A method for identifying mechanical parameters of geotechnical materials
Through the combination of digital speckle-related methods and finite unit method, the rapid, real-time and accuracy of the acquisition of mechanical parameters of geotechnical materials is solved, and the efficient identification of mechanical parameters of geotechnical materials is achieved, and the engineering is improved. Design and construction accuracy.
Patent Information
- Application Number
- CN202110719352.2
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2021-06-28
- Publication Date
- 2024-11-26
- Estimated Expiration
- Not applicable · inactive patent
AI Technical Summary
It is difficult for the existing technology to quickly, in real time and accurately obtain the mechanical parameters of geotechnical materials, which affects the accuracy of engineering design and construction.
The digital speckle correlation method and the finite element method are used to combine the improved artificial fish school optimization algorithm, and the mechanical parameters of geotechnical materials are inversely calculated by using the error between the displacement measurement value and the simulation value as the objective function.
It achieves rapid, real-time and accurate acquisition of mechanical parameters of geotechnical materials, improves the accuracy of engineering design and construction, and has important theoretical analysis and engineering application significance.
Smart Images

Figure CN113962118B_ABST
Abstract
Description
Technical Field
[0001] The invention relates to the technical field of geotechnical material mechanical parameter testing, and in particular to a geotechnical material mechanical parameter identification method based on a digital speckle correlation method and a finite element method. Background Art
[0002] With the application and promotion of computing technology in geotechnical engineering, numerical analysis and calculation methods involving geotechnical materials have been widely used in engineering design, construction, operation and maintenance of tunnels, slopes, foundation pits, and engineering disaster prediction and prevention. However, although the numerical analysis method has a more reasonable calculation model, it is very sensitive to the mechanical parameters of geotechnical materials. Therefore, the mechanical parameters of geotechnical materials are an important factor affecting the accuracy of theoretical analysis and numerical simulation results in engineering applications. Therefore, it is of great significance to quickly, real-timely and accurately determine the real-time changes of the mechanical parameters of geotechnical materials for engineering design and construction as well as predict the stability of the project.
[0003] At present, scholars at home and abroad have proposed a variety of methods for determining geotechnical mechanics parameters, such as field test method and parameter inversion method. Due to the limitations of engineering costs, test sites and test time, the data available from field test methods are very limited, and it is impossible to provide accurate mechanical parameters for engineering design and construction and predict the stability of the project. The parameter inversion method can easily adjust the scale and accuracy of the measurement, has better environmental adaptability and other advantages, and has a wider application prospect in engineering practice. Therefore, conducting inversion method research and finding a fast, real-time and accurate method for obtaining the mechanical parameters of geotechnical materials is of great significance to the theoretical analysis and engineering application of geotechnical engineering. Summary of the invention
[0004] The technical problem to be solved by the present invention is to provide a method for identifying mechanical parameters of geotechnical materials based on a digital speckle correlation method and a finite element method, so that the geomechanical parameters of geotechnical materials can be obtained quickly, in real time and accurately.
[0005] The technical solution adopted by the present invention to solve the above technical problems is: a method for identifying mechanical parameters of geotechnical materials, characterized in that it comprises the following steps:
[0006] S1. Determine the displacement measurement value of the measuring point
[0007] The digital speckle correlation method is used as the experimental observation method, and the displacement field in the experimental process is obtained as the measurement value;
[0008] S2. Determine the displacement simulation value of the measuring point
[0009] According to the same specimen size and boundary conditions as the experiment, a finite element model was established, and the displacement field calculated by the finite element method was used as the simulation value;
[0010] S3. Determine the objective function:
[0011] The error between the displacement measurement value and the simulation value is taken as the objective function, that is:
[0012]
[0013] Where: P is the vector of mechanical parameters to be inverted, m is the number of points of interest, are the horizontal displacement response and vertical displacement response of the i-th point in the finite element model, are the horizontal displacement and vertical displacement of the ith point in the digital speckle displacement result, respectively.
[0014] S4. Select the mechanical parameters of the geotechnical material to be inverted and determine its value range, that is,
[0015] P1≤P≤P2
[0016] S5. Use the improved artificial fish swarm optimization algorithm to inversely solve the mechanical parameters of geotechnical materials, determine whether the global optimal solution is within the set error limit, or terminate the algorithm when the upper limit of the number of iterations is reached, and obtain the optimal value that can characterize the mechanical parameters of geotechnical materials.
[0017] S5, initialize the fish school, and initially set the algorithm parameters including the number of artificial fish N, the artificial fish visual field Visual, the crowding factor δ, the maximum number of attempts Try-number, and the maximum number of iterations Max-Iteration of the artificial fish school;
[0018] S6. Generate the same population size as set within the range of the mechanical parameters to be inverted in are the initial values of the mechanical parameters corresponding to each artificial fish, and the artificial fish group is normalized;
[0019] S7. Calculate the fitness value and objective function value of each artificial fish in the current fish school, select the behavior to be performed by each artificial fish, including foraging, flocking, chasing and random, and record the calculated optimal value, i.e. the minimum value of the objective function, on the bulletin board, then select a behavior that is closer to the optimal value, and update and generate a new fish school.
[0020] S8, if the objective function value corresponding to the artificial fish of the updated fish school is not better than the optimal value in the bulletin board, the bulletin board value is not changed, the artificial fish performs random behavior, and repeats step 7 until the number of iterations reaches the maximum number of iterations, then the calculation is stopped, and the optimal value in the current bulletin board, that is, the smallest artificial fish, is selected as the final result, and a denormalization process is performed on it. The artificial fish corresponding to the minimum objective function value is the mechanical parameter value of the geotechnical material;
[0021] In one embodiment, a method for identifying mechanical parameters of geotechnical materials comprises the following steps: using a digital speckle correlation method as an experimental observation means to obtain a displacement field in the experimental process as a measurement value; using a finite element method to obtain a displacement field as a simulation value; using the error between the displacement measurement value and the simulation value as an objective function; selecting and determining its value range; using an improved artificial fish swarm optimization algorithm to inversely determine the mechanical parameters of the geotechnical materials, determining whether the global optimal solution is within a set error limit or the algorithm ends when the upper limit of the number of iterations is reached, and obtaining the optimal value that can characterize the mechanical parameters of the geotechnical materials.
[0022] In one embodiment, the digital speckle correlation method specifically includes a CCD camera that can be mounted on a pan-tilt tripod, and the CCD camera is connected to a computer via a network cable.
[0023] In one embodiment, a CCD camera is used to continuously photograph the entire experimental process, and two-dimensional full-field displacement field measurement is achieved through a correlation coefficient function.
[0024] In one embodiment, the correlation coefficient function is:
[0025]
[0026] Where f(x,y) and g(x',y') are the grayscale of the image sub-area before and after deformation, respectively. and is the average grayscale in the sub-region, and M is the size of the sub-region.
[0027] According to formula (1), the positions of corresponding points in the reference image and the deformed image are matched to calculate the full-field two-dimensional displacement field.
[0028] In one embodiment, a finite element model is established and numerical analysis is performed according to the same sample size and boundary conditions as the experiment.
[0029] In one embodiment, the displacement results calculated by the finite element model are extracted as simulation values.
[0030] In one embodiment, the objective function is:
[0031]
[0032] Where: P is the vector of mechanical parameters to be inverted, m is the number of points of interest, are the horizontal displacement response and vertical displacement response of the i-th point in the finite element model, are the horizontal displacement and vertical displacement of the ith point in the digital speckle displacement result, respectively.
[0033] In one embodiment, the inverse calculation of mechanical parameters by improving the artificial fish swarm optimization algorithm specifically includes randomly generating a population size of the same number as the setting within the value range of the mechanical parameters to be inverted, calculating the objective function value of each individual in the population, selecting the behavior to be performed by each individual, and generating new individuals after updating until the global optimal solution reaches the set error limit or the algorithm ends when the upper limit of the number of iterations is reached.
[0034] The present invention adopts a method based on digital speckle correlation method and finite element method, combined with an improved artificial fish swarm optimization algorithm to obtain the mechanical parameters of geotechnical materials, which can achieve fast, real-time and accurate acquisition of the mechanical parameters of geotechnical materials, and is of great significance to the theoretical analysis and engineering application of geotechnical engineering. BRIEF DESCRIPTION OF THE DRAWINGS
[0035] Figure 1 is a flow chart of the method of the present invention;
[0036] Figure 2 This is a schematic diagram of the displacement calculation principle of the digital speckle correlation method of the present invention;
[0037] Figure 3 It is a flow chart of the improved fish school optimization algorithm of the present invention.
[0038] Figure 4 is a schematic diagram of the finite element model;
[0039] Figure 5 It is the displacement cloud diagram of the measured displacement field and the finite element simulation displacement cloud diagram of the inversion result (unit: mm)
[0040] Figure 6 It is the displacement field contour map of the measured displacement field and the inversion result finite element simulation displacement field (unit: mm)
[0041] Figure 7 It is a comparison chart of the vertical displacement measurement value of the horizontal measuring point and the finite element simulation value of the inversion result (unit: mm);
[0042] Figure 8 It is a comparison chart of the vertical displacement measurement value of the vertical measuring point and the finite element simulation value of the inversion result (unit: mm); DETAILED DESCRIPTION
[0043] In order to more clearly illustrate the implementation cases of the present invention and the technical solutions in the prior art, the present invention will be further described below in conjunction with specific examples and drawings.
[0044] The invention relates to a method for identifying mechanical parameters of geotechnical materials by using a digital speckle correlation method and a finite element method, which is suitable for quickly, real-time and accurately determining mechanical parameters of geotechnical materials in laboratories and on-site.
[0045] like Figure 1 As shown in FIG, the method includes: using photogrammetry as a deformation field measurement method, combining digital speckle correlation method to achieve full-field displacement measurement of the material to be inverted; according to the boundary conditions and geometric shape of the material to be inverted, establishing a corresponding numerical model in the finite element (such as Figure 4 ), read the displacement calculation results in the finite element result file through MATLAB calling PYTHON as the displacement simulation value. Establish the objective function according to the relationship between the displacement measurement value and the simulation value, and continuously adjust the mechanical parameters to be inverted in the finite element model by improving the artificial fish swarm algorithm, wherein the improved artificial fish swarm algorithm is implemented in MATLAB, and the modification of the mechanical parameters is realized by MATLAB calling PYTHON to modify the input file. Finally, the objective function composed of the displacement simulation value of the finite element model at the monitoring point and the experimental displacement measurement value is minimized, and the mechanical parameters of the material are characterized by the mechanical parameters in the finite element model at this time. The entire mechanical parameter inversion process is completed by MATLAB-PYTHON-ABAQUS. The other software is driven by MATLAB in a coordinated manner, a reasonable parameter range is set, and the inversion of mechanical parameters can be completed by optimizing the conditions.
[0046] Preferably, the objective function is:
[0047]
[0048] Where: P is the vector of mechanical parameters to be inverted, m is the number of points of interest, are the horizontal displacement response and vertical displacement response of the i-th point in the finite element model, are the horizontal displacement and vertical displacement of the ith point in the digital speckle displacement result, respectively.
[0049] like Figure 2As shown in the figure, the basic principle of the digital speckle correlation method is to obtain deformation information by tracking the movement of the same point in the images before and after deformation (referred to as the reference image and the deformed image, respectively). The distribution of a single pixel point has no regularity, so when calculating the displacement of a pixel point, an image sub-area centered on the point is selected to improve the accuracy of the displacement calculation. Therefore, by accurately matching the positions of the corresponding points in the reference image sub-area and the target image sub-area, the position and shape changes of the sub-area can be obtained, and these changes reflect the changes in the displacement of the object at that point.
[0050] In the reference image before deformation, there is a point P(x0,y0). With this point as the center, M pixels are extended to the surrounding areas. The enclosed area is the reference image sub-area. In the deformed image, the size of the reference image sub-area is searched in the deformed image, and the similarity of the two image sub-areas is calculated by the correlation function. The image sub-area corresponding to the extreme value of the correlation coefficient is the target image sub-area, and the difference between the central pixel coordinates of the reference image sub-area and the target image sub-area is the displacement information of the point.
[0051]
[0052] Preferably, the correlation coefficient function is:
[0053]
[0054] Where f(x,y) and g(x',y') are the grayscale of the image sub-area before and after deformation, respectively. and is the average grayscale in the sub-region, and M is the size of the sub-region.
[0055] According to formula (1), the positions of corresponding points in the reference image and the deformed image are matched to calculate the full-field two-dimensional displacement field.
[0056] like Figure 3The figure shows the optimization principle of the improved artificial fish swarm algorithm. The artificial fish swarm optimization algorithm simulates the foraging behavior of fish swarms to achieve optimization based on the characteristic that the place with the largest number of fish swarms in the water area is the place with the most nutrients in the water area. The algorithm mainly uses the four basic behaviors of fish: foraging, clustering, chasing and random behavior. It adopts a top-down optimization mode to start from the bottom-level behavior of the constructed individual, and through the local optimization of each individual in the fish swarm, the global optimal value is highlighted in the group. From the optimization mechanism of the standard artificial fish swarm algorithm, it can be found that the entire search process of the algorithm is based on the "memory" ability of each artificial fish and the information sharing mechanism between artificial fish swarms, so as to achieve a solution process with a certain degree of randomness but always able to achieve the optimal value. In the standard artificial fish swarm algorithm, the initial position of the artificial fish is a random value within the constraint range of the parameters to be optimized, but in the inversion of the mechanical parameters, there are orders of magnitude differences between the values of the four mechanical parameters, which will lead to inconsistent optimization processes of various mechanical parameters during the optimization process, and low efficiency in the optimization process of individual mechanical parameters. During the entire optimization process, due to the continuous changes in the individual extreme values and group optimal values of the artificial fish swarm, the algorithm has a strong search capability and high efficiency in the early stage. However, as the optimization process progresses, the closer it is to the optimal value of the mechanical parameter to be inverted, the smaller the optimization effect of the artificial fish swarm is, or even stagnant. The specific steps to improve the artificial fish swarm algorithm are as follows:
[0057] (1) Initialization settings, including population size, initial position of artificial fish, field of view of artificial fish, step length, crowding factor, number of repetitions, and maximum number of iterations;
[0058] (2) Normalize the initial position of the artificial fish. The standard artificial fish swarm algorithm does not consider the discreteness of parameters in the optimization problem. Since there are differences in the order of magnitude between the parameters to be inverted, it is necessary to normalize the initial position of the artificial fish and transform the positions of the parameters to be inverted to between [0,1] to avoid the impact of large differences in the order of magnitude of the data on the optimization results; convert the dimensional expression into a dimensionless expression to facilitate the comparison of parameters of different units and different magnitudes; and avoid numerical problems caused by too large numbers.
[0059]
[0060] In the formula, x i Represents the input parameter value to be inverted; x min Indicates the lower limit of the search for the inversion parameter; x max Indicates the upper limit of the search for the parameters to be inverted.
[0061] (3) Calculate the fitness value of each individual in the initial fish swarm, take the optimal artificial fish state and its value and assign it to the bulletin board;
[0062] (4) Evaluate each individual and select the behavior to be performed, including foraging, flocking, chasing, and random behavior; foraging behavior is a behavior that describes the tendency of fish to move toward food. They use vision to perceive the amount of food or food concentration in the water to choose the direction of action:
[0063]
[0064] In the formula, and They represent the states of the artificial fish’s attempts at the tth and t+1th times respectively; X j Indicates another state of the artificial fish; Step indicates the step length of the artificial fish; Rand indicates a random number between 0 and 1.
[0065] Schooling behavior describes the behavior of a large number of fish gathering in groups to collectively forage for food and avoid enemies. This is a way of survival that they have developed during the evolution process:
[0066]
[0067] Where, X c Indicates the center position of the artificial fish.
[0068] Tailing behavior describes that when a fish finds food, the fish nearby will follow, causing the fish farther away to follow. The artificial fish will explore the optimal position of the surrounding fish school. When the objective function value of the optimal position is greater than the objective function value of the current position and it is not very crowded, the current position will move one step toward the optimal fish position. The basic formula is the same as the grouping behavior formula.
[0069] Random behavior is a default behavior of foraging behavior, describing the random movement of the artificial fish within the field of view.
[0070]
[0071] (5) Add elimination strategies. In a group, individuals that can quickly adapt to the environment usually have higher fitness, so they have a higher survival rate than other individuals. Individuals with low fitness cannot quickly adapt to the environment and will be eliminated during the evolution process. Based on this, an elimination strategy is adopted to eliminate individuals with low fitness in all fish swarms, and then certain rules are adopted to generate new fish swarms based on fish swarms with high fitness to make up for the eliminated individuals with low fitness, so as to ensure the dynamic balance of the fish population size. In the elimination strategy, 80% of the fish swarms that dominate are controlled, and the remaining 20% of the fish swarms are eliminated in each iteration process, and the eliminated fish swarms are generated based on the 80% of the fish swarms. Through this process, the size of the fish swarms with higher fitness is expanded, and the fish swarms with lower fitness are eliminated, thereby accelerating the convergence speed of the algorithm iteration.
[0072] (6) Execute the behavior of artificial fish and generate new fish schools;
[0073] (7) Evaluate all individuals. If an individual is better than the billboard, update the billboard to that individual;
[0074] (8) Dynamically change the step size of each iteration. The step size of the basic artificial fish swarm algorithm is fixed. As an important parameter that affects the accuracy of the algorithm, the larger the step size, the lower the accuracy of the algorithm. The smaller the step size, the higher the accuracy of the algorithm, but it requires a lot of computing time. Therefore, the improved artificial fish swarm algorithm uses a larger step size in the early stage and a smaller step size in the later stage of the algorithm. The purpose of the improvement is to allow the artificial fish swarm to have a larger field of view in the early stage of the iteration, thereby accelerating the convergence speed of the iteration and enhancing the global search ability of the algorithm; the step size gradually decreases in the later stage of the iteration as the number of iterations increases, thereby enhancing the local search ability and optimization accuracy.
[0075] In a more preferred embodiment, the step length is calculated using the following formula:
[0076]
[0077] In the formula, MaxIteration represents the maximum number of iterations, t represents the current number of iterations, and step0 represents the initial step value.
[0078] Experiments have shown that by using the above formula, the change in step size can ensure a balance between the number of iterations and the change in iteration step size, thereby ensuring the calculation speed and accuracy.
[0079] (9) The algorithm ends when the optimal solution on the bulletin board is within the satisfactory error limit or reaches the upper limit of the number of iterations, otherwise go to step 3.
[0080] (10) Perform denormalization on the fish school on the bulletin board. Denormalization is the inverse operation of normalization. The actual value of each mechanical parameter is obtained through denormalization.
[0081] The present invention is further verified from an example below.
[0082] Similarity model test is a model test based on similarity theory. It is a method to study the laws by using the similarities and similar characteristics between things. The model and the prototype are similar not only in physical quantities, but also in physical essence. The main difference is that the size ratio of each physical quantity is different. Similarity model test can not only simulate many studies that cannot be realized in reality, but also the test method is repeatable. Therefore, it has become an important means of scientific research in recent years. The test takes the IV grade surrounding rock tunnel as the research target, and selects similar materials with barite, quartz sand, gypsum, detergent and water as raw materials to simulate the tunnel surrounding rock. The selected material ratio is barite: quartz sand: gypsum: detergent: water = 12:4:2:1:1. According to the principle of similarity, the requirements of physical and mechanical parameters such as stress, strain, elastic modulus, etc. are met. According to the size of the model bench, the geometric similarity ratio is selected as 100. According to the principle of similarity, the similarity ratios of various physical and mechanical parameters are as follows: bulk density similarity ratio C γ =1; Poisson's ratio similarity ratio C u =1; Similarity ratio of cohesion, stress and elastic modulus C C =C σ =C E =100.
[0083] The test system consists of a loading system and an image acquisition system. The loading system is a multi-channel electro-hydraulic servo similar simulation material test bench. The loading method is set by a computer program. The size of the bench model is 1080mm×1080mm×30mm. The plexiglass plate is fixed in front of the bench by bolts to simulate the tunnel plane strain model and facilitate observation. During the test, the test method of "digging the tunnel first and then loading" was adopted. The image acquisition system consists of a CCD camera, a light source, a computer, a tripod, a converter, etc. The converter is connected to the CCD camera and the computer through a data cable, and then connected to a power supply. The software for acquiring digital images in the computer is opened. The test piece is adjusted to the center of the digital image by adjusting the posture of the CCD camera lens and the height of the tripod. The CCD camera acquires the model speckle area image during the entire test process. The acquisition rate is 2 frames / s, and the image resolution is 1600×1200 pixels. The object plane resolution of the camera is determined according to the measurement accuracy. The object plane resolution of this experiment is 0.564mm / pixel, the measurement accuracy is 0.01pixel, and the actual displacement accuracy is 0.00564mm. The light source uses a white light source, which illuminates the surface of the test piece during the test to ensure that the grayscale of the image remains unchanged. The computer is the control center of the entire system, which acquires images in real time during the experiment, saves and processes the images, and outputs the final results. The test tunnel is placed in the center of the model, where the tunnel diameter D = 150mm, and the 450mm×450mm speckle area with the tunnel as the center is sprayed as the observation area. A layer of black primer is first evenly sprayed on the observation area. After the black primer is dried, white paint is sprayed on the black primer to form speckles. The main steps of the test are: After the materials weighed according to the ratio are stirred evenly, they are layered and paved from bottom to top in the test bench model box and compacted. In order to reduce the friction resistance between the model and the inner wall of the box, a layer of transparent film is pasted in the box and lubricated with grease. Then the steel template on the front side of the bench is removed, and the tunnel is excavated with iron catalpa according to the design, and the speckle is sprayed around the tunnel. After the spray paint is completely dried, the plexiglass plate is installed on the front side of the bench. Then set the loading mode of the loading system, debug the image acquisition system, let the test model be in the center of the image, and adjust the focal length and aperture of the lens to make the model image in the picture reach the clearest state, and finally start the loading system and image acquisition system at the same time to start data acquisition.
[0084] The model and monitoring point locations of similar model tests are as follows: Figure 6As shown. According to the specimen size and boundary conditions of the similar model test, a finite element model is established in ABAQUS, with the upper left corner of the model as the coordinate origin, the horizontal direction as the x direction, the vertical direction as the y direction, and the direction perpendicular to the paper as the z direction. The unit type of the finite element grid is C3D8. In the finite element model, the speckle area of the specimen is divided into a denser grid. The points marked in red are monitoring point sets, and the area where the monitoring point sets are located is the observation area. The monitoring point set is defined as a set for the subsequent output of monitoring point coordinates and displacement field data. The measurement value of the monitoring point set first needs to be calculated by DSCM to obtain the displacement field of the speckle image, and then the pixel coordinates of the monitoring point set on the speckle image are calculated according to the coordinates of the monitoring point set, the image resolution, and the object plane resolution in the finite element model. Finally, the displacement value of the pixel coordinates of the monitoring point set in the speckle image displacement field is obtained by interpolation through the interp2 function, which is the displacement measurement value.
[0085] The improved artificial fish swarm optimization algorithm is used to invert the mechanical parameters of similar models. The number of fish swarms of the fish swarm optimization algorithm is set to 8, the maximum number of iterations is 40, the maximum number of trials is 4, the perception distance is 0.1, the crowding factor is 0.618, and the step size is 0.05. The discrete problem of the parameters to be inverted is optimized by normalization. Since there are orders of magnitude differences between the four parameters to be inverted, the input data is normalized and the parameters to be inverted are transformed to [0,1] to avoid the impact of large differences in the order of magnitude of the data on the optimization results. After the inversion is completed, the data is denormalized. For the update of the step size, a larger step size is used in the early stage of iterative optimization, which is conducive to the algorithm to quickly jump out of the local optimal solution. A smaller step size is used in the later stage of iterative optimization, which is conducive to the algorithm to achieve higher accuracy.
[0086] The errors between the mechanical parameters of the similar model obtained by inversion using the mechanical parameter identification method of geotechnical materials based on digital speckle correlation method and finite element method and the reference values are shown in Table 1. The errors between the mechanical parameters obtained by inversion of the mechanical parameters of geotechnical materials based on digital speckle correlation method and finite element method and the reference values of the mechanical parameters of the model materials are very small. All the inversion mechanical parameter errors are within 1.5%, which is within the acceptable range.
[0087] Table 1 Inversion results of mechanical parameters after denormalization
[0088]
[0089] The displacement field of the specimen surface is obtained by the digital speckle correlation method. The mechanical parameters of the geotechnical materials (elastic modulus, Poisson's ratio, cohesion, and internal friction angle) obtained by inversion are used as the corresponding parameter inputs in the numerical simulation. The displacement field cloud map of the model surface in the finite element model is obtained through forward analysis, and the displacement measurement values and displacement simulation values of the measuring points in the horizontal and vertical directions are extracted for comparison. Figure 5-8 .
[0090] Depend on Figure 5-8 It can be seen that the mechanical parameters obtained by inversion based on the digital speckle correlation method and the finite element method have consistent distribution characteristics in the test displacement field cloud map and the numerical simulation displacement field cloud map, and are symmetrically distributed; the displacement magnitude is the same and decreases from top to bottom. The vertical displacements of the points on the vertical line from top to bottom are 1-12 and the points on the horizontal line from left to right are 13-24 in the test displacement field cloud map and the finite element simulation displacement field cloud map are extracted respectively. The measured values and numerical simulation values at the vertical measuring points and the horizontal measuring points are basically the same, with a difference of 0.05-0.2mm, a small difference, and basically the same change trend. The vertical displacement of the points on the vertical line is from top to bottom, from large to small; the vertical displacement of the points on the horizontal line is symmetrically distributed. It is verified that the mechanical parameter identification method of geotechnical materials based on the digital speckle correlation method and the finite element method is effective and reliable.
[0091] In summary, the present invention is based on the digital speckle correlation method and the finite element method, uses the digital speckle correlation method as an observation method, analyzes and obtains the displacement field as the measurement value, uses the displacement field obtained by the finite element method as the simulation value, uses the error between the displacement measurement value and the simulation value as the objective function, uses the artificial fish swarm optimization algorithm to optimize and solve the objective function, and proposes an inversion method for the mechanical parameters of geotechnical materials. The mechanical parameters of geotechnical materials in similar model tests were inverted and analyzed, the calculation results are reasonable, the inversion accuracy is high, the displacement measurement value cloud map and the simulation value cloud map are well matched, and the overall deformation characteristics of the model can be described. In addition, the measurement values and numerical simulation values at the vertical and horizontal measuring points are basically the same, and the change trends are consistent. This shows that this method is effective and reliable.
[0092] The above examples are only used to illustrate the design ideas and features of the present invention, and their purpose is to allow people to better understand the content of the present invention. The protection scope of the present invention is not limited to the above examples. Therefore, all equivalent changes or modifications made based on the principles and design ideas disclosed by the present invention are within the protection scope of the present invention.
Claims
1. A method for identifying mechanical parameters of geotechnical materials, comprising the following steps: 1) Using digital speckle correlation method as the experimental observation method, the displacement field in the experimental process is obtained as the measurement value; 2) The displacement field obtained by the finite element method is used as the simulation value; 3) Using the error between displacement measurement and simulation as the objective function; 4) Select the mechanical parameters of the geotechnical material to be inverted and determine its value range; 5) The improved artificial fish swarm optimization algorithm is used to inversely solve the mechanical parameters of geotechnical materials, and the algorithm is terminated when the global optimal solution is within the set error limit or the upper limit of the number of iterations is reached, so as to obtain the optimal value that can characterize the mechanical parameters of geotechnical materials; Among them, a CCD camera is used to continuously shoot the entire experimental process, and the two-dimensional full-field displacement field measurement is realized through the correlation coefficient function; The correlation coefficient function is: Where f(x,y) and g(x',y') are the grayscale of the image sub-area before and after deformation, respectively. and is the average grayscale in the sub-region, and M is the size of the sub-region; according to formula (1), the positions of the corresponding points in the reference image and the deformed image are matched to calculate the full-field two-dimensional displacement field; The objective function is: Where: P is the vector of mechanical parameters to be inverted, m is the number of points of interest, are the horizontal displacement response and vertical displacement response of the i-th point in the finite element model, are the horizontal displacement and vertical displacement of the ith point in the digital speckle displacement result respectively; The improved artificial fish swarm optimization algorithm includes: normalizing the initial position of the artificial fish; dynamically changing the step size of each iteration; the step size is calculated using the following formula Where MaxIteration represents the maximum number of iterations, and t represents the current number of iterations.
2. A method for identifying mechanical parameters of geotechnical materials according to claim 1, characterized in that: The digital speckle correlation method specifically includes a CCD camera that can be mounted on a pan / tilt tripod, and the CCD camera is connected to a computer via a network cable.
3. A method for identifying mechanical parameters of geotechnical materials according to claim 1, characterized in that: According to the same sample size and boundary conditions as the experiment, a finite element model was established and numerical analysis was carried out.
4. A method for identifying mechanical parameters of geotechnical materials according to claim 1, characterized in that: The displacement results calculated by the finite element model are extracted as simulation values.
5. A method for identifying mechanical parameters of geotechnical materials according to claim 1, characterized in that: The inverse analysis of mechanical parameters by improving the artificial fish swarm optimization algorithm specifically includes randomly generating a population size of the same number as the setting within the value range of the mechanical parameters to be inverted, calculating the objective function value of each individual in the population, selecting the behavior to be performed by each individual, and generating new individuals after updating until the global optimal solution reaches the set error limit or the algorithm ends when the upper limit of the number of iterations is reached.
