A method for predicting the behavior response of foundation soil in power engineering projects

Abstract

The present invention discloses a method for predicting the behavior response of foundation soil of an electric power project, the method comprising the following steps: step 1, collecting existing data and determining a monitoring arrangement scheme and a foundation soil behavior response calculation model; step 2, collecting on-site monitoring data through real-time feedback technology of monitoring data; step 3, using single-type monitoring data and multi-type monitoring data to perform Bayesian updates on soil parameters; step 4, combining the calculation model, predicting the behavior response of foundation soil based on the updated soil parameters. Compared with directly using the soil parameter values ​​measured by site survey, this method integrates monitoring data based on Bayesian theory, providing a possible way for prediction, early warning and ensuring the long-term safety of the foundation; currently, monitoring data is usually only used for simple comparison with the design warning value, and this method can make full use of monitoring data information to update the probability distribution of soil parameters, thereby more accurately predicting the long-term response of foundation soil of an electric power project.

Description

Technical Field

[0001] The invention relates to the technical field of civil engineering, and in particular to a method for predicting the behavior response of foundation soil in an electric power engineering project. Background Art

[0002] With the growth of economy and the acceleration of urbanization, the progress of power engineering construction is also accelerating. With the construction and development of high-voltage transmission lines in my country, in the process of building and using large power plants, buildings, and transmission towers, buildings (structures) are affected by their own loads, mechanical vibrations, groundwater level changes and other factors, and the foundation will have behavioral responses including foundation soil settlement and pore water pressure changes. Among them, the behavioral response of foundation soil settlement is the most important. When the uneven settlement is too large, it is very easy to cause the building (structure) to tilt, the structure to crack, and even cause related engineering accidents. Therefore, the calculation and analysis of foundation settlement is particularly important.

[0003] In the process of power engineering survey and construction, complex geological conditions are often encountered, the soil parameters in the underground area are complex, the groundwater distribution is uncertain, and the physical and mechanical properties of multiple geological structures, strata and soil are uncertain, making it difficult to determine the appropriate soil parameters and calculation model inputs for the foundation soil. The usual behavior response prediction method is to use the soil parameters measured by the site survey as input, combined with the calculation model, to calculate the behavior response such as the settlement of the foundation soil. However, the behavior response of the foundation soil predicted by this method is often quite different from the behavior response obtained by actual monitoring. In addition to the survey data, monitoring data is another type of data that can be obtained from the site. Introducing monitoring data on the basis of survey information provides a new possibility for prediction, early warning and long-term safety of the structure. However, the current monitoring data is usually only used for simple comparison with the design warning value, which is not helpful for improving the accuracy of the prediction of the behavior response such as the settlement of the foundation soil. In order to achieve the safety of the entire process of the project, it is the main technical problem solved by the present invention to make full use of the monitoring data to update the soil parameter values, so as to more accurately predict the long-term response of the foundation soil of the power engineering. Summary of the invention

[0004] The purpose of the present invention is to overcome the defects of the above-mentioned prior art and provide a behavioral response method for foundation soil settlement based on Bayesian update. The present invention uses field monitoring data and updates the probability density distribution of soil parameters based on Bayesian theory, thereby obtaining a real-time updated foundation soil settlement prediction and its probability distribution, ensuring the safe operation of engineering projects and improving safety prevention and control capabilities.

[0005] In order to achieve the above technical objectives, the present invention adopts the following technical solutions:

[0006] The present invention provides a method for predicting the behavior response of foundation soil in a power engineering project, which comprises the following steps:

[0007] Step 1: Collect existing data and determine the monitoring layout plan and the foundation soil behavior response standard method calculation model;

[0008] Step 2: Collect on-site monitoring data through real-time monitoring data feedback technology;

[0009] Step 3, using single-type monitoring data and multi-type monitoring data to perform Bayesian update on soil parameters;

[0010] Step 4: Combined with the foundation soil behavior response standard method calculation model, the foundation soil behavior response is predicted based on the updated soil parameters.

[0011] As a preferred embodiment of the present invention, the specific method of step 1 is as follows:

[0012] The prior distribution of soil parameters is obtained by integrating survey data, literature and engineering experience; the monitoring layout plan is determined based on the deformation of foundation soil and foundation settlement and the actual construction site profile;

[0013] The method for determining the calculation model of the code method for foundation soil behavior response is as follows: on the basis of the final settlement calculated by the code method, the consolidation of layered soil is considered.

[0014] As a preferred solution of the present invention, the monitoring arrangement solution includes foundation soil settlement monitoring and pore water pressure monitoring.

[0015] As a preferred solution of the present invention, the method for determining the foundation soil behavior response standard method calculation model is specifically as follows:

[0016] First, the standard method calculation formula is as follows:

[0017]

[0018] s-final deformation of foundation, mm;

[0019] s'-foundation deformation calculated by layer summation method;

[0020] ψ s - Empirical coefficient for settlement calculation;

[0021] n-the number of soil layers divided within the depth range of foundation settlement calculation;

[0022] P0-quasi-permanent value of additional pressure at the bottom of the foundation (kPa);

[0023] E si -Compression modulus of the i-th soil layer below the foundation bottom (MPa);

[0024] z i 、z i-1-Distance from the bottom of the foundation to the bottom of the i-th soil layer and the i-1-th soil layer (m);

[0025] - Average additional stress coefficient within the range from the calculation point of the foundation bottom to the bottom of the i-th soil layer and the i-1-th soil layer;

[0026] Secondly, on the basis of calculating the final settlement by the standard method, the consolidation of the foundation soil layer is considered, and the chemical equivalent layer method is used to calculate the standard method calculation model. The calculation method is as follows:

[0027] In a multi-layer soil layer with uniform layers, select a layer from each layer and the consolidation coefficient C ve As the parameter of the entire soil layer, change the thickness of other soil layers except this layer at the same time, and add up the changed soil layer thickness to get the total chemical equivalent layer thickness H. c , the actual soil layer is regarded as having a consolidation coefficient of C ve , thickness is H c Calculation of homogeneous soil layer;

[0028] For n layers of consolidated soil, the thickness and consolidation coefficient of each soil layer are H i and C vi (i=1,2,……,n); take the consolidation coefficient C of the first layer of soil (i=1) v1 As the consolidation parameter of the entire soil layer, C ve =C v1 ; Before changing the index, at a certain time t, the time factor of the i-th layer of soil (i = 2, 3, ..., n) is

[0029]

[0030] Let C vi =C v1 , and if we want to make it reach the same degree of consolidation at the same time t, then:

[0031]

[0032] In the formula, H i ' is the original thickness H i The thickness of solution (i=2, 3, ..., n);

[0033] The time factor T of soil samples with the same soil quality v Equal, so there is

[0034]

[0035]

[0036] Therefore, the chemical equivalent layer method can be used for multi-layer foundations. The consolidation of layered soil is calculated as a homogeneous soil layer, and the actual soil layer is regarded as a consolidation coefficient of C.ve =C v1 , thickness is Calculation of homogeneous soil layer.

[0037] As a preferred embodiment of the present invention, the step 2 comprises:

[0038] Intelligent online monitoring is used to achieve real-time, continuous and complete data collection, and the data include but are not limited to pore water pressure and foundation soil settlement.

[0039] As a preferred embodiment of the present invention, the specific method of step 3 is as follows:

[0040] Step 3.1: Combined with the prior distribution of soil parameters obtained in step 1, the probability density function determined by the prior information is the prior distribution of unknown soil parameters;

[0041] Step 3.2: Consider the monitoring information with measurement errors by establishing a likelihood function. For multi-type monitoring data, it is assumed that the measurement errors of each type of monitoring data are independent of each other. A likelihood function is established for each type of monitoring data, and a joint likelihood function is established, where the joint likelihood function is the product of each likelihood function.

[0042] Step 3.3: Obtain the posterior distribution and posterior statistical values ​​of soil parameters after Bayesian update through Markov chain Monte Carlo (MCMC) simulation.

[0043] As a preferred embodiment of the present invention, the updated soil parameters are deformation modulus and consolidation coefficient.

[0044] As a preferred embodiment of the present invention, the specific method of step 4 is as follows:

[0045] For the soil response specification method calculation model, the soil parameters obtained based on the single-type monitoring data or the multi-type monitoring data update are input respectively to obtain the corresponding foundation soil behavior response curve, which is compared with the existing monitoring data on site, so as to choose to use the single-type or multi-type monitoring data update scheme to predict the subsequent foundation soil behavior response.

[0046] As a preferred solution of the present invention, the posterior distribution of soil parameters in step 3.3 is the model input of the soil response standard method calculation model in step 4.

[0047] The present invention utilizes the online monitoring information of the Internet of Things to construct a monitoring information fusion algorithm based on Bayesian theory, dynamically updates the probability distribution of soil parameters, and realizes the update of subsequent foundation soil behavior response prediction based on continuously acquired monitoring data.

[0048] Compared with the prior art, the present invention has at least the following advantages:

[0049] (1) Aiming at the problem that it is difficult to accurately predict the foundation settlement of power engineering, a foundation settlement prediction method based on the Bayesian updating algorithm is proposed. The prior information of soil parameters is established from existing information such as survey data, literature and engineering experience; the likelihood function is constructed using monitoring data and a calculation model that takes into account multi-layer soil consolidation, and the posterior distribution of soil parameters is obtained in combination with the Markov chain Monte Carlo simulation method, thereby effectively updating the soil parameters. The present invention can make full use of multi-source information, effectively combine existing information and multi-type monitoring data, and thus accurately predict the settlement of power engineering buildings (structures).

[0050] (2) The present invention adopts intelligent online monitoring to realize a real-time, continuous and complete data acquisition system, which solves the data discontinuity and inefficiency problems of "few times and long cycles" in manual observation; and by utilizing the Internet of Things online monitoring data system, a monitoring information fusion algorithm based on Bayesian theory is constructed, combined with real-time monitoring data, to dynamically update the probability distribution of soil parameters and calculation model parameters, and realize the dynamic update of subsequent settlement and deformation prediction of foundation soil based on continuously acquired monitoring data, thereby realizing the intelligence and efficiency of engineering projects. BRIEF DESCRIPTION OF THE DRAWINGS

[0051] Figure 1 A technical roadmap of a method for predicting the behavior response of foundation soil in a power engineering project provided in an embodiment of the present invention;

[0052] Figure 2 A schematic diagram of soil deformation modulus and consolidation coefficient results using multiple types of monitoring data provided in an embodiment of the present invention;

[0053] Figure 3 A schematic diagram of a settlement prediction result using a single type of monitoring data provided in an embodiment of the present invention;

[0054] Figure 4 It is a schematic diagram of the settlement and pore water pressure prediction results using multiple types of monitoring data provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0055] The present invention is described in detail below in conjunction with the accompanying drawings and specific embodiments. Note that the following embodiments are merely illustrative in nature, and the present invention is not intended to limit its applicable objects or uses, and the present invention is not limited to the following embodiments.

[0056] Reference Figure 1 As shown, an embodiment of the present invention provides a method for predicting the behavior response of foundation soil in a power engineering project, comprising:

[0057] Step 1: Collect existing data, determine the monitoring layout plan and calculation model;

[0058] Step 2: Collect on-site monitoring data, such as pore water pressure and foundation soil settlement, through real-time monitoring data feedback technology;

[0059] Step 3, using single-type monitoring data and multi-type monitoring data to perform Bayesian update on soil parameters;

[0060] Step 4: Combined with the calculation model, the foundation soil behavior response is predicted based on the updated soil parameters.

[0061] Among them, in step 1, the prior distribution of soil parameters is obtained by integrating survey data, literature and engineering experience; for foundation soil deformation and foundation settlement, the monitoring layout plan including pore water pressure and settlement is determined in combination with the actual construction site profile. The calculation model is determined according to the monitoring layout plan. The specific method is to consider the consolidation of layered soil (Formula 2) on the basis of calculating the final settlement in the standard method (Formula 1). Among them, the consolidation principle of layered soil is mainly as follows:

[0062] Natural sedimentary soil generally has a layered structure. If the consolidation characteristics of each layer are quite different, it is advisable to consider it as a layered foundation. That is, according to the actual soil layer profile, the foundation is divided into several horizontal layers, and the soil properties of each layer are considered to be uniform. In actual engineering applications, the chemical equivalent layer method is often used for model calculation:

[0063] For uniform soil layers with the same soil quality but different thicknesses, the time factor T required to reach the same degree of consolidation is v The chemical equivalent layer method changes the thickness of the soil layer and treats the multi-layer soil layer as a single homogeneous soil layer. In the multi-layer soil layer with uniform layers, the consolidation coefficient C of one layer is selected from each layer. ve As the parameter of the entire soil layer, change the thickness of other soil layers except this layer at the same time, and add up the changed soil layer thickness to get the total chemical equivalent layer thickness H. c , then the actual soil layer is regarded as having a consolidation coefficient of C ve , thickness is H c Calculation of homogeneous soil layer.

[0064] Specifically, for n layers of consolidated soil, the thickness and consolidation coefficient of each soil layer are H i and C vi (i=1,2,……,n). Take the consolidation coefficient C of the first layer of soil (i=1) v1 As the consolidation parameter of the entire soil layer, C ve =C v1 Before changing the index, at a certain time t, the time factor of the i-th layer of soil (i = 2, 3, ..., n) is

[0065]

[0066] Let Cvi =C v1 , and if we want to make it reach the same degree of consolidation at the same time t, then:

[0067]

[0068] In the formula, H i ' is the original thickness H i The dissolution thickness (i=2, 3,…, n).

[0069] The time factor T of soil samples with the same soil quality v Equal, so there is

[0070]

[0071]

[0072] Therefore, the chemical equivalent layer method can be used for multi-layer foundations. The consolidation is calculated based on homogeneous soil layers, and the thickness of the soil layer is The consolidation coefficient is C ve =C v1 .

[0073] If there is a permeable layer in the stratification, the layer should be regarded as a free drainage surface, and the consolidation process of the soil layer between the two permeable layers should be calculated separately, and the compression amount at the same time should be superimposed.

[0074] In step 2, intelligent online monitoring is used to achieve real-time, continuous and complete data collection and data collection, which solves the data discontinuity and inefficiency problems of "few times and long cycles" in manual observation. While realizing the intelligent operation of the engineering project, multiple types of real-time monitoring data are obtained.

[0075] In step 3, relying on intelligent online monitoring, using the online monitoring information of the Internet of Things and the prior distribution of the soil parameters obtained, the monitoring information with measurement errors is considered through the likelihood function, and combined with the Markov Chain Monte Carlo (MCMC) simulation, a monitoring information fusion algorithm based on Bayesian theory is constructed. By combining real-time monitoring data, the probability posterior distribution of soil parameters and calculation model parameters is dynamically updated, so as to achieve the purpose of real-time updating of soil parameters. Among them, the specific Markov Chain Monte Carlo (MCMC) simulation is as follows:

[0076] The Markov chain Monte Carlo (MCMC) algorithm can be used to solve the Bayesian formula to obtain a posterior distribution that conforms to the target probability density distribution. The present invention adopts the Metropolis-Hastings (MH) sampling method, which has the advantage that it can extract samples from any probability distribution; it allows the Markov chain to combine certain acceptance / rejection criteria during the random sampling process, so that the Markov chain converges to obtain the target distribution.

[0077] The specific sampling process of the MH sampling method is as follows:

[0078] 1. Sampling from the prior distribution to obtain the initial sample Θ 0 ;

[0079] 2. When iteration step t = 1, 2, ...:

[0080] a) At step t, according to the proposed distribution Q(Θ * |Θ t-1 ), from which we randomly sample the proposed samples Θ * ;

[0081] b) Randomly sample from uniform distribution U[0,1] to generate a random number u;

[0082] c) Calculate the ratio r of the probability density

[0083]

[0084] d) Determine the acceptance rate α

[0085] α=min(r,1) (4)

[0086] e) Decide whether to accept the proposed sample based on the following criteria:

[0087]

[0088] In step 4, the posterior distribution of the parameters updated in step 3 is used to predict the subsequent behavioral responses such as foundation soil settlement based on the continuously acquired monitoring data.

[0089] The following is a specific example to illustrate the prediction method provided by the present invention:

[0090] This specific embodiment is a power project in Taizhou, Zhejiang.

[0091] First, the soil parameters and deformation modulus E are obtained by integrating survey data, literature and engineering experience. s and the consolidation coefficient c v Prior distribution. According to the actual situation of the project, the monitoring layout plan is determined. In this embodiment, 4 towers are selected as pilots to carry out the foundation soil settlement and deformation monitoring test; one measuring point is set for each foundation, and pore water pressure gauges and layered settlement gauges are arranged vertically in layers. The calculation model of this embodiment considers the consolidation of layered soil on the basis of calculating the soil settlement by the standard method. The settlement and pore pressure of the foundation soil are predicted by the calculation model.

[0092] Secondly, the real-time monitoring data (in this embodiment, mainly pore water pressure and settlement monitoring) is collected through the intelligent online monitoring system.

[0093] Then, based on the monitoring information fusion algorithm of Bayesian theory, single-type monitoring data and multi-type monitoring data fusion are considered respectively. When single-type monitoring data is used to predict settlement: according to step 1, the prior distribution of the deformation modulus and consolidation coefficient of the soil is determined. In this embodiment, a non-informative prior distribution is used; considering the measurement error of settlement monitoring, a likelihood function is established. In this embodiment, the settlement measurement error is taken as 0.001m and the pore pressure measurement error is taken as 1kPa; the posterior distribution of soil parameters (deformation modulus and consolidation coefficient) is obtained by Markov chain Monte Carlo (MCMC) simulation. When multi-type monitoring data is used to predict settlement: according to step 1, the prior distribution of the deformation modulus and consolidation coefficient of the soil is determined; for each monitoring data, a likelihood function considering the measurement error is established. Since different monitoring data come from different measuring instruments, it is considered that the measurement errors of settlement monitoring and pore pressure monitoring are independent of each other, and the overall likelihood function is the product of the likelihood functions corresponding to each type of monitoring; the posterior distribution of soil parameters (deformation modulus and consolidation coefficient) is obtained by MCMC simulation. This step, combined with the intelligent monitoring system, can realize the automatic update of soil parameters based on the continuously acquired monitoring data.

[0094] Finally, based on the computational model and combined with the updated posterior distribution of soil parameters, the settlement of foundation soil and the development of pore water pressure are predicted.

[0095] Figure 2 To update the results of soil deformation modulus and consolidation coefficient using settlement and pore water pressure monitoring. It can be seen that the variance of the posterior distribution is significantly smaller than the prior distribution, that is, the distribution area of ​​the posterior distribution is narrower. Therefore, this method can effectively reduce the uncertainty of soil parameters (deformation modulus and consolidation coefficient) by integrating field monitoring data.

[0096] Figure 3 This is the prediction of foundation soil settlement using a single type of monitoring data (settlement monitoring). It can be seen that the difference between the predicted mean and the monitoring data mean is small and the agreement is good. Therefore, it can be concluded that the use of a single type of monitoring data (settlement monitoring) can effectively predict foundation soil settlement.

[0097] Figure 4The monitored settlement and pore water pressure are combined by a multi-type information fusion method based on Bayesian theory to jointly update soil parameters, thereby predicting subsequent settlement and pore pressure changes. It can be seen that the predicted settlement mean and pore pressure mean are very close to the on-site monitoring values, and the prediction results are in good agreement with the monitoring results; the predicted settlement mean using multi-type monitoring data (monitored surface settlement and monitored pore water pressure) is closer to the monitoring value than using single-type monitoring data, and the confidence interval obtained using multi-type monitoring data is significantly narrower than the confidence interval obtained using single-type monitoring data, that is, the variance is smaller and the uncertainty is smaller. It can be explained that in this embodiment, using multi-type monitoring data is better than using single-type monitoring data.

Claims

1. A method for predicting the behavior response of foundation soil in power engineering, characterized in that: The steps include: Step 1: Collect existing data and determine the monitoring layout plan and the foundation soil behavior response standard method calculation model; The specific method of step 1 is as follows: The prior distribution of soil parameters is obtained by integrating survey data, literature and engineering experience; the monitoring layout plan is determined based on the deformation of foundation soil and foundation settlement and the actual construction site profile; The method for determining the calculation model of the foundation soil behavior response standard method is as follows: on the basis of calculating the final settlement by the standard method, the consolidation of the layered soil is considered; The method for determining the calculation model of the foundation soil behavior response standard method is specifically as follows: First, the calculation formula of the standard method is as follows: s-final deformation of foundation, mm; s'-foundation deformation calculated by layer summation method; ψ s - Empirical coefficient for settlement calculation; n-the number of soil layers divided within the depth range of foundation settlement calculation; P0-quasi-permanent value of additional pressure at the bottom of the foundation; E si -Compression modulus of the i-th layer of soil below the foundation bottom; z i 、z i-1 -The distance from the bottom of the foundation to the bottom of the i-th soil layer and the i-1-th soil layer; The average additional stress coefficient within the range from the calculation point of the foundation bottom to the bottom of the i-th soil layer and the i-1-th soil layer; Secondly, on the basis of calculating the final settlement by the standard method, considering the consolidation of the foundation soil layer, the chemical equivalent layer method is used to calculate the standard method calculation model. The calculation method is as follows: In a multi-layer soil layer with uniform layers, select a layer from each layer and the consolidation coefficient C ve As the parameter of the entire soil layer, change the thickness of other soil layers except this layer at the same time, and add up the changed soil layer thickness to get the total chemical equivalent layer thickness H. c , the actual soil layer is regarded as having a consolidation coefficient of C ve , thickness is H c Calculation of homogeneous soil layer; For n layers of consolidated soil, the thickness and consolidation coefficient of each soil layer are H i and C vi , i = 1, 2, ..., n; take the consolidation coefficient C of the first layer of soil v1 As the consolidation parameter of the entire soil layer, C ve =C v1 ; Before changing the index, at a certain time t, the time factor of the i-th layer of soil is Let C vi =C v1 , and if we want to make it reach the same degree of consolidation at the same time t, then: In the formula, H i ' is the original thickness H i The thickness of the solution; The time factor T of soil samples with the same soil quality v Equal, so there is Therefore, the chemical equivalent layer method can be used for multi-layer foundations. The consolidation of layered soil is calculated as a homogeneous soil layer, and the actual soil layer is regarded as a consolidation coefficient of C. ve =C v1 , thickness is Calculation of homogeneous soil layer; Step 2: Collect on-site monitoring data through real-time monitoring data feedback technology; Step 3, using single-type monitoring data and multi-type monitoring data to perform Bayesian update on soil parameters; The specific method of step 3 is as follows: Step 3.1: Combined with the prior distribution of soil parameters obtained in step 1, the probability density function determined by the prior information is the prior distribution of unknown soil parameters; Step 3.2: Consider the monitoring information with measurement errors by establishing a likelihood function. For multi-type monitoring data, it is assumed that the measurement errors of each type of monitoring data are independent of each other. A likelihood function is established for each type of monitoring data, and a joint likelihood function is established, where the joint likelihood function is the product of each likelihood function. Step 3.3: Obtain the posterior distribution and posterior statistical values ​​of soil parameters after Bayesian update through Markov chain Monte Carlo (MCMC) simulation; Step 4: Combined with the foundation soil behavior response standard method calculation model, the foundation soil behavior response is predicted based on the updated soil parameters.

2. A method for predicting the behavior response of foundation soil in a power engineering project according to claim 1, characterized in that: The monitoring arrangement scheme includes a foundation soil settlement monitoring arrangement scheme and a pore water pressure monitoring arrangement scheme.

3. A method for predicting the behavior response of foundation soil in a power engineering project according to claim 1, characterized in that: The step 2 comprises: Intelligent online monitoring is used to achieve real-time, continuous and complete data collection, and the data include but are not limited to pore water pressure and foundation soil settlement.

4. A method for predicting the behavior response of foundation soil in a power engineering project according to claim 1, characterized in that: The updated soil parameters are deformation modulus and consolidation coefficient.

5. A method for predicting the behavior response of foundation soil in a power engineering project according to claim 1, characterized in that: The specific method of step 4 is as follows: For the calculation model of the standard method for foundation soil behavior response, the soil parameters obtained based on the update of single-type monitoring data or multi-type monitoring data are input respectively to obtain the corresponding foundation soil behavior response curve, which is compared with the existing monitoring data on site, so as to choose to use the single-type or multi-type monitoring data update scheme to predict the subsequent foundation soil behavior response.

6. A method for predicting the behavior response of foundation soil in a power engineering project according to claim 1, characterized in that: The posterior distribution of soil parameters in step 3.3 is the model input of the foundation soil behavior response standard method calculation model in step 4.

Images