Aggregate asphalt adhesiveness rotation tester and adhesiveness evaluation method thereof

Abstract

The invention relates to an aggregate asphalt adhesiveness rotation tester and an adhesiveness evaluation method thereof. At present, in the processing process of manually dividing three-channel thresholds of different aggregates in an RGB color space, the accuracy is poor, and the operation process is tedious; a testing assembly is arranged in an inner cavity, a first supporting plate and a second supporting plate are arranged on a bottom plate, a transmission chain wheel assembly is arranged on the first supporting plate, a plurality of roller bodies are arranged between the first supporting plate and the second supporting plate, one end of each roller body is connected with the transmission chain wheel assembly, and the other end of each roller body is hinged to the second supporting plate. Each roller body is driven by the transmission chain wheel assembly to rotate between the first supporting plate and the second supporting plate, and a clamping gap matched with a test piece test bottle is formed between every two adjacent roller bodies; according to the aggregate asphalt adhesion performance evaluation method, image data collected by the aggregate asphalt adhesion rotation tester is used as basic data, and the basic data is subjected to the quantitative evaluation process of the aggregate asphalt adhesion performance.

Description

Technical Field

[0001] The present invention specifically relates to an aggregate asphalt adhesion rotation tester and an adhesion performance evaluation method thereof. Background Art

[0002] At present, the common evaluation methods for the adhesion between aggregate and asphalt in my country include those proposed in the Test Procedure for Asphalt and Asphalt Mixtures for Highway Engineering (JTG E20-2011): the boiling method applicable to aggregates with a maximum particle size greater than 13.2 mm; the immersion method applicable to aggregates with a maximum particle size less than or equal to 13.2 mm. In contrast, the rotating bottle method has the advantages of dynamic simulation of actual working conditions, good repeatability, controllable test conditions, wide application range, simplicity and speed. According to the EU standard EN 12697-11A, the EU countries generally use the rotating bottle method to evaluate the adhesion between aggregate and asphalt. According to the Test Procedure for Asphalt and Asphalt Mixtures for Highway Engineering (Draft for Comments), it is recommended to use the rotating bottle method to evaluate the adhesion between asphalt and aggregate (T 0675-201X). It can be seen that the industry has a high degree of recognition for the rotating bottle method.

[0003] At present, the traditional evaluation standard of asphalt adhesion by the rotating bottle method is that two experimenters use visual inspection to evaluate the coverage of asphalt, with an accuracy of 5%. This method has the problem of large errors and insufficient accuracy. At present, the improvement of it in the academic community is mainly to take pictures of the results and manually divide the thresholds of the three channels in the RGB color space for different aggregates. The specific problems are mainly:

[0004] First, the accuracy is poor: the same type of aggregate has different colors, and the background, aggregate, and asphalt are not well distinguished. The specific interference factors are shadows and water. Shadows and asphalt are both black, which will reduce the accuracy during recognition. When the aggregate is taken out of the water, it is inevitable that it will be stained with water, and the water will reflect light and affect the results. In addition, some aggregates are similar in color to asphalt, which will also cause an impact;

[0005] Secondly, a threshold value needs to be set in advance for each type of aggregate, which is a cumbersome process and it is difficult to eliminate redundant steps.

[0006] In short, the accuracy of visual inspection for asphalt coverage is poor, and there is a lack of standardized and unified evaluation methods for asphalt coverage. Summary of the invention

[0007] In order to overcome the defects of the prior art, an aggregate asphalt adhesion rotation tester and an adhesion performance evaluation method thereof are provided to solve the above problems.

[0008] An aggregate asphalt adhesion rotation tester comprises a bottom plate, a shell and a test assembly, wherein the shell is arranged on the bottom plate, an inner cavity is formed between the shell and the bottom plate, and the test assembly is arranged in the inner cavity;

[0009] The test assembly includes a first support plate, a second support plate, a plurality of rollers, a transmission sprocket assembly and a driving member. The first support plate and the second support plate are vertically arranged in parallel on the bottom plate. The transmission sprocket assembly is arranged on the first support plate, and the transmission sprocket assembly is connected to the driving member. The plurality of rollers are arranged between the first support plate and the second support plate. One end of each roller passes through the first support plate and is connected to the transmission sprocket assembly. The other end of each roller is hinged to the second support plate. Each roller rotates between the first support plate and the second support plate under the drive of the transmission sprocket assembly. A clamping gap that matches the test bottle of the specimen is formed between every two adjacent rollers.

[0010] As a preferred solution: a first opening is processed on the shell, and an opening and closing cover is hinged on the first opening.

[0011] As a preferred solution: a second opening is processed on the shell, and the second opening is provided with a visual window.

[0012] As a preferred solution: the transmission sprocket assembly includes a chain, a transition gear, a driving gear and multiple passive gears, the passive gears are arranged in one-to-one correspondence with the roller bodies, each passive gear is mounted on its corresponding roller body, one end of the roller body passes through the first support plate and is connected to its corresponding passive gear, a transition gear is arranged on the outer wall of the first support plate, the driving gear is mounted on the power output shaft of the driving member, and a chain is arranged between the driving gear, the transition gear and the multiple passive gears.

[0013] As a preferred solution: a strip-shaped support plate is arranged on the outer wall of the first support plate along its length direction, and a chain is arranged on the strip-shaped support plate.

[0014] As a preferred solution: it also includes a multi-position irradiation shadowless lamp assembly, which is arranged on the inner wall of the shell, and the multi-position irradiation shadowless lamp assembly is cooperated with a paving board, which is arranged directly below the multi-position irradiation shadowless lamp assembly, and the paving board is arranged on the bottom plate. The multi-position irradiation shadowless lamp assembly includes a main support plate, a plurality of bar tracks and a plurality of self-moving light sources. The main support plate is provided with a plurality of bar tracks, and each bar track is provided with a number of self-moving light sources; each self-moving light source includes a support frame, a motor, a gear and a light source, and the support frame is arranged on its corresponding bar track, one side of the support frame is slidably matched with one side of the bar track, and the other side of the support frame is provided with a motor, and a gear is mounted on the output shaft of the motor, and the gear is meshed with the bar track, and a light source is hinged at the bottom of the support frame, and the light source is emitted in a downward direction. Driven by the support frame, the light source reciprocates along the length direction of the bar track.

[0015] As a preferred solution: it also includes a multi-position irradiation shadowless lamp assembly, which is arranged outside the shell, and the multi-position irradiation shadowless lamp assembly includes a main support plate, multiple strip tracks and multiple self-moving light sources. The bottom of the main support plate is provided with multiple strip tracks, and each strip track is provided with several self-moving light sources; each self-moving light source includes a support frame, a motor, a gear and a light source, and the support frame is arranged on its corresponding strip track, one side of the support frame is slidably matched with one side of the strip track, and the other side of the support frame is provided with a motor, and a gear is mounted on the output shaft of the motor, and the gear is meshed with the strip track, and the light source is hinged at the bottom of the support frame, and the light source is emitted in a downward direction. Driven by the support frame, the light source reciprocates along the length direction of the strip track.

[0016] As a preferred solution: a polyester reflective film sheet is provided at the bottom of the main support plate, a plurality of strip tracks are arranged in an array on the polyester reflective film sheet, and an embedded camera is provided at the center of the bottom of the main support plate.

[0017] A method for evaluating the adhesion performance of aggregate asphalt is implemented by using the above-mentioned aggregate asphalt adhesion rotation tester. The method for evaluating the adhesion performance of aggregate asphalt is to use the image data collected by the aggregate asphalt adhesion rotation tester as basic data, and to conduct a quantitative evaluation process of the adhesion performance of aggregate asphalt on the basic data, specifically:

[0018] The quantitative evaluation process through basic data is to form a training set with the image data obtained by the aggregate asphalt adhesion rotation tester. A predetermined number of images are randomly cut out from the training set to form a classification. The class center point of each category is defined as the arithmetic mean of the pixel coordinates of all pixels in the category. The calculation formula of the arithmetic mean is:

[0019]

[0020] In the above formula, (x i ,y i ) is the coordinate of the i-th pixel in category C, N C is the number of pixels in category C; a weight coefficient w is introduced in each category, and the sampling probability of each category is adjusted according to the weight coefficient w. The sampling probability P of each category is calculated according to its corresponding weight coefficient w. The calculation formula of the sampling probability P is:

[0021]

[0022] According to the formula of the sampling area A of each category, the size of the sampling area is calculated by the weight coefficient w and the area of ​​the category area. The size formula of the sampling area of ​​each category is:

[0023] A 沥青 =P沥青 A 沥青 =0.5333A 沥青 ;

[0024] A 集料 =P 集料 A 集料 =0.4A 集料 ;

[0025] A 背景 =P 背景 A 背景 =0.0667A 背景 ;

[0026] The loss function calculation process is beneficial to improve the segmentation accuracy:

[0027] The local contrast loss is introduced in the atrous space pyramid pooling module. The key weight coefficients α = 0.23, β = 0.36, γ = 0.41, τ = 0.07 are obtained in the experiment by combining the cross entropy loss with the DICE loss. The loss function is obtained based on the key weight coefficients α = 0.23, β = 0.36, γ = 0.41, τ = 0.07. The loss function formula is:

[0028]

[0029] In the above formula, is the total loss function, is the cross entropy loss function, is the DICE loss function; is the local contrast loss function; τ is a hyperparameter that controls the smoothness of feature similarity distribution in contrastive learning; N represents the total number of pixels in the image, C represents the number of categories, including the number of backgrounds, the number of asphalts, and the number of aggregates, and y i,c Indicates whether pixel i belongs to category c, p i,c represents the predicted probability that pixel i belongs to category c; p i represents the model's predicted probability for pixel i, y i represents the actual label of pixel i; z i and z j is a positive sample pair of feature vectors, z k is the negative sample feature vector, sim(z i ,z j ) is the cosine similarity;

[0030] The quantitative evaluation conclusion of aggregate asphalt adhesion performance is obtained through the calculation of loss function: when the task and data scale are determined, When it is lower than 0.2, it indicates that the model is in a good fitting state, indicating that the asphalt coverage is in a good state, and the corresponding aggregate asphalt adhesion performance is good; when If it is still greater than 1.0 after twenty rounds of iterative training, it usually means that the model has not converged or is underfitting, indicating that the asphalt coverage is in a poor state and the corresponding aggregate asphalt adhesion performance is poor. In the future, it will be necessary to increase the number of training rounds or adjust the network structure and hyperparameters to re-acquire the basic data for calculation.

[0031] As a preferred solution: the quantitative evaluation process of aggregate asphalt adhesion performance using basic data also includes the calculation process of inter-category separation index and intra-category aggregation index. The specific calculation process is:

[0032] Inter-class separation index

[0033] Intra-class clustering index

[0034] In the above formula, and are the centroids of categories c1 and c2, x i is the feature vector of the i-th sample in the category;

[0035] Through multiple verifications of the model, it was found that the threshold of the inter-category separability index is 2.0, and the threshold of the intra-category aggregation index is 0.8; when the inter-category separability index is greater than 2.0 and the intra-category aggregation index is less than 0.8, it means that the features of different categories have a high degree of distinction in space, and the distribution of features within the same category is compact, which is conducive to the model's accurate prediction of the residual asphalt coverage.

[0036] The beneficial effects of the present invention are:

[0037] 1. The aggregate asphalt adhesion rotation tester of the present invention can simulate the dynamic effects under actual road conditions through the cooperation between the base plate, the shell and the test components, and realize the accurate evaluation of the adhesion between aggregate and asphalt. At the same time, it can also dynamically simulate different working conditions, which can be beneficial to the subsequent accurate testing of the degree of asphalt peeling on the aggregate surface, and provide favorable image data conditions for the subsequent continuous and accurate evaluation of the coverage of aggregate asphalt.

[0038] 2. The aggregate asphalt adhesion rotary tester of the present invention has good repeatability, controllable test conditions, wide application range, simplicity and speed, which is conducive to the accurate and rapid acquisition of a large number of concentrated image data, realizing large-scale evaluation, and facilitating the subsequent comprehensive and reliable evaluation results of aggregate asphalt coverage and adhesion performance.

[0039] 3. The evaluation standard formed by the aggregate asphalt adhesion performance evaluation method in the present invention also improves the defects of the international rotating bottle method. It realizes the dynamic simulation and accurate evaluation of the adhesion between aggregate and asphalt. Through the rotating bottle test, the degree of asphalt peeling on the aggregate surface is measured, which overcomes the shortcomings of the traditional water boiling method and water immersion method, and provides a more scientific, simple and reliable test method suitable for popularization and use. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] Figure 1 It is a schematic diagram of the three-dimensional structure of the present invention;

[0041] Figure 2 It is a side view structural diagram of the connection relationship between the bar track, the support frame, the motor, the gear and the light source;

[0042] Figure 3 It is a side view structural diagram of the connection relationship between the first support plate, the chain, the transition gear, the passive gear, the active gear, the driving member, the test bottle of the specimen and the strip support plate, and only part of the chain is shown in the figure;

[0043] Figure 4 It is a three-dimensional structural schematic diagram of the connection relationship between the bottom plate, the first support plate, the second support plate, the roller body, the transmission sprocket assembly, the driving member, the test bottle of the specimen and the strip support plate;

[0044] Figure 5 A schematic diagram of a top view of the connection relationship between the main support plate, the strip track, the self-moving light source, the embedded camera and the shadowless array light source bar;

[0045] Figure 6 A side view structural diagram showing the connection relationship between the main support plate, the strip track, the support frame, the motor, the gear, the light source and the polyester reflective film sheet;

[0046] Figure 7 A schematic diagram of the top view of the connection relationship between the bottom plate, the first support plate, the roller body, the transmission sprocket assembly, the driving member, the strip support plate and the paving plate;

[0047] Figure 8 It is a top view structural schematic diagram of the connection relationship between the roller body and the test bottle of the specimen;

[0048] Fig. 9 It is a side view structural schematic diagram of the connection relationship between the bottom plate, the first support plate, the second support plate, the roller body, the test bottle of the specimen and the support feet;

[0049] Fig.10 A schematic diagram of the side view of the connection between the base plate, the housing, the legs and the handle;

[0050] Fig.11 It is an operation flow chart of the method for evaluating the adhesion performance of aggregate asphalt in the present invention.

[0051] In the figure: 1-bottom plate; 2-housing; 3-first support plate; 4-second support plate; 5-roller; 6-drive sprocket assembly; 6-1-chain; 6-2-transition gear; 6-3-passive gear; 6-4-driving gear; 7-driving member; 8-clamping gap; 9-multi-position irradiation shadowless lamp assembly; 9-1-main support plate; 9-2-strip track; 9-3-self-moving light source; 9-3-1-support frame; 9-3-2-motor; 9-3-3-gear; 9-3-4-light source; 10-test piece test bottle; 11-first opening; 12-opening and closing cover; 13-second opening; 14-visible window; 15-strip support plate; 16-spreading board; 17-polyester reflective film sheet; 18-embedded camera; 19-shadowless array light source bar; 20-support foot; 21-heat dissipation hole; 22-handle. DETAILED DESCRIPTION

[0052] The following describes the embodiments of the present invention through specific examples, and those skilled in the art can easily understand other advantages and effects of the present invention from the contents disclosed in this specification. The present invention can also be implemented or applied through other different specific embodiments, and the details in this specification can also be modified or changed in various ways based on different viewpoints and applications without departing from the spirit of the present invention.

[0053] Specific implementation method 1: Combination Figure 1 , Figure 2 , Figure 3 , Figure 4 , Figure 5 , Figure 6 , Figure 7 , Figure 8 , Fig. 9 and Fig.10 The present embodiment is described. The aggregate asphalt adhesion rotation tester of the present embodiment comprises a base plate 1, a housing 2 and a test assembly, wherein the test assembly is arranged on the base plate 1, the housing 2 is arranged on the base plate 1, an inner cavity is formed between the housing 2 and the base plate 1, and the test assembly is arranged in the inner cavity;

[0054] The test component includes a first support plate 3, a second support plate 4, a plurality of rollers 5, a transmission sprocket assembly 6 and a driving member 7. The first support plate 3 and the second support plate 4 are vertically arranged side by side on the bottom plate 1. The transmission sprocket assembly 6 is arranged on the first support plate 3. The transmission sprocket assembly 6 is connected to the driving member 7. A plurality of rollers 5 are arranged between the first support plate 3 and the second support plate 4. One end of each roller 5 passes through the first support plate 3 and is connected to the transmission sprocket assembly 6. The other end of each roller 5 is hinged to the second support plate 4. Each roller 5 makes a self-rotation movement between the first support plate 3 and the second support plate 4 driven by the transmission sprocket assembly 6. A clamping gap 8 for fitting the test bottle 10 of the specimen is formed between every two adjacent rollers 5.

[0055] Further, an outer shell 2 and a test component are respectively arranged above the bottom plate 1. Four feet 20 are arranged below the bottom plate 1. The four feet 20 are arranged near the four angles of the bottom plate 1. The feet 20 provide stable support for this rotary tester. The test component is arranged in the outer shell 2. A plurality of heat dissipation holes 21 are respectively machined on both sides of the outer shell 2 along its thickness direction. The heat dissipation holes 21 can effectively discharge the heat generated by the test component during operation, ensuring that the temperature of this rotary tester can be stable, thereby avoiding the influence of temperature change on the test results. The test bottle 10 of the specimen is a bottle body containing asphalt and aggregate. The test bottle 10 of the specimen is arranged in the clamping gap 8 of each roller 5. Each roller 5 makes a self-rotation movement driven by the transmission sprocket assembly 6, so as to ensure that the test bottle 10 of the specimen can realize the function of continuous rotation, and ensure that the asphalt and aggregate in the test bottle 10 of the specimen can be evenly mixed together during the rotation process.

[0056] Specific embodiment two: This embodiment is a further limitation of specific embodiment one. A first opening 11 is machined on the outer shell 2, and an opening and closing cover plate 12 is hinged on the first opening 11.

[0057] Further, a handle 22 is arranged on the outer shell 2. The handle 22 is in a C shape. One side of the handle 22 is connected to the outer shell 2, and the other side of the handle 22 is suspended. The staff can open or close the opening and closing cover plate 12 through the handle 22.

[0058] Specific embodiment three: This embodiment is a further limitation of specific embodiment one or two. A second opening 13 is machined on the outer shell 2, and a viewing window 14 is arranged on the second opening 13.

[0059] Further, the staff uses the handle 22 to open the opening and closing cover plate 12, and places the test bottle 10 of the specimen between two adjacent rollers 5, and then closes the opening and closing cover plate 12. The staff observes the test results through the viewing window 14.

[0060] Specific embodiment four: This embodiment is a further limitation of specific embodiments one, two or three. The transmission sprocket assembly 6 includes a chain 6-1, a transition gear 6-2, a driving gear 6-4 and a plurality of passive gears 6-3. The passive gears 6-3 are arranged in a one-to-one correspondence with the roller bodies 5. Each passive gear 6-3 is sleeved on its corresponding roller body 5. One end of the roller body 5 passes through the first support plate 3 and is connected to its corresponding passive gear 6-3. A transition gear 6-2 is arranged on the outer wall of the first support plate 3. A driving gear 6-4 is sleeved on the power output shaft of the driving member 7. A chain 6-1 is arranged between the driving gear 6-4, the transition gear 6-2 and the plurality of passive gears 6-3.

[0061] Furthermore, the roller body 5 is horizontally arranged between the first support plate 3 and the second support plate 4, and one end of the roller body 5 passes through the first support plate 3 and is connected to the passive gear 6-3. At the same time, a transition gear 6-2 is arranged on the outer wall of the first support plate 3, wherein the driving member 7 has the same working principle as the motor in the prior art, and the driving member 7 can drive the active gear 6-4 to rotate, and the active gear 6-4 cooperates with the transition gear 6-2 and multiple passive gears 6-3 through the chain 6-1 to ensure that the roller body 5 realizes rotational motion, wherein the driving member 7 is prone to generate heat during operation, and the heat can be discharged through the heat dissipation holes 21 on the outer shell 2 to ensure the temperature stability in the rotation tester.

[0062] Specific implementation mode five: This implementation mode is a further limitation of specific implementation modes one, two, three or four. The outer wall of the first support plate 3 is provided with a strip support plate 15 along its length direction, and the strip support plate 15 is provided with a chain 6 - 1 .

[0063] Furthermore, the chain 6-1 is respectively meshed with the transition gear 6-2, the driving gear 6-4 and multiple passive gears 6-3, wherein the chain 6-1 is arranged on a strip support plate 15, and the strip support plate 15 ensures that the chain 6-1 can run stably and avoid falling off, wherein the chain 6-1 has the same working principle as the chain in the prior art.

[0064] Specific embodiment six: This embodiment is a further limitation of specific embodiments one, two, three, four or five, and also includes a multi-position irradiation shadowless lamp assembly 9. One arrangement form of the multi-position irradiation shadowless lamp assembly 9 is that the multi-position irradiation shadowless lamp assembly 9 is arranged on the inner wall of the outer shell 2, and the multi-position irradiation shadowless lamp assembly 9 is cooperated with a paving board 16. The paving board 16 is arranged directly below the multi-position irradiation shadowless lamp assembly 9, and the paving board 16 is arranged on the bottom plate 1. The multi-position irradiation shadowless lamp assembly 9 includes a main support plate 9-1, a plurality of strip rails 9-2 and a plurality of self-moving light sources 9-3. The main support plate 9-1 is provided with a plurality of strip rails 9-2, and each strip rail 9-2 is provided with a plurality of self-moving light sources 9-3; each self-moving The light source 9-3 includes a support frame 9-3-1, a motor 9-3-2, a gear 9-3-3 and a light source 9-3-4. The support frame 9-3-1 is arranged on the corresponding strip track 9-2. One side of the support frame 9-3-1 is slidably matched with one side of the strip track 9-2. The other side of the support frame 9-3-1 is provided with a motor 9-3-2. The output shaft of the motor 9-3-2 is provided with a gear 9-3-3. The gear 9-3-3 is meshed with the strip track 9-2. The bottom of the support frame 9-3-1 is hinged with a light source 9-3-4. The light source 9-3-4 emits light in a direction directly downward or tilted downward. The light source 9-3-4 moves back and forth along the length direction of the strip track 9-2 driven by the support frame 9-3-1.

[0065] Furthermore, the multi-position irradiation shadowless lamp assembly 9 also includes a plurality of shadowless array light source bars 19, and the plurality of shadowless array light source bars 19 are respectively arranged near both sides and both ends of the main support plate 9-1, and each strip track 9-2 is arranged in the shadowless array light source bar 19, and each strip track 9-2 is provided with a plurality of self-moving light sources 9-3; the motor 9-3-2 drives the support frame 9-3-1 to reciprocate on the strip track 9-2, thereby ensuring that the light source 9-3-4 on the support frame 9-3-1 can reciprocate, wherein the motor 9-3-2 has the same working principle as the motor in the prior art.

[0066] In this embodiment, a plurality of support blocks are provided on one side of the main support plate 9-1 facing the bottom plate 1, and the support blocks are vertically arranged at both ends of the strip track 9-2, thereby ensuring that a plurality of self-moving light sources 9-3 can slide back and forth on the strip track 9-2.

[0067] The configuration of the multi-position irradiation shadowless lamp assembly 9 in this embodiment is conducive to the accurate acquisition of basic data in the adhesion evaluation method, and its significance lies in eliminating the interference of the external environment on the captured image and providing a standardized image acquisition environment. Among them, multiple self-moving light sources 9-3 and multiple shadowless array light source bars 19 cooperate with each other to form a uniform light source array structure. Multiple shadowless array light source bars 19 can also be provided with a lifting mechanism, which is used to raise or lower the position of the shadowless array light source bar 19. The lifting mechanism formed by the cooperation of gears and racks can be used, and other existing lifting mechanisms can also be replaced. The initial installation height of the shadowless array light source bar 19 is set to 50cm, allowing the light source height to be adjusted between 30cm and 60cm to accommodate shooting objects of different sizes.

[0068] Furthermore, the specific parameters of the shadowless array light source strip 19 are:

[0069] Luminous flux: The luminous flux of each shadowless array light source strip 19 should be between 800 and 1200 lumens lm to ensure sufficient lighting effect.

[0070] Color temperature: The color temperature of the shadowless array light source bar 19 is set to 5000K to 6500K, which is close to the color temperature of daylight to provide a natural lighting effect, which is suitable for aggregate shooting conditions.

[0071] Light source type: The shadowless array light source bar 19 is an LED light source to provide an efficient, low-heat lighting solution to prevent overheating from affecting the normal performance of the aggregate and embedded camera 18.

[0072] Tilt angle of front and rear light sources: set to 15 degrees, with the illumination direction toward the center; the left and right light sources are also tilted 15 degrees to ensure a cross-lighting effect.

[0073] In this embodiment, the specific form of the mutual cooperation between the main support plate 9-1, the strip track 9-2 and the self-moving light source 9-3 is: the self-moving light source 9-3 can adjust the position and angle of the light source in real time according to different shooting requirements. The setting of the self-moving light source 9-3 ensures that the lighting conditions can be flexibly adjusted during the shooting process to adapt to the characteristics of the aggregate, so as to obtain the best lighting effect. The self-moving light source 9-3 system better adapts to the reflective characteristics of different materials, further improving the image quality and data availability. The self-moving light source 9-3 is fixed on the main support plate 9-1 through the strip track 9-2, and a total of 8 strip tracks 9-2 are provided, and each strip track 9-2 can be installed with multiple self-moving light sources 9-3. When used specifically, the relevant parameters of the light source 9-3-4 in the self-moving light source 9-3 are: the self-moving light source 9-3 has six degrees of freedom, the adjustment angle range is -30 degrees to +30 degrees, and supports 360-degree rotation function. The control system uses wireless remote control for adjustment, and integrates a light intensity sensor to monitor the lighting effect in real time. The light source type is LED, the luminous flux is set at 800 to 1200 lumens, and the color temperature range is 5000K to 6500K.

[0074] The specific situation of the laying process of the polyester reflective film sheet 17 in this embodiment is as follows: the polyester reflective film sheet 17 is used as a shadowless structure, specifically, the polyester reflective film sheet 17 is formed by using high-quality polypropylene scattering paint on the inner surface of the shell 2, so that the reflectivity of the polyester reflective film sheet 17 should not be less than 85%, and the coating thickness is 5mm. This material can effectively diffuse the incident light to ensure that it evenly illuminates the entire shooting area, thereby significantly reducing the generation of local highlight areas and shadows, and improving the overall uniformity and quality of the image. The color of the bottom plate 1 is green, and the specific RGB value is (0,255,0). After experimental verification, the green background performs best in the RGB color space processed by the subsequent software, enhancing the image recognition effect.

[0075] In this embodiment, the embedded camera 18 can achieve standardized shooting. Specifically, the embedded camera 18 is embedded and connected to the main support plate 9-1, so that the embedded camera 18 can adapt to different types of aggregates. This design significantly simplifies the operation process and reduces human errors, thereby improving the unified standardization effect of shooting. According to the different characteristics of the aggregate, the embedded camera 18 will automatically configure the focal length, aperture, ISO, white balance and exposure time to ensure the best image quality in a variety of environments. The shooting form realized by the embedded camera 18 not only improves the accuracy of data collection, but also provides strong support for subsequent image recognition and analysis, and provides favorable support for the comprehensive evaluation of the adhesion between aggregates and asphalt.

[0076] The relevant setting parameters of the embedded camera 18 are:

[0077] When the shutter speed of the embedded camera 18 is 1 / 80s, the corresponding ISO sensitivity is ISO 400, the focal length is 24mm, and the aperture is f / 1.6;

[0078] When the shutter speed of the embedded camera 18 is 1 / 60s, the corresponding ISO sensitivity is ISO200, the focal length is 20mm, and the aperture is f / 1.8;

[0079] When the shutter speed of the embedded camera 18 is 1 / 90s, the corresponding ISO sensitivity is ISO300, the focal length is 24mm, and the aperture is f / 1.5;

[0080] The above parameter configuration method is selected according to different situations and is suitable for comprehensively acquiring effective and stable basic image data.

[0081] Specific embodiment seven: This embodiment is a further limitation of specific embodiments one, two, three, four, five or six, and also includes a multi-position irradiation shadowless lamp assembly 9. Another arrangement form of the multi-position irradiation shadowless lamp assembly 9 is that the multi-position irradiation shadowless lamp assembly 9 is arranged outside the housing 2, and cooperates with a chassis and another paving board 16 to realize the independent acquisition process of a large number of shadowless images without relying on the bottom plate 1 and the housing 2, which is conducive to the accurate and effective acquisition of basic data. The multi-position irradiation shadowless lamp assembly 9 includes a main support plate 9-1, a plurality of strip rails 9-2 and a plurality of self-moving light sources 9-3. A plurality of strip rails 9-2 are arranged at the bottom of the main support plate 9-1, and a plurality of self-moving light sources 9-3 are arranged on each strip rail 9-2; Each self-moving light source 9-3 includes a support frame 9-3-1, a motor 9-3-2, a gear 9-3-3 and a light source 9-3-4. The support frame 9-3-1 is arranged on its corresponding strip track 9-2. One side of the support frame 9-3-1 is slidably matched with one side of the strip track 9-2. The other side of the support frame 9-3-1 is provided with a motor 9-3-2. The output shaft of the motor 9-3-2 is provided with a gear 9-3-3. The gear 9-3-3 is meshed with the strip track 9-2. The bottom of the support frame 9-3-1 is hinged with a light source 9-3-4. The light source 9-3-4 emits light in a downward direction. Driven by the support frame 9-3-1, the light source 9-3-4 reciprocates along the length direction of the strip track 9-2.

[0082] Specific embodiment eight: This embodiment is a further limitation of specific embodiments one, two, three, four, five, six or seven. A polyester reflective film sheet 17 is provided at the bottom of the main support plate 9-1, and a plurality of strip tracks 9-2 are arranged in an array on the polyester reflective film sheet 17. An embedded camera 18 is provided at the center of the bottom of the main support plate 9-1.

[0083] Furthermore, the light source 9-3-4 and the shadowless array light source bar 19 illuminate the paving board 16 and can be reflected on the polyester reflective film sheet 17, thereby ensuring sufficient light inside the rotation tester. When the embedded camera 18 is shooting, the high quality of the captured image is ensured, and the final photo taken will be used for subsequent image recognition processing, providing a reliable data basis for image recognition.

[0084] The installation process and working principle of the aggregate asphalt adhesion rotation tester in the present invention are as follows: the outer shell 2 is installed on the base plate 1, and the test assembly is installed in the outer shell 2, wherein the first support plate 3 and the second support plate 4 on the test assembly are vertically installed on the base plate 1, one end of the multiple rollers 5 pass through the first support plate 3 and are connected to the transmission sprocket assembly 6, and the other ends of the multiple rollers 5 are hinged on the second support plate 4, and the driving member 7 is started. The driving member 7 drives the multiple rollers 5 to rotate on the first support plate 3 and the second support plate 4 through the transmission sprocket assembly 6. The staff puts the specimen test bottle 10 in the clamping gap 8 between the two rollers 5, and the specimen test bottle 10 realizes rotational motion.

[0085] Specific implementation method nine: Combination Figures 1 to 11 This embodiment is described. In this embodiment, the aggregate asphalt adhesion rotation tester includes a base plate 1, a shell 2 and a test assembly. The test assembly is arranged on the base plate 1, the shell 2 is arranged on the base plate 1, an inner cavity is formed between the shell 2 and the base plate 1, and the test assembly is arranged in the inner cavity;

[0086] The test assembly includes a first support plate 3, a second support plate 4, a plurality of rollers 5, a transmission sprocket assembly 6 and a driving member 7. The first support plate 3 and the second support plate 4 are vertically arranged in parallel on the base plate 1. The transmission sprocket assembly 6 is arranged on the first support plate 3, and the transmission sprocket assembly 6 is connected to the driving member 7. The plurality of rollers 5 are arranged between the first support plate 3 and the second support plate 4. One end of each roller 5 passes through the first support plate 3 and is connected to the transmission sprocket assembly 6, and the other end of each roller 5 is hinged to the second support plate 4. Each roller 5 rotates between the first support plate 3 and the second support plate 4 under the drive of the transmission sprocket assembly 6, and a clamping gap 8 for matching the test bottle 10 of the specimen is formed between every two adjacent rollers 5.

[0087] In this embodiment, the aggregate asphalt adhesion performance evaluation method implemented by the aggregate asphalt adhesion rotation tester is to use the image data collected by the aggregate asphalt adhesion rotation tester as basic data, model the basic data, and form a relevant model, so as to complete the quantitative evaluation process of the aggregate asphalt adhesion performance according to the model, that is, to use the image data collected by the aggregate asphalt adhesion rotation tester as basic data, and to quantitatively evaluate the basic data through the image recognition method based on deep learning, specifically:

[0088] The quantitative evaluation process through basic data is to form a training set with the image data obtained by the aggregate asphalt adhesion rotation tester. A predetermined number of images are randomly cut out from the training set to form a classification. The class center point of each category is defined as the arithmetic mean of the pixel coordinates of all pixels in the category. The calculation formula of the arithmetic mean is:

[0089]

[0090] In the above formula, (x i ,y i ) is the coordinate of the i-th pixel in category C, N C is the number of pixels in category C; a weight coefficient w is introduced in each category, and the sampling probability of each category is adjusted according to the weight coefficient w. The sampling probability P of each category is calculated according to its corresponding weight coefficient w. The calculation formula of the sampling probability P is:

[0091]

[0092]

[0093] According to the formula of the sampling area A of each category, the size of the sampling area is calculated by the weight coefficient w and the area of ​​the category area. The size formula of the sampling area of ​​each category is:

[0094] A 沥青 =P 沥青 A 沥青 =0.5333A 沥青 ;

[0095] A 集料 =P 集料 A 集料 =0.4A 集料 ;

[0096] A 背景 =P 背景 A 背景 =0.0667A 背景 ;

[0097] The loss function calculation process is beneficial to improve the segmentation accuracy:

[0098] The local contrast loss is introduced in the atrous space pyramid pooling module. The key weight coefficients α = 0.23, β = 0.36, γ = 0.41, τ = 0.07 are obtained in the experiment by combining the cross entropy loss with the DICE loss. The loss function is obtained based on the key weight coefficients α = 0.23, β = 0.36, γ = 0.41, τ = 0.07. The loss function formula is:

[0099]

[0100] In the above formula, is the total loss function, is the cross entropy loss function, is the DICE loss function, is the local contrast loss function, the total loss function Combined with the cross entropy loss function DICE loss function And the local contrast loss function Realize the optimization process of the entire model; cross entropy loss function Used to measure the difference between the predicted probability distribution and the true label distribution. DICE loss function DICE loss is used to measure the quality of overlap between the model prediction and the true label area. Local contrast loss function The main purpose of establishing it is to optimize the sample distribution in the feature space so that the feature vectors of samples of the same category are more similar and the feature vectors of samples of different categories are more distinguishable. τ is an important hyperparameter in contrastive learning, which is used to control the smoothness of the feature similarity distribution; τ is a hyperparameter that controls the smoothness of the feature similarity distribution in contrastive learning; N represents the total number of pixels in the image, C represents the number of categories, including the number of backgrounds, the number of asphalts, and the number of aggregates, y i,c Indicates whether pixel i belongs to category c, p i,c represents the predicted probability that pixel i belongs to category c; p i represents the model's predicted probability for pixel i, y i represents the actual label of pixel i; z i and z j is a positive sample pair of feature vectors, z k is the negative sample feature vector, sim(z i ,z j ) is the cosine similarity;

[0101] The quantitative evaluation conclusion of aggregate asphalt adhesion performance is obtained through the calculation of loss function: when the task and data scale are determined, When it is lower than 0.2, it indicates that the model is in a good fitting state, indicating that the asphalt coverage is in a good state, and the corresponding aggregate asphalt adhesion performance is good; when If it is still greater than 1.0 after twenty rounds of iterative training, it usually means that the model has not converged or is underfitting, indicating that the asphalt coverage is in a poor state and the corresponding aggregate asphalt adhesion performance is poor. In the future, it will be necessary to increase the number of training rounds or adjust the network structure and hyperparameters to re-acquire the basic data for calculation.

[0102] Specific implementation method 10: This implementation method is a further limitation of specific implementation method 9. In the actual training process, It will converge to a relatively stable value range, depending on the different limitations of the task and data size. Generally, if it is lower than 0.2, it is considered a good fit; if it is greater than 1.0 for a long time, it usually means that the model has not converged or is underfitting, and it is necessary to increase the number of training rounds or adjust the network structure, hyperparameters, etc. When it converges to 0.147, it shows that it has an excellent and good fitting effect and can accurately predict the coverage of asphalt.

[0103] The aggregate asphalt adhesion performance evaluation method of the present invention can accurately measure the asphalt coverage, and further evaluate the adhesion performance based on this, which is summarized as follows after a lot of research and calculation:

[0104] When the coverage is greater than or equal to 90%, the adhesion performance is generally considered to be very good; when the coverage is between 80% and 90%, it can be considered to be good adhesion performance; when the coverage is between 70% and 80%, it indicates that the adhesion performance is at a medium level; and when the coverage is less than 70%, the adhesion performance is generally considered to be poor.

[0105] Specific implementation eleven: This implementation is a further limitation of specific implementation nine or ten. In this implementation, the aggregate asphalt adhesion performance evaluation method is to use basic data to perform a quantitative evaluation of aggregate asphalt adhesion performance. The process also includes the calculation process of the inter-category separation index and the intra-category aggregation index. The specific calculation process is:

[0106] Inter-class separation index

[0107] Intra-class clustering index

[0108] In the above formula, and are the centroids of categories c1 and c2, x i is the feature vector of the i-th sample in the category.

[0109] The inter-class separation index is the same as the loss function. Its main function is to evaluate the accuracy of the evaluation results in the aggregate asphalt adhesion performance evaluation method. Through multiple verifications of the model, the threshold of the inter-class separation index is 2.0, and the threshold of the intra-class aggregation index is 0.8; when the index is greater than 2.0, it means that the features of different categories have a high degree of distinction in space; at the same time, the threshold of the intra-class aggregation index is set to 0.8: when the index is less than 0.8, it means that the distribution of features within the same category is more compact, so that higher recognition accuracy and stability can be obtained in practical applications. At this threshold, the model can accurately predict the residual asphalt coverage.

[0110] Specific embodiment 12: This embodiment is a further limitation of specific embodiments 9, 10 or 11. In this embodiment, the multi-position irradiation shadowless lamp assembly 9 is used to spread the aggregate and asphalt mixture after the rotating bottle test on the paving board 16 after the rotating bottle test. The multi-position irradiation shadowless lamp assembly 9 irradiates and collects relevant images of the aggregate and asphalt mixture on the paving board 16, thereby forming basic data for studying the adhesion status of the aggregate and asphalt, so as to facilitate subsequent image recognition and analysis. The inner cavity provides an independent and controllable shooting environment to avoid interference from external light and background. The shadowless array light source bar 19 and the multiple self-moving light sources 9-3 cooperate with each other to realize the arrangement structure of the dynamic and static circumferential light sources, which can ensure uniform light during shooting and eliminate shadow effects. The self-moving light source 9-3 can adjust the light source position and angle according to different shooting requirements to enhance the adaptability of light. The visual window 14 is convenient for observation and adjustment of shooting angles.

[0111] The specific operation process of this aggregate asphalt adhesion rotation tester is as follows:

[0112] Step 1: After the prepared aggregate and asphalt mixture specimens are fully cooled, the staff puts them into the aggregate asphalt bottle 10 to ensure that the aggregate is evenly coated with asphalt and fixed on multiple rollers 5. Multiple rollers 5 are installed between the second support plate 4 and the first support plate 3, and the aggregate asphalt bottle 10 is installed between every two adjacent rollers 5 to ensure the stability of the aggregate asphalt bottle 10 driven by the multiple rollers 5 during the rotation process.

[0113] Step 2: Set the rotation speed V and rotation time T of the aggregate asphalt bottle 10 through the control system. Set the speed required by the test through the speed control button. The rotation speed and time should be adjusted according to the test requirements and specifications to simulate different working conditions.

[0114] Step 3: Start the driving member 7, which drives the driving gear 6-4 to rotate. The driving gear 6-4 drives the transition gear 6-2 and multiple passive gears 6-3 to rotate through the chain 6-1, ensuring that the roller body 5 rotates and transmits power to the aggregate asphalt bottle 10, ensuring that the aggregate asphalt bottle 10 rotates at the set speed. The aggregate and asphalt mixture move in the water medium environment in the aggregate asphalt bottle 10, simulating the working conditions of asphalt and aggregate in the actual road environment.

[0115] Step 4: The rotation speed and time of the device during operation are monitored and recorded in real time through the control system and display screen, and the asphalt peeling in the aggregate asphalt bottle 10 is observed through the visual window 14. After the specified rotation time, the asphalt mixture is taken out and the degree of asphalt peeling on the aggregate surface is observed to evaluate the asphalt-aggregate adhesion.

[0116] Step 5: After the test of the aggregate asphalt bottle 10 is completed, the asphalt mixture is moved to the paving board 16 and photographed by the embedded camera 18 on the multi-position irradiation shadowless lamp assembly 9. The housing 2 provides an independent shooting environment. The embedded camera 18 is illuminated by the multi-position irradiation shadowless lamp assembly 9 during shooting to eliminate shadow interference and ensure the clarity and consistency of the image. The multi-position irradiation shadowless lamp assembly 9 ensures that the object is in the best position to obtain high-quality images. These images will be used for subsequent analysis, and the deep learning-based method proposed in the present invention is used to further evaluate the adhesion of asphalt-aggregate.

[0117] The present invention proposes a standard form of image acquisition, standardizes the acquisition process, avoids interference caused by external factors during image acquisition, completely eliminates the interference of shadows through the multi-position irradiation shadowless lamp assembly 9, and simultaneously forms quantitative shooting distance, material and color of the base plate 1, and picture shooting parameters, thereby improving the standardized quality of image acquisition.

[0118] The present invention develops image recognition software based on computer vision technology and deep learning technology. The present invention provides a preprocessing solution for collecting pictures, and collects and annotates large-scale data sets, solving the problem that there are no similar data sets in the relevant fields. At the same time, the present invention uses this data set to train a semantic segmentation model with ResNet-101 as the backbone network model, providing a high-precision automatic recognition solution. It solves the problem of low precision, high error and cumbersomeness of previous evaluation methods.

[0119] The adhesion evaluation method of the present invention applies deep learning technology to the evaluation process of the rotating bottle test, breaking through the traditional threshold segmentation-based method. Compared with the common manual labeling and empirical judgment in the industry, a deep learning semantic segmentation model with ResNet-101 as the backbone network was proposed and introduced for the first time to realize the automatic processing of images in the rolling bottle test. This adhesion evaluation method optimizes the evaluation process through deep learning, solves the subjective problem caused by the reliance on manual experience in traditional methods, and improves the objectivity and accuracy of the evaluation results. The introduction of this technology provides a more accurate and efficient evaluation method for the rolling bottle test.

[0120] The adhesion evaluation method in the present invention is based on the semantic segmentation model of the ResNet-101 architecture, which has obvious advantages in solving the problem of class imbalance and improving segmentation accuracy. First, combined with ResNet-101 as the feature extraction backbone network, it can more effectively capture local information in complex images compared with traditional U-Net or ResNet-50 or other shallow networks. Secondly, a self-created class center sampling method is proposed to ensure that samples of all categories are effectively sampled in each training cycle, especially by repeated sampling to increase the training probability of asphalt pixels, which successfully alleviates the class imbalance problem and accelerates the training process. At the same time, the introduction of local contrast loss and ASPP void space pyramid pooling module enables the model to finely process image features at multiple scales, further improving the boundary processing accuracy and overall segmentation effect. A new loss function is introduced to further optimize the classification performance and effectively cope with the challenge of class imbalance. This adhesion evaluation method combines sampling strategy, loss function and network architecture to not only improve the segmentation accuracy, but also enhance the generalization ability of the model.

[0121] Combination Fig.11 As shown, the adhesion evaluation method of the present invention also includes an image recognition algorithm, which adopts a strategy based on centroid sampling to effectively solve the imbalance problem of asphalt and aggregate categories in the training data set. This method significantly enhances the model's recognition ability on different categories by randomly sampling the centroid of the category in each training cycle and cutting out a fixed-size image block from the center point of the class. Especially when dealing with small sample categories, this method greatly improves the generalization ability of the model.

[0122] During the experiment, the present invention randomly cut out 1200 images from the training set, each with 600×600 pixels, and verified that the image size of 600×600 pixels achieved the best balance between efficiency and accuracy. This size not only improves the speed and efficiency of training, but also solves the problem of imbalanced pixels among the three categories of residual asphalt, aggregate, and background while retaining sufficient detail information. By reasonably selecting the image size, the model can maintain high-precision performance between different categories and avoid the impact of category imbalance. The choice of 600×600 pixels not only takes into account the optimization of computing resources, but also improves the accuracy of the model.

[0123] The class center point of each category is defined as the arithmetic mean of the pixel coordinates of all pixels in that category. The specific formula is as follows:

[0124]

[0125] Among them, (x i ,y i ) is the coordinate of the i-th pixel in category C, N C is the number of pixels in class C.

[0126] In order to further optimize the sampling process, the present invention introduces a weight coefficient w for each category and adjusts the sampling probability of each category according to these weight coefficients. The sampling probability P of each category is calculated according to its weight coefficient w. The sampling probability formula proposed by the present invention is as follows:

[0127]

[0128] In addition, the present invention also provides a formula for the area A of the sampling area of ​​each category, and calculates the size of the sampling area by the weight coefficient and the area of ​​the category area. The formula for the size of the sampling area of ​​each category proposed by the present invention is as follows:

[0129] A 沥青 =P 沥青 A 沥青 =0.5333A 沥青

[0130] A 集料 =P 集料 A 集料 =0.4A 集料

[0131] A 背景 =P 背景 A 背景 =0.0667A 背景

[0132] The present invention proposes a new self-created loss function to improve segmentation accuracy, specifically:

[0133] In the present invention, the local contrast loss method is introduced into the hollow space pyramid pooling module of the deep learning model to further optimize the discrimination of different categories in the feature space. This method effectively improves the segmentation accuracy of the model in boundary details and enhances the recognition ability of the interface area between asphalt and aggregate. It is used in conjunction with the cross entropy loss and the DICE loss, and the key weight coefficients α = 0.23, β = 0.36, γ = 0.41, τ = 0.07 are obtained in a large number of experiments. The loss function proposed by the present invention is shown in the formula below:

[0134]

[0135] Where: N is the total number of pixels in the image, C is the number of categories (such as background, asphalt, aggregate), y i,c Indicates whether pixel i belongs to category c, p i,c represents the predicted probability that pixel i belongs to category c. i represents the model's predicted probability for pixel i, y i Ah represents the true label of pixel i. z i and z j is a positive sample pair of feature vectors, z k is the negative sample feature vector, sim(z i ,z j ) is the cosine similarity.

[0136]

[0137] A set of optimal parameter values ​​were proposed based on the experimental data, and the Adam optimizer was optimized.

[0138] The present invention improves and optimizes the traditional Adam optimizer and adjusts the key parameters in the optimizer to better meet the needs of specific tasks. After a large number of experiments and fitting the experimental data, a set of optimal parameter values ​​are obtained. These adjustments help to improve the convergence speed and accuracy of the model.

[0139] The specific formula of the optimized Adam optimizer is as follows:

[0140] First-order moment (momentum) update formula:

[0141] m t =0.85m t-1 +0.15g t

[0142] Second-order moment (momentum) update formula:

[0143] Deviation correction formula:

[0144]

[0145] Parameter update formula:

[0146]

[0147] By adaptively adjusting these parameters, the present invention solves the efficiency problem of the traditional Adam optimizer in certain specific tasks through the fitting optimization of a large amount of experimental data. The adjusted Adam optimizer can not only improve the model performance, but also accelerate the convergence process, avoiding the slow convergence or instability problems in traditional methods.

[0148] In the present invention, the performance evaluation of the model not only relies on traditional indicators such as precision and recall, but also applies the two evaluation indicators of inter-class separation and intra-class aggregation to the evaluation of aggregate and asphalt adhesion for the first time, which can provide a deeper understanding of the performance of the model in the feature space and significantly optimize the classification effect of aggregate and asphalt adhesion features. By introducing these two indicators, the model's ability to separate different categories (residual asphalt, aggregate and background) and the consistency of samples of the same category can be effectively evaluated, thereby improving the model's discrimination and accuracy, especially under complex conditions such as the present invention.

[0149]

[0150] and are the centroids of categories c1 and c2, x i is the feature vector of the i-th sample in the category.

[0151] This paper is the first to apply inter-category separation and intra-category aggregation to the asphalt-aggregate adhesion algorithm, providing a new perspective and technical path when traditional methods cannot effectively solve the problem. This paper significantly improves the model's ability to identify the characteristics of aggregate-asphalt adhesion through an algorithm adhesion evaluation method based on deep learning, providing favorable data support for research in related fields.

Claims

1. An aggregate asphalt adhesion rotation tester, characterized in that: It comprises a base plate (1), a shell (2) and a test component, wherein the shell (2) is arranged on the base plate (1), an inner cavity is formed between the shell (2) and the base plate (1), and the test component is arranged in the inner cavity; The test assembly comprises a first support plate (3), a second support plate (4), a plurality of rollers (5), a transmission sprocket assembly (6) and a driving member (7); the first support plate (3) and the second support plate (4) are arranged vertically in parallel on the bottom plate (1); the transmission sprocket assembly (6) is arranged on the first support plate (3); the transmission sprocket assembly (6) is connected to the driving member (7); the plurality of rollers (5) are arranged between the first support plate (3) and the second support plate (4); one end of each roller (5) passes through the first support plate (3) and is connected to the transmission sprocket assembly (6); the other end of each roller (5) is hinged to the second support plate (4); each roller (5) rotates between the first support plate (3) and the second support plate (4) under the drive of the transmission sprocket assembly (6); and a clamping gap (8) for matching the test bottle (10) of the test piece is formed between each two adjacent rollers (5).

2. The aggregate asphalt adhesion rotation tester according to claim 1, characterized in that: A first opening (11) is formed on the outer shell (2), and an opening and closing cover plate (12) is hingedly connected to the first opening (11).

3. The aggregate asphalt adhesion rotation tester according to claim 1 or 2, characterized in that: A second opening (13) is processed on the outer shell (2), and the second opening (13) is provided with a viewing window (14).

4. The aggregate asphalt adhesion rotation tester according to claim 3, characterized in that: The transmission sprocket assembly (6) comprises a chain (6-1), a transition gear (6-2), a driving gear (6-4) and a plurality of passive gears (6-3); the passive gears (6-3) are arranged in one-to-one correspondence with the roller bodies (5); each passive gear (6-3) is sleeved on its corresponding roller body (5); one end of the roller body (5) passes through a first support plate (3) and is connected to its corresponding passive gear (6-3); a transition gear (6-2) is arranged on the outer wall of the first support plate (3); a driving gear (6-4) is sleeved on the power output shaft of the driving member (7); and a chain (6-1) is arranged between the driving gear (6-4), the transition gear (6-2) and the plurality of passive gears (6-3).

5. The aggregate asphalt adhesion rotation tester according to claim 4, characterized in that: A strip-shaped support plate (15) is arranged on the outer wall of the first support plate (3) along its length direction, and a chain (6-1) is arranged on the strip-shaped support plate (15).

6. The aggregate asphalt adhesion rotation tester according to claim 1, characterized in that: The invention also comprises a multi-position irradiation shadowless lamp assembly (9), wherein the multi-position irradiation shadowless lamp assembly (9) is arranged on the inner wall of the housing (2), and the multi-position irradiation shadowless lamp assembly (9) is provided with a paving board (16), and the paving board (16) is provided directly below the multi-position irradiation shadowless lamp assembly (9), and the paving board (16) is provided on the bottom plate (1), and the multi-position irradiation shadowless lamp assembly (9) comprises a main support plate (9-1), a plurality of bar tracks (9-2) and a plurality of self-moving light sources (9-3), wherein the main support plate (9-1) is provided with a plurality of bar tracks (9-2), and each bar track (9-2) is provided with a plurality of self-moving light sources (9-3); each self-moving light source (9-3) comprises a support frame (9-3-1), a motor (9-3-2 ), gears (9-3-3) and light sources (9-3-4), the support frame (9-3-1) being arranged on the corresponding strip track (9-2), one side of the support frame (9-3-1) being slidably matched with one side of the strip track (9-2), the other side of the support frame (9-3-1) being provided with a motor (9-3-2), the output shaft of the motor (9-3-2) being provided with a gear (9-3-3), the gear (9-3-3) being meshed with the strip track (9-2), the bottom of the support frame (9-3-1) being hinged with a light source (9-3-4), the light of the light source (9-3-4) being emitted in a downward direction, and the light source (9-3-4) being driven by the support frame (9-3-1) to reciprocate along the length direction of the strip track (9-2).

7. The aggregate asphalt adhesion rotation tester according to claim 1, characterized in that: The invention also comprises a multi-position irradiation shadowless lamp assembly (9), which is arranged outside the housing (2), and comprises a main support plate (9-1), a plurality of bar-shaped rails (9-2) and a plurality of self-moving light sources (9-3). The bottom of the main support plate (9-1) is provided with a plurality of bar-shaped rails (9-2), and each bar-shaped rail (9-2) is provided with a plurality of self-moving light sources (9-3); each self-moving light source (9-3) comprises a support frame (9-3-1), a motor (9-3-2), a gear (9-3-3) and a light source (9-3-4), and the support frame (9-3-1) is provided with a plurality of bar-shaped rails (9-2). The support frame (9-3-1) is arranged on the corresponding bar track (9-2), one side of the support frame (9-3-1) is slidably matched with one side of the bar track (9-2), a motor (9-3-2) is arranged on the other side of the support frame (9-3-1), a gear (9-3-3) is mounted on the output shaft of the motor (9-3-2), the gear (9-3-3) is meshed with the bar track (9-2), a light source (9-3-4) is hingedly connected to the bottom of the support frame (9-3-1), the light of the light source (9-3-4) is emitted in a downward direction, and the light source (9-3-4) moves back and forth along the length direction of the bar track (9-2) driven by the support frame (9-3-1).

8. The aggregate asphalt adhesion rotation tester according to claim 7, characterized in that: A polyester reflective film sheet (17) is provided at the bottom of the main support plate (9-1), a plurality of strip tracks (9-2) are arranged in an array on the polyester reflective film sheet (17), and an embedded camera (18) is provided at the center of the bottom of the main support plate (9-1).

9. A method for evaluating the adhesion performance of aggregate asphalt, implemented by using the aggregate asphalt adhesion rotary tester according to claim 8, characterized in that: The method for evaluating the adhesion performance of aggregate asphalt is to use the image data collected by the aggregate asphalt adhesion rotation tester as the basic data, and to conduct a quantitative evaluation process of the adhesion of aggregate asphalt using the basic data, specifically: The quantitative evaluation process through basic data is to form a training set with the image data obtained by the aggregate asphalt adhesion rotation tester. A predetermined number of images are randomly cut out from the training set to form a classification. The class center point of each category is defined as the arithmetic mean of the pixel coordinates of all pixels in the category. The calculation formula of the arithmetic mean is: In the above formula, (x i ,y i ) is the coordinate of the i-th pixel in category C, N C is the number of pixels in category C; Then, a weight coefficient w is introduced in each category. The sampling probability of each category is adjusted according to the weight coefficient w. The sampling probability P of each category is calculated according to its corresponding weight coefficient w. The calculation formula of the sampling probability P is: According to the formula of the sampling area A of each category, the size of the sampling area is calculated by the weight coefficient w and the area of ​​the category area. The size formula of the sampling area of ​​each category is: <h2 style=";text-align:left;direction:ltr">A<h2 style=";text-align:left;direction:ltr"> 沥青 <h2 style=";text-align:left;direction:ltr"> =P<h2 style=";text-align:left;direction:ltr"> 沥青 <h2 style=";text-align:left;direction:ltr"> A<h2 style=";text-align:left;direction:ltr"> 沥青 <h2 style=";text-align:left;direction:ltr"> =0.5333A<h2 style=";text-align:left;direction:ltr"> 沥青 <h2 style=";text-align:left;direction:ltr"> ; A 集料 =P 集料 A 集料 =0.4A 集料 ; A 背景 =P 背景 A 背景 =0.0667A 背景 ; The loss function calculation process is beneficial to improve the segmentation accuracy: The local contrast loss is introduced in the atrous space pyramid pooling module. The key weight coefficients α = 0.23, β = 0.36, γ = 0.41, τ = 0.07 are obtained in the experiment by combining the cross entropy loss with the DICE loss. The loss function is obtained based on the key weight coefficients α = 0.23, β = 0.36, γ = 0.41, τ = 0.

07. The loss function formula is: In the above formula, is the total loss function, is the cross entropy loss function, is the DICE loss function; is the local contrast loss function; τ is a hyperparameter that controls the smoothness of feature similarity distribution in contrastive learning; N represents the total number of pixels in the image, C represents the number of categories, including the number of backgrounds, the number of asphalts, and the number of aggregates, and y i,c Indicates whether pixel i belongs to category c, p i,c represents the predicted probability that pixel i belongs to category c; p i represents the model's predicted probability for pixel i, y i represents the actual label of pixel i; z i and z j is a positive sample pair of feature vectors, z k is the negative sample feature vector, sim(z i ,z j ) is the cosine similarity; The quantitative evaluation conclusion of aggregate asphalt adhesion is obtained through the calculation of loss function: when the task and data scale are determined, When it is lower than 0.2, it indicates that the model is in a good fitting state, indicating that the asphalt coverage is in a good state, and the corresponding aggregate asphalt adhesion performance is good; when If it is still greater than 1.0 after twenty rounds of iterative training, it usually means that the model has not converged or is underfitting, indicating that the asphalt coverage is in a poor state and the corresponding aggregate asphalt adhesion is poor. In the future, it will be necessary to increase the number of training rounds or adjust the network structure and hyperparameters to re-acquire the basic data for calculation.

10. The method for evaluating the adhesion performance of aggregate asphalt according to claim 9, characterized in that: The quantitative evaluation process of aggregate asphalt adhesion using basic data also includes the calculation process of inter-category separation index and intra-category aggregation index. The specific calculation process is as follows: Inter-class separation index Intra-class clustering index In the above formula, and are the centroids of categories c1 and c2, x i is the feature vector of the i-th sample in the category; Through multiple verifications of the model, it was found that the threshold of the inter-category separability index is 2.0, and the threshold of the intra-category aggregation index is 0.8; when the inter-category separability index is greater than 2.0 and the intra-category aggregation index is less than 0.8, it means that the features of different categories have a high degree of distinction in space, and the distribution of features within the same category is compact, which is conducive to the model's accurate prediction of the residual asphalt coverage.

Images