed, and the proportion of pixels at each grayscale level is calculated. Then, all possible thresholds c are traversed, and the pixels are classified according to the threshold c. Pixels with grayscale values ​​≤ threshold c are classified as background pixels, and pixels with grayscale values ​​> threshold c are classified as foreground pixels. Obtain all background and foreground pixels, and calculate the inter-class variance between the two classes. = ,in The percentage of background pixels. Foreground pixel percentage and The average gray values ​​of the two classes are respectively, and the selection is... The largest c is used as the optimal binarization threshold T1. Regions with pixel values ​​> T1 in the grayscale image are set to white, and regions with pixel values ​​≤ T1 are set to black, thus generating a distribution feature map.

6. The adaptive control method for intelligent mixing of roadbed solidification soil in open-cut tunnel sections according to claim 1, characterized in that, Step four includes: Obtain all windows and label them as i, where i = 1, 2, ..., N, and N represents the number of windows. Calculate the proportion of the hardener to the total area of ​​each window, R. i =W i ÷T i , where R i W represents the proportion of the curing agent area to the i-window. i T represents the area of ​​the curing agent within window i. i Let T be the area of ​​window i. For windows with merged edges, adjust T according to the actual physical area ratio. i Next, the variance of the curing agent ratio for all windows is calculated using the following formula: = ,in The variance of the curing agent ratio for all windows is given by N, where N is the number of windows. The obtained variance Comparative analysis with a set threshold, when variance If the mixture is ≤ a set threshold, it is considered to be mixed evenly. When a threshold is set, the mixture is judged to be uneven, and the variance evaluation result is obtained.

7. The adaptive control method for intelligent mixing of roadbed solidification soil in open-cut tunnel sections according to claim 1, characterized in that, Step five includes: If the global grayscale average value is greater than the set threshold, or the local variance is greater than the set threshold, parameter adjustment will be triggered if either of the two conditions is met. If both the global and local assessments are qualified, the current road mixer parameters will be maintained and construction will continue. If both the global and local assessments are unqualified, basic adjustments will be made based on the global assessment results first, and then fine-tuning will be carried out in combination with the local variance magnitude. Analyze the trigger parameter adjustments to obtain the base speed v0, and adjust the speed based on different trigger conditions. If only a global evaluation is performed, it will be deemed unqualified. =v0x( -Global threshold) / Global threshold x 0.5, if only a local evaluation fails, =v0x( -Set threshold) / Set threshold x 0.8, if both are not up to standard, =0.6×[v0x( -Global threshold) / Global threshold x 0.5] + 0.4 × [v0x( -Set threshold) / Set threshold x 0.8], adjusted rotational speed v = v0 + v must be limited to the safe range of the equipment.

8. An adaptive control system for intelligent mixing of roadbed solidification soil in open-cut tunnel sections, used to execute the method described in any one of claims 1 to 7, characterized in that, include: The subgrade solidified soil image acquisition unit is used to acquire image data of the solidified soil after mixing; The subgrade solidified soil image preprocessing unit is used to perform perspective correction, cropping and grayscale conversion on the image of the solidified soil after mixing to obtain a grayscale image of the mixing area. The global mixing effect evaluation unit is used to preliminarily evaluate the global mixing effect of the roadbed solidification soil by comparing the global grayscale average value of the grayscale map of the mixing area with a set threshold. The local evaluation unit for mixing effect is used to obtain the distribution area of ​​curing agent by binarizing the grayscale image of the mixing area, calculating the variance of curing agent distribution within the image segmentation window, and evaluating the mixing effect. The mixing parameter adjustment module is used to adjust the road mixer parameters in real time based on the comprehensive evaluation results of the mixing effect, and to record areas with poor mixing.

9. A computer device, comprising a memory, a processor, a controller, and a computer program stored in the memory and executable on the processor, characterized in that, The processor executes a computer program to implement the method described in any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program that, when executed by a processor, implements the method according to any one of claims 1 to 7.

Images