Method, device and equipment for predicting bearing capacity of offshore wind power single pile and medium

Abstract

The invention relates to the technical field of geotechnical engineering, and discloses a method, device and equipment for predicting the bearing capacity of an offshore wind power single pile and a medium, and the method comprises the steps that actual measurement data of different stratums where the single pile is located are obtained based on a pore pressure static sounding test; determining side friction resistance and unit pile end resistance of different stratums where the single pile is located based on measured data; the BP neural network model is trained based on preset single pile attribute data, the side friction resistance and the single pile end resistance, and a single pile bearing capacity prediction model is obtained; optimizing the single-pile bearing capacity prediction model based on a genetic algorithm to obtain a data processing type single-pile bearing capacity prediction model with gradually increased fitness; and solving the data processing type single-pile bearing capacity prediction model with gradually increased fitness to obtain an optimal solution of the single-pile bearing capacity. According to the method, the purpose that the accurate single-pile bearing capacity can be obtained only by conducting the pore pressure static sounding test subsequently is achieved, and the prediction accuracy is improved.

Description

Technical Field

[0001] The present invention relates to the field of geotechnical engineering technology, and in particular to a method, device, equipment and medium for predicting the bearing capacity of an offshore wind power single pile. Background Art

[0002] As global energy is becoming increasingly scarce, new renewable energy sources have received great attention. Power generation methods have always been a hot topic of research. The cost of solar power generation is relatively high, and the availability of hydropower generation is relatively saturated. Wind power generation is currently a more promising power generation method. Offshore wind power projects have performed very well in terms of the convenience of grid access, the stability of power generation, and land conservation. They are a very promising power generation industry. According to statistics, about 200 million kilowatts of wind power can be installed offshore. The offshore wind power generation market is stable, the wind speed is at an average level all year round and can be used for power generation, and the power generation of generators can be strongly guaranteed.

[0003] As a representative of micro-disturbance in-situ testing technology, piezocone penetration test (CPTU) has the advantages of being applicable to deep-water environments, fast and accurate, and obtaining a large amount of data, and is crucial in offshore survey applications. CPTU test results can not only be used to provide sea area geotechnical parameters, but also to predict the compressive and horizontal bearing capacity of pile foundations; in other words, the CPTU data interpretation results are an indispensable parameter source in the current offshore wind power pile foundation bearing capacity design calculation, and are of great significance to the optimization design of offshore wind turbine pile foundations. However, the current research on interpretation methods is still in its infancy and is difficult to meet engineering needs.

[0004] The penetration mechanism of CPTU technology is complex and the penetration theory is not clear enough, which makes the existing CPTU data interpretation method rely heavily on professional and technical personnel to speculate the value of the empirical coefficient based on their own experience; in addition, the amount of CPTU data is large and there are many influencing factors, while the empirical formula / coefficient has low accuracy due to the lack of theoretical basis, and is highly regional and greatly affected by human factors. Once the application area or scenario changes, the empirical formula cannot be actively updated, and there is a significant lag and application limitations.

[0005] At present, there are the following deficiencies in the prediction methods of offshore pile foundation bearing capacity: the pile foundation horizontal bearing capacity prediction method with publication number CN112926117A, a BAS-BP-based method, needs to rely on a large number of data samples, while the pile foundation bearing capacity data samples provided by the existing load tests of offshore wind power are relatively limited. Therefore, the insufficient number of samples will lead to inaccurate prediction results of the pile foundation bearing capacity; the method for predicting pile foundation bearing capacity based on multifunctional pore pressure static penetration technology with publication number CN101979783A relies on specifications and involves a large number of empirical coefficients, so it is impossible to accurately predict the pile foundation bearing capacity, and the prediction method does not involve intelligent algorithms, so the data processing is very complicated.

[0006] Therefore, a method is required to accurately predict the bearing capacity of a single offshore wind turbine pile based on a small number of actual values ​​from static load tests on offshore wind turbine pile foundations. Summary of the invention

[0007] In view of this, the present invention provides a method, device, equipment and medium for predicting the bearing capacity of an offshore wind power single pile, so as to solve the problem of inaccurate prediction of the bearing capacity of a single pile due to insufficient measured sample data.

[0008] In a first aspect, the present invention provides a method for predicting the bearing capacity of an offshore wind power single pile, the method comprising:

[0009] Obtain measured data of different strata where the single pile is located based on pore pressure static penetration test;

[0010] Determine the side friction resistance and unit pile end resistance of different strata where the single pile is located based on measured data;

[0011] The BP neural network model is trained based on the preset single pile attribute data, side friction resistance and single pile end resistance to obtain a single pile bearing capacity prediction model;

[0012] The single pile bearing capacity prediction model is optimized based on genetic algorithm, and a data processing type single pile bearing capacity prediction model with gradually increasing fitness is obtained;

[0013] The data processing type single pile bearing capacity prediction model with gradually increasing fitness is solved to obtain the optimal solution of the single pile bearing capacity.

[0014] The method for predicting the bearing capacity of a single offshore wind power pile provided by the present invention combines the relevant characteristics of the stratum in the area where the project is located, obtains the measured data of different strata where the single pile is located based on the pore pressure static penetration test, and predicts the bearing capacity of the single pile based on the BP neural network and the GA (Genetic Algorithm, GA) optimization algorithm (i.e., the GA-BP neural network), thereby achieving the purpose of obtaining a relatively accurate single pile bearing capacity by only performing a subsequent pore pressure static penetration test, improving the accuracy of the prediction, and solving the problem of inaccurate prediction of the single pile bearing capacity due to the small amount of measured sample data.

[0015] In an optional embodiment, the measured data include cone tip resistance, side wall friction resistance and pore water pressure;

[0016] Determining the side friction resistance and unit pile end resistance of different strata of a single pile based on measured data includes:

[0017] Determine the preliminary modified cone tip resistance based on the cone tip resistance and pore water pressure;

[0018] Determine the effective cone tip resistance based on the preliminary corrected cone tip resistance and pore water pressure;

[0019] The side friction resistance is determined based on the side wall friction resistance, and the single pile end resistance is determined based on the effective cone tip resistance.

[0020] The method for predicting the bearing capacity of a single offshore wind power pile provided by the present invention converts the cone tip resistance, side wall friction resistance and pore water pressure of the measured data into the side friction resistance and end resistance of the single pile, providing a data basis for the subsequent calculation of the bearing capacity of the single pile.

[0021] In an optional implementation, the preset single pile attribute data includes a single pile diameter and a single pile length;

[0022] The BP neural network model is trained based on the preset single pile attribute data, side friction resistance and single pile end resistance, and the single pile bearing capacity prediction model is obtained, including:

[0023] The single pile diameter, single pile length, side friction resistance and single pile end resistance are taken as the data sample set;

[0024] Based on the data sample set and using the gradient descent method, the BP neural network model is trained to obtain a single pile bearing capacity prediction model.

[0025] The method for predicting the bearing capacity of a single offshore wind power pile provided by the present invention uses the single pile diameter, single pile length, side friction resistance and single pile end resistance as a data sample set; based on the data sample set and using the gradient descent method, a BP neural network model is trained to obtain a single pile bearing capacity prediction model, thereby achieving the purpose of training the BP neural network model with a small data sample set and then obtaining a single pile bearing capacity prediction model, and providing conditions for subsequent prediction of the single pile bearing capacity.

[0026] In an optional implementation, the BP neural network model is trained based on a data sample set and using a gradient descent method to obtain a single pile bearing capacity prediction model, including:

[0027] Divide the data sample set into a training sample set and a verification sample set;

[0028] Based on the training sample set, the BP neural network model is trained using the gradient descent method to obtain the training results;

[0029] Input the verification sample set into the BP neural network model for verification to obtain the verification result;

[0030] The error between the training result and the verification result is calculated. When the error does not decrease or increase after a preset number of consecutive times, the training of the BP neural network model is stopped, and the BP neural network model obtained after the last training is used as the single pile bearing capacity prediction model.

[0031] The method for predicting the bearing capacity of a single offshore wind power pile provided by the present invention verifies the BP neural network model through a verification sample set, calculates the error between the training result and the verification result, stops training the BP neural network model when the error does not decrease or increase after a preset number of consecutive times, and uses the BP neural network model obtained after the last training as a single pile bearing capacity prediction model, thereby improving the training speed and training accuracy of the BP neural network model.

[0032] In an optional implementation, the single pile bearing capacity prediction model is expressed by the following formula:

[0033] Q ult =Q s +Q b =Σf pi A si +q b A b

[0034] Among them, Q ult is the ultimate bearing capacity of a single pile, Q s is the pile side bearing capacity, Q b is the pile end bearing capacity, f pi is the unit lateral friction of the ith layer of the stratum where the single pile is located, A si is the lateral area of ​​a single pile in the i-th stratum, q b is the unit pile end resistance, A b is the pile tip area, which is calculated from the single pile diameter and single pile length.

[0035] In an optional implementation, the single pile bearing capacity prediction model is optimized based on a genetic algorithm to obtain a data processing type single pile bearing capacity prediction model with gradually increasing fitness, including:

[0036] Evaluate the applicability of the single pile bearing capacity prediction model based on the error;

[0037] The fitness is used to optimize the symbol string population in the evolution process of the genetic algorithm, and a data processing single pile bearing capacity prediction model with gradually increasing fitness is obtained.

[0038] The method for predicting the bearing capacity of a single pile of offshore wind power provided by the present invention adds a genetic algorithm to a BP neural network model, evaluates the applicability of the single pile bearing capacity prediction model through errors in a verification process, optimizes the symbol string population in the evolution process of the genetic algorithm using the applicability, and obtains a data processing type single pile bearing capacity prediction model with gradually increasing fitness, thereby achieving optimization of the single pile bearing capacity prediction model, improving the prediction accuracy of the single pile bearing capacity, and avoiding the problem of low prediction accuracy due to the lack of theoretical basis for empirical formulas or coefficients.

[0039] In an optional implementation, solving the data processing type single pile bearing capacity prediction model with gradually increasing fitness to obtain the optimal solution of the single pile bearing capacity includes:

[0040] Determine the number of input layer nodes, hidden layer nodes and output layer nodes of the data processing type single pile bearing capacity prediction model with gradually increasing fitness;

[0041] Based on the number of input layer nodes, hidden layer nodes and output layer nodes, a genetic algorithm is used to perform a global search on the data processing type single pile bearing capacity prediction model with gradually increasing fitness, and the weight and threshold with the best fitness are obtained.

[0042] Based on the weight and threshold with the best fitness, the local optimization algorithm is used to calculate the optimal solution of the single pile bearing capacity.

[0043] The method for predicting the bearing capacity of a single offshore wind power pile provided by the present invention determines the number of input layer nodes, the number of hidden layer nodes and the number of output layer nodes of a data processing type single pile bearing capacity prediction model with gradually increasing fitness, thereby shortening the prediction time of the single pile bearing capacity; based on the number of input layer nodes, the number of hidden layer nodes and the number of output layer nodes, a genetic algorithm is used to perform a global search on the data processing type single pile bearing capacity prediction model with gradually increasing fitness, thereby obtaining a weight and a threshold value with the best fitness; based on the weight and the threshold value with the best fitness, a local optimization algorithm is used to calculate and obtain an optimal solution for the bearing capacity of the single pile; based on the prediction of the bearing capacity of the pile foundation by a BP neural network, a GA genetic algorithm is used to perform optimization, thereby obtaining an optimal solution for the bearing capacity of the single pile with high accuracy, thereby improving the prediction accuracy.

[0044] In a second aspect, the present invention provides a device for predicting the bearing capacity of an offshore wind power single pile, the device comprising:

[0045] A measured data acquisition module is used to obtain measured data of different strata where the single pile is located based on the pore pressure static penetration test;

[0046] Resistance calculation module, used to determine the side friction resistance and unit pile end resistance of different strata where the single pile is located based on measured data;

[0047] A model training module is used to train the BP neural network model based on preset single pile attribute data, side friction resistance and single pile end resistance to obtain a single pile bearing capacity prediction model;

[0048] A model optimization module is used to optimize the single pile bearing capacity prediction model based on a genetic algorithm to obtain a data processing type single pile bearing capacity prediction model with gradually increasing fitness;

[0049] The model solving module is used to solve the data processing type single pile bearing capacity prediction model with gradually increasing fitness to obtain the optimal solution of the single pile bearing capacity.

[0050] In a third aspect, the present invention provides a computer device, comprising: a memory and a processor, the memory and the processor are communicatively connected to each other, the memory stores computer instructions, and the processor executes the method for predicting the bearing capacity of a single offshore wind power pile according to the first aspect or any corresponding embodiment thereof by executing the computer instructions.

[0051] In a fourth aspect, the present invention provides a computer-readable storage medium having computer instructions stored thereon, the computer instructions being used to enable a computer to execute the method for predicting the bearing capacity of a single offshore wind power pile according to the first aspect or any corresponding embodiment thereof. BRIEF DESCRIPTION OF THE DRAWINGS

[0052] In order to more clearly illustrate the specific implementation methods of the present invention or the technical solutions in the prior art, the drawings required for use in the specific implementation methods or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are some implementation methods of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0053] Figure 1 is a flow chart of a method for predicting the bearing capacity of an offshore wind power single pile according to an embodiment of the present invention;

[0054] Figure 2 is a flow chart of another method for predicting the bearing capacity of an offshore wind power single pile according to an embodiment of the present invention;

[0055] Figure 3 is a flow chart of another method for predicting the bearing capacity of a single offshore wind power pile according to an embodiment of the present invention;

[0056] Figure 4 is a flow chart of another method for predicting the bearing capacity of a single offshore wind power pile according to an embodiment of the present invention;

[0057] Figure 5 Schematic diagram of a neural network according to an embodiment of the present invention

[0058] Figure 6 is a structural block diagram of a device for predicting the bearing capacity of an offshore wind power single pile according to an embodiment of the present invention;

[0059] Figure 7 It is a schematic diagram of the hardware structure of a computer device according to an embodiment of the present invention. DETAILED DESCRIPTION

[0060] In order to make the purpose, technical solution and advantages of the embodiments of the present invention clearer, the technical solution in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative work are within the scope of protection of the present invention.

[0061] According to an embodiment of the present invention, an embodiment of a method for predicting the bearing capacity of a single offshore wind turbine pile is provided. It should be noted that the steps shown in the flowchart of the accompanying drawings can be executed in a computer system such as a set of computer executable instructions, and although a logical order is shown in the flowchart, in some cases, the steps shown or described can be executed in an order different from that shown here.

[0062] In this embodiment, a method for predicting the bearing capacity of an offshore wind power single pile is provided, which can be used in a mobile terminal, such as a central processing unit, a server, etc. Figure 1 is a flow chart of a method for predicting the bearing capacity of an offshore wind power single pile according to an embodiment of the present invention. Figure 1 As shown, the process includes the following steps:

[0063] Step S101, obtaining measured data of different strata where the single pile is located based on the pore pressure static penetration test.

[0064] Specifically, the pore pressure static penetration test is used to combine the sensor element for measuring pore water pressure with the standard static penetration test, so that the pore water pressure can be measured at the same time as the penetration resistance; when the penetration is stopped, the dissipation of excess pore water pressure can also be measured. The penetration resistance includes the cone tip resistance q of different formations c and side wall friction f s The cone tip resistance q of the single pile in different strata is obtained through the pore pressure static penetration test. c , side wall friction f s And measured data such as pore water pressure u2.

[0065] It should be noted that while obtaining the measured data of different strata where the single pile is located, the corresponding real-time measured penetration depth must be recorded. The penetration depth can roughly determine the thickness of different strata. The changes in the penetration resistance of the same strata are basically consistent. If the change is too large, the change of the strata can be judged. The purpose of recording the penetration depth is to roughly determine the location of different strata.

[0066] Step S102, determining the side friction resistance and unit pile end resistance of different strata where the single pile is located based on the measured data.

[0067] Specifically, based on the cone tip resistance q c , side wall friction f s The side friction resistance and unit pile end resistance of the single pile in different strata are calculated based on the measured data such as pore water pressure u2.

[0068] Step S103, training the BP neural network model based on the preset single pile attribute data, the side friction resistance and the single pile end resistance to obtain a single pile bearing capacity prediction model.

[0069] Specifically, the single pile attribute data includes the single pile diameter, single pile length, pile body inclination, pile top elevation, pile tip elevation and single pile number, etc. A data sample set is constructed, and the pile diameter, pile length, single pile end resistance and modified side friction resistance are used as the input values ​​of the BP neural network model. The BP neural network model is trained to obtain a single pile bearing capacity prediction model based on the BP neural network model. The single pile bearing capacity is also called the single pile vertical bearing capacity.

[0070] Step S104, optimizing the single pile bearing capacity prediction model based on a genetic algorithm to obtain a data processing type single pile bearing capacity prediction model with gradually increasing fitness.

[0071] Specifically, the single pile bearing capacity prediction model is optimized by using a GA (Genetic Algorithm, GA) genetic algorithm, specifically: on the basis of the single pile bearing capacity prediction model in the field, a mathematical method is used to convert the solution process of the single pile bearing capacity prediction model into a process similar to the crossover and mutation of chromosome genes in biological evolution by using computer simulation calculations, so as to generate a GMDH (Group Method of Data Handling) type neural network model with a gradually increasing fitness Φ, that is, a data processing type single pile bearing capacity prediction model with a gradually increasing fitness.

[0072] Step S105, solving the data processing type single pile bearing capacity prediction model with gradually increasing fitness to obtain the optimal solution for the single pile bearing capacity.

[0073] Specifically, based on the data processing single pile bearing capacity prediction model with gradually increasing fitness, the existing data population fitting effect can reach the optimal solution when the data volume is relatively limited, so that a more accurate single pile bearing capacity can be obtained from the CPTU test data in the site.

[0074] The method for predicting the bearing capacity of a single offshore wind power pile provided in this embodiment combines the relevant characteristics of the strata in the area where the project is located, obtains the measured data of different strata where the single pile is located based on the pore pressure static penetration test, and predicts the bearing capacity of the single pile based on the BP neural network and GA (Genetic Algorithm, GA) optimization algorithm (i.e., GA-BP neural network), thereby achieving the purpose of obtaining a relatively accurate single pile bearing capacity by only conducting a subsequent pore pressure static penetration test, improving the accuracy of the prediction, and solving the problem of inaccurate prediction of the single pile bearing capacity due to the small amount of measured sample data.

[0075] In this embodiment, a method for predicting the bearing capacity of an offshore wind power single pile is provided, which can be used in a mobile terminal, such as a central processing unit, a server, etc. Figure 2 is a flow chart of a method for predicting the bearing capacity of an offshore wind power single pile according to an embodiment of the present invention. Figure 2 As shown, the process includes the following steps:

[0076] Step S201, based on the pore pressure static penetration test, obtain the measured data of different strata where the single pile is located. Figure 1 Step S101 of the illustrated embodiment will not be described in detail here.

[0077] Step S202, determining the side friction resistance and unit pile end resistance of different strata where the single pile is located based on the measured data.

[0078] Specifically, the measured data include the cone tip resistance q c , side wall friction f s and pore water pressure u2, the measured data are corrected and preprocessed to obtain the preliminary corrected cone tip resistance q t and effective cone tip resistance q E , thereby determining the modified single pile lateral friction f p and the pile end resistance q b , and collect the ultimate bearing capacity Q of the single pile at the corresponding drilling hole ult Establish the single pile lateral friction f p and the pile end resistance q b The ultimate bearing capacity of a single pile Q ult Preliminary relationship between values.

[0079] Specifically, the above step S202 includes:

[0080] Step S2021, determining a preliminary corrected cone tip resistance based on the cone tip resistance and the pore water pressure.

[0081] Specifically, the initial correction of the cone tip resistance q t It is expressed by the following formula:

[0082] q t =q c +u2(1-a) (1);

[0083] Among them, q t is the initial modified cone tip resistance, q c is the cone tip resistance, u2 is the pore water pressure, and a is the net area ratio of the penetrometer used in the pore pressure static penetration test.

[0084] Step S2022, determining the effective cone tip resistance based on the preliminary corrected cone tip resistance and the pore water pressure.

[0085] Specifically, the effective cone tip resistance q E It is expressed by the following formula:

[0086] q E =q t -u2 (2);

[0087] Among them, q E is the effective cone tip resistance.

[0088] Step S2023, determining the side friction resistance based on the side wall friction resistance, and determining the single pile end resistance based on the effective cone tip resistance.

[0089] Specifically, considering the influence of the single pile size effect, the modified single pile side friction f p It can be calculated by combining the empirical formula in the Technical Specification for Building Pile Foundations (JGJ94-2008). Single pile end resistance q b Within the depth range affected by the pile tip, q E The geometric mean of is obtained, namely:

[0090]

[0091] Where n is the number of strata that the probe passes through. E1 q E2 q E3 ……q En They are the effective cone tip resistance of the first layer, the second layer, the third layer and the nth layer respectively.

[0092] Step S203, training the BP neural network model based on the preset single pile attribute data, the side friction resistance and the single pile end resistance to obtain a single pile bearing capacity prediction model.

[0093] Specifically, the preset single pile attribute data includes the single pile diameter and the single pile length. The above step S202 includes:

[0094] Step S2031, taking the single pile diameter, single pile length, side friction resistance and single pile end resistance as a data sample set.

[0095] Specifically, the single pile diameter, single pile length, side friction resistance and single pile end resistance are input into the BP neural network model as a data sample set.

[0096] Step S2032: Based on the data sample set, the BP neural network model is trained using the gradient descent method to obtain a single pile bearing capacity prediction model.

[0097] In some optional embodiments, such as Figure 4 As shown, the above step S2032 includes:

[0098] Step a1: divide the data sample set into a training sample set and a validation sample set.

[0099] Specifically, when building a model for the vertical pile bearing capacity of the target site, the number of epochs that the entire training sample set training is completely traversed by the BP neural network model cannot be less than the total number of single pile data sample sets. Epoch is the training cycle, which represents the verification error check of consecutive training cycles.

[0100] Step a2: training the BP neural network model based on the training sample set and using the gradient descent method to obtain a training result.

[0101] Specifically, the function Gradient of the gradient descent method is used in the training process and in the iterative verification process. During training, the training sample set training set is used for training, and a training result is obtained after each training.

[0102] Step a3: input the verification sample set into the BP neural network model for verification to obtain the verification result.

[0103] Specifically, each time training is performed, the sample data in the validation set is automatically input into the neural network for verification to obtain a verification result.

[0104] Step a4, calculate the error between the training result and the verification result. When the error does not decrease or increase after a preset number of consecutive times, stop training the BP neural network model, and use the BP neural network model obtained after the last training as a single pile bearing capacity prediction model.

[0105] Specifically, after the validation set validation set is input, an error will be obtained based on the training results and the validation results. Previously, a step number must be set for the validation set validation set. For example, the default is 6 epochs (6 validation error checks). Then, it is determined whether the error does not decrease after 6 consecutive tests. If it does not decrease or even increases, it means that the error of the training set training set training set is no longer decreasing. At this time, the training needs to be stopped, otherwise it may fall into over-learning. The BP neural network model obtained after the last training is used as a single pile bearing capacity prediction model. Among them, epoch is the training cycle, which represents the validation error check of 6 consecutive training cycles (6 epochs). If the validation error does not improve after these 6 validation checks, the training process will automatically stop to prevent the model from overfitting.

[0106] The single pile bearing capacity prediction model is expressed as follows:

[0107] Q ult =Q s +Q b =Σf pi A si +q b A b

[0108] Among them, Q ult is the ultimate bearing capacity of a single pile, Q s is the pile side bearing capacity, Q b is the pile end bearing capacity, f pi is the unit lateral friction of the ith layer of the stratum where the single pile is located, A si is the lateral area of ​​a single pile in the i-th stratum. The force per unit area multiplied by the unit lateral area gives the surface force. The circumference of a single pile multiplied by the stratum thickness gives the lateral area of ​​the single pile. b is the single pile end resistance, A b is the pile tip area, which is calculated from the single pile diameter and single pile length.

[0109] The method for predicting the bearing capacity of a single offshore wind power pile provided in this embodiment verifies the BP neural network model through a verification sample set, calculates the error between the training result and the verification result, and stops training the BP neural network model when the error does not decrease or increase after a preset number of consecutive times. The BP neural network model obtained after the last training is used as a single pile bearing capacity prediction model, thereby improving the training speed and training accuracy of the BP neural network model.

[0110] Step S204: Optimize the single pile bearing capacity prediction model based on the genetic algorithm to obtain a data processing type single pile bearing capacity prediction model with gradually increasing fitness. Figure 1Step S104 of the illustrated embodiment will not be described in detail here.

[0111] Step S205, solving the data processing type single pile bearing capacity prediction model with gradually increasing fitness to obtain the optimal solution for the single pile bearing capacity. Figure 1 Step S105 of the illustrated embodiment will not be described in detail here.

[0112] The method for predicting the bearing capacity of a single offshore wind power pile provided in this embodiment converts the cone tip resistance, side wall friction resistance and pore water pressure of the measured data into the side friction resistance and end resistance of the single pile, providing a data basis for the subsequent calculation of the bearing capacity of the single pile, with the single pile diameter, single pile length, side friction resistance and end resistance of the single pile as the data sample set; based on the data sample set and using the gradient descent method to train the BP neural network model, a single pile bearing capacity prediction model is obtained, achieving the purpose of training the BP neural network model with a small data sample set, and then obtaining the single pile bearing capacity prediction model, providing conditions for the subsequent prediction of the single pile bearing capacity.

[0113] In this embodiment, a method for predicting the bearing capacity of an offshore wind power single pile is provided, which can be used in a mobile terminal, such as a central processing unit, a server, etc. Figure 3 is a flow chart of a method for predicting the bearing capacity of an offshore wind power single pile according to an embodiment of the present invention. Figure 3 As shown, the process includes the following steps:

[0114] Step S301, based on the pore pressure static penetration test, obtain the measured data of different strata where the single pile is located. Figure 2 Step S201 of the illustrated embodiment will not be described in detail here.

[0115] Step S302: Determine the side friction resistance and unit pile end resistance of different strata where the single pile is located based on the measured data. Figure 2 Step S202 of the illustrated embodiment will not be described in detail here.

[0116] Step S303: training the BP neural network model based on the preset single pile attribute data, side friction resistance and single pile end resistance to obtain a single pile bearing capacity prediction model. Figure 2 Step S203 of the illustrated embodiment will not be described in detail here.

[0117] Step S304, optimizing the single pile bearing capacity prediction model based on a genetic algorithm to obtain a data processing type single pile bearing capacity prediction model with gradually increasing fitness.

[0118] Specifically, the above step S304 includes:

[0119] Step S3041, evaluating the applicability of the single pile bearing capacity prediction model based on the error.

[0120] Specifically, the genetic algorithm is combined with the design of the BP neural network model. The fitness Φ of each whole string of symbolic numbers representing the BP neural network model is evaluated in the following form: Φ = 1 / E, where E is the mean square error. The fitness Φ is maximized to minimize the evolution process. Among them, each whole string of symbolic numbers represents the encoding of the BP neural network model, which is an encoding of the BP neural network structure and parameters, and is used for the individual representation of the genetic algorithm. At the beginning of the genetic algorithm, a population is randomly generated, and each population member is a symbolic string, and each symbolic string represents a BP neural network model.

[0121] Step S3042, using the fitness to optimize the symbol string population in the evolution process of the genetic algorithm, to obtain a data processing type single pile bearing capacity prediction model with gradually increasing fitness.

[0122] Specifically, Figure 4 As shown in the figure, the genetic algorithm evolution process starts with the initial population of randomly generated symbol strings, each of which is a candidate solution. Then, the entire symbol string population is gradually improved by using the roulette wheel to perform genetic operations of selection, crossover and mutation. By optimizing the symbol string population in the genetic algorithm evolution process using fitness, a GMDH-type neural network model with gradually increasing fitness Φ is generated, that is, a data processing-type single pile bearing capacity prediction model with gradually increasing fitness, until no further significant improvement can be achieved, which can make the existing data population fitting effect reach the optimal solution when the amount of data is relatively limited.

[0123] Step S305, solving the data processing type single pile bearing capacity prediction model with gradually increasing fitness to obtain the optimal solution for the single pile bearing capacity.

[0124] Specifically, the above step S305 includes:

[0125] Step S3051, determining the number of input layer nodes, the number of hidden layer nodes and the number of output layer nodes of the data processing type single pile bearing capacity prediction model with gradually increasing fitness.

[0126] Specifically, Figure 5 As shown in the figure, GA genetic algorithm is used for optimization based on the prediction of pile foundation bearing capacity by BP neural network (i.e. GA-BP neural network). The optimized structure is divided into input layer, hidden layer and output layer. The number of nodes in the input layer depends on the type of input parameters. The number of nodes in the hidden layer determines the training time and prediction accuracy of the neural network. The number of nodes is first selected according to the following formula:

[0127] l<n-1 (4);

[0128]

[0129] l = log2 n (6);

[0130] Then use the trial and error method to further select and finally determine the optimal number of nodes.

[0131] Among them, n is the number of input layer nodes; l is the number of hidden layer nodes; m is the number of output layer nodes; a is a constant between 0 and 10.

[0132] Step S3052, based on the number of input layer nodes, the number of hidden layer nodes and the number of output layer nodes, a genetic algorithm is used to perform a global search on the data processing type single pile bearing capacity prediction model with gradually increasing fitness to obtain the weight and threshold with the best fitness.

[0133] Specifically, weights are parameters that connect the neurons in the BP neural network model. Each connection is assigned a weight to determine the degree of influence of the input value on the output. During the training process of the BP neural network model, the weights are constantly adjusted to minimize the prediction error. A larger weight means that the input has a greater impact on the output, and a smaller weight means a smaller impact. The purpose of optimizing the weights is to make the BP neural network model fit the data better, thereby improving the prediction accuracy of the BP neural network model.

[0134] The threshold is a constant term added to the linear combination output of neurons in the BP neural network model, allowing the model to shift the output without relying on the input. The threshold can be regarded as the bias value of the activation function, which enables the model to better fit nonlinear relationships.

[0135] The purpose of obtaining the optimal weights and thresholds for fitness is to maximize the fitness function (also the performance or accuracy of the model) by continuously adjusting these parameters. By finding the optimal weights and thresholds, the bearing capacity of a single pile can be predicted more accurately.

[0136] In some optional implementations, the above step S3052 includes:

[0137] Step b1, population initialization: the original data of the input single pile diameter, single pile length, side friction resistance and single pile end resistance are converted into binary codes and randomly form an initial population.

[0138] Step b2, fitness optimization: the initial weights and thresholds of the data processing type single pile bearing capacity prediction model with gradually increasing fitness are obtained according to the individuals, and are transferred into the data processing type single pile bearing capacity prediction model with gradually increasing fitness for model training, and the weights and thresholds of the optimal fitness are found according to the absolute value of the error.

[0139] Step b3, iteration: After multiple iterations of fitness optimization, selection, crossover, and mutation, the iteration termination condition is finally met and the model iteration is completed.

[0140] Step S3053, based on the weight and threshold with the best suitability, a local optimization algorithm is used to calculate the optimal solution for the single pile bearing capacity.

[0141] Specifically, the input parameters are brought into the model, and the inversion data with the maximum fitness is output as the result, i.e., the optimal solution for the single pile bearing capacity.

[0142] It should be noted that when each group uses the optimization algorithm to repeatedly train and adjust the data processing type single pile bearing capacity prediction model with gradually increasing fitness in the simulated environment, the same conditions are repeated 10 times, and the results are averaged to ensure the stability and reliability of the results. In addition, since the numerical range and unit of the environmental parameters affecting carbon flux vary greatly, the model training and prediction effects will have a significant impact, so it is necessary to normalize the measured data obtained from the test.

[0143] The method for predicting the bearing capacity of a single offshore wind power pile provided in this embodiment adds a genetic algorithm to a BP neural network model, and evaluates the applicability of the single pile bearing capacity prediction model through errors in the verification process, and uses the applicability to optimize the symbol string population in the evolution process of the genetic algorithm to obtain a data processing type single pile bearing capacity prediction model with gradually increasing fitness, thereby achieving optimization of the single pile bearing capacity prediction model, improving the prediction accuracy of the single pile bearing capacity, and avoiding the problem of low prediction accuracy due to the lack of theoretical basis for empirical formulas or coefficients. The number of input layer nodes, hidden layer nodes and output layer nodes of the data processing type single pile bearing capacity prediction model with gradually increasing fitness is determined, which shortens the prediction time of the single pile bearing capacity. Based on the number of input layer nodes, hidden layer nodes and output layer nodes, a genetic algorithm is used to perform a global search on the data processing type single pile bearing capacity prediction model with gradually increasing fitness, and the weights and thresholds with the best fitness are obtained. Based on the weights and thresholds with the best fitness, a local optimization algorithm is used to calculate the optimal solution of the single pile bearing capacity. Based on the BP neural network pile foundation bearing capacity prediction, a GA genetic algorithm is used for optimization to obtain the optimal solution of the single pile bearing capacity with high accuracy, thereby improving the prediction accuracy.

[0144] As one or more specific application embodiments of the embodiments of the present invention, the prediction method of the bearing capacity of a single offshore wind power pile provided by the present invention is further described in detail in combination with the prediction of the bearing capacity of two test piles in the target area, as follows:

[0145] Step 1: Plane layout of test pile positions and test pile specification parameters:

[0146] According to the design requirements, there are two groups of static load test piles, and the positions of the two groups of test piles are arranged in the same way. Each group of test piles consists of 1 test pile, 4 anchor piles and 2 reference piles. The test piles and anchor piles use Φ2000mm open steel pipe piles, and the reference piles are Φ1200mm open steel pipe piles. The installation elevations of the fiber optic sensors are: ZK6-SZ is +4m to -73m, and ZK8-SZ is +4m to -75m. The test task requirements and the types, specifications and technical parameters of all test piles, anchor piles and reference piles in the test piles are shown in Tables 1 and 2 below.

[0147] Table 1 Test task requirements

[0148]

[0149] Table 2 Test pile parameters

[0150]

[0151] Step 2: Axial compressive static load test:

[0152] The ZK6-SZ pile is set as the test pile. During the test, the top of the ZK6-SZ pile is in a free and unconstrained state. The test loading process is loaded at 2400kN (kN is a unit of kilonewton) per level to 36000kN, meeting the maximum load termination loading condition. When the test is loaded from zero to 36000kN, the total settlement of the pile top is 50.79mm, and the load-settlement curve of the loading section remains basically linear. After graded unloading, the residual settlement of the pile top after unloading to zero is 7.42mm. According to the specification, the recommended value of the vertical compressive ultimate bearing capacity of a single ZK6-SZ pile is not less than 36000kN.

[0153] The ZK8-SZ pile was set as the test pile. During the test, the top of the ZK8-SZ pile was in a free and unconstrained state. When the test loading process was loaded at 2400kN per level to 33600kN, the test platform made a severe abnormal noise. Considering the safety of the test, it was decided to stop loading and unload in stages. When the test was loaded from zero to 33600kN, the total settlement of the pile top was 46.24mm, and the load-settlement curve of the loading section remained basically linear. The residual settlement of the pile top after unloading to zero was 7.59mm. According to the specification, the recommended value of the vertical compressive ultimate bearing capacity of a single ZK8-SZ pile is not less than 33600kN.

[0154] Step 3: Data sample set preprocessing:

[0155] First, the original CPTU test result data is checked for non-emptiness, and all empty values ​​are replaced with 0. Since this project is an offshore wind power project, the bearing capacity of the test piles often fails to reach the ultimate bearing value of the test piles, and due to the high test cost, it is temporarily impossible to conduct pile foundation bearing capacity tests on enough test piles. Based on the site characteristics of the project location and the empirical formula of the specification "Technical Specification for Building Pile Foundations" (JGJ94-2008), the characteristic values ​​of the data samples are pre-processed, and the corresponding ultimate bearing capacity is calculated as the output value of the data sample. In this method, f p By modifying the cone tip resistance q t The effective cone tip resistance q after subtracting u2 E Get, where q t =q c +(1-a n )u2,a n is the net area ratio of the penetrometer. b Within the depth range affected by the pile tip, q E The geometric mean of n is the number of strata that the probe passes through. The size effect of the pile is also taken into account. s In this method it is used to determine the soil layer type and the associated calculation coefficients.

[0156] Step 4: Single pile foundation bearing capacity prediction model based on BP neural network:

[0157] Through the process of step S202, the CPTU measured data of 50 holes and the parameter information of the two piles ZK6-SZ and ZK8-SZ are processed, and the 100 single pile bearing capacity sample data obtained by processing are trained, and the training test ratio is 4: 1. First, two models are established for the two piles ZK6-SZ and ZK8-SZ for training and testing, and the prediction models of the two piles ZK6-SZ and ZK8-SZ are obtained. Then, the data of the two piles are combined to establish a new site prediction model to predict the bearing capacity of the single pile in the site. The differences between the prediction models of the two piles ZK6-SZ and ZK8-SZ and the site prediction model results are compared, and then optimization is performed based on the GA genetic algorithm, and finally a more reasonable and accurate single pile bearing capacity prediction of the site can be obtained.

[0158] The cost of the bearing capacity test of offshore wind power piles is very high, and often the test piles fail to reach the ultimate bearing capacity value. This embodiment starts from the existing specifications and empirical formulas, combined with the stratum-related characteristics of the project area, based on BP neural network and GA optimization algorithm (i.e. GA-BP neural network), to predict the pile foundation bearing capacity of the two test piles in the field, so that the single pile bearing capacity in the field can be obtained by only conducting a pore pressure static penetration test, thereby reducing the construction cost of the project. CPTU has a large amount of data, very complex calculations, and many influencing factors. In addition, due to the lack of theoretical basis for empirical formulas / coefficients, the accuracy is low. This method uses BP neural network to analyze SCPTU (Seismic Piezocone Penetration Test, SCPTU) data to predict pile foundation bearing capacity, which not only avoids complex theoretical calculations, but also has high accuracy, so it has practical promotion significance. The offshore wind power single pile bearing capacity prediction method provided in this embodiment can combine the existing measured data and theoretical calculations, and realize the prediction and analysis of rock and soil properties, pile foundation bearing capacity and other data through data fitting and algorithm application, thereby reducing the workload of data monitoring while improving the accuracy of theoretical calculations, and can also maintain the high precision of nonlinear algorithms. It is very suitable for processing the relationship between the ultimate bearing capacity of a single pile and its influencing factors, and has high research value and development prospects.

[0159] In this embodiment, a prediction device for the bearing capacity of a single offshore wind power pile is also provided, which is used to implement the above-mentioned embodiments and preferred implementation modes, and will not be repeated hereafter. As used below, the term "module" can be a combination of software and / or hardware that implements a predetermined function. Although the device described in the following embodiments is preferably implemented in software, the implementation of hardware, or a combination of software and hardware, is also possible and conceivable.

[0160] This embodiment provides a device for predicting the bearing capacity of a single offshore wind power pile. Figure 6 As shown, including:

[0161] The measured data acquisition module 601 is used to obtain measured data of different strata where the single pile is located based on the pore pressure static penetration test.

[0162] The resistance calculation module 602 is used to determine the side friction resistance and unit pile end resistance of different strata where the single pile is located based on the measured data.

[0163] The model training module 603 is used to train the BP neural network model based on the preset single pile attribute data, side friction resistance and single pile end resistance to obtain a single pile bearing capacity prediction model.

[0164] The model optimization module 604 is used to optimize the single pile bearing capacity prediction model based on a genetic algorithm to obtain a data processing type single pile bearing capacity prediction model with gradually increasing fitness.

[0165] The model solving module 605 is used to solve the data processing type single pile bearing capacity prediction model with gradually increasing fitness to obtain the optimal solution of the single pile bearing capacity.

[0166] In some optional implementations, the measured data include cone tip resistance, side wall friction resistance and pore water pressure; the resistance calculation module 602 includes:

[0167] The preliminary modified cone tip resistance calculation unit is used to determine the preliminary modified cone tip resistance based on the cone tip resistance and the pore water pressure.

[0168] The effective cone tip resistance unit is used to determine the effective cone tip resistance based on the preliminary corrected cone tip resistance and the pore water pressure.

[0169] The side friction resistance and end resistance calculation unit is used to determine the side friction resistance based on the side wall friction resistance and determine the single pile end resistance based on the effective cone tip resistance.

[0170] In some optional implementations, the preset single pile attribute data includes a single pile diameter and a single pile length, and the model training module 603 includes:

[0171] The data set sample determination unit is used to take the single pile diameter, single pile length, side friction resistance and single pile end resistance as the data sample set.

[0172] A model training unit is used to train the BP neural network model based on the data sample set and using the gradient descent method to obtain a single pile bearing capacity prediction model. In some optional embodiments, the model training unit includes:

[0173] The sample division subunit is used to divide the data sample set into a training sample set and a verification sample set.

[0174] The training subunit is used to train the BP neural network model based on the training sample set and adopt the gradient descent method to obtain the training result.

[0175] The verification subunit is used to input the verification sample set into the BP neural network model for verification to obtain the verification result.

[0176] The judgment subunit is used to calculate the error between the training result and the verification result. When the error does not decrease or increase after a preset number of consecutive times, the training of the BP neural network model is stopped, and the BP neural network model obtained after the last training is used as the single pile bearing capacity prediction model.

[0177] In some optional implementations, the model optimization module 604 includes:

[0178] The evaluation unit is used to evaluate the applicability of the single pile bearing capacity prediction model based on the error.

[0179] The optimization unit is used to optimize the symbol string population in the evolution process of the genetic algorithm by using the fitness, so as to obtain a data processing type single pile bearing capacity prediction model with gradually increasing fitness.

[0180] In some optional implementations, the model solving module 605 includes:

[0181] The node number determination unit is used to determine the number of input layer nodes, the number of hidden layer nodes and the number of output layer nodes of the data processing type single pile bearing capacity prediction model with gradually increasing fitness.

[0182] The global search unit is used to perform a global search on the data processing type single pile bearing capacity prediction model with gradually increasing fitness based on the number of input layer nodes, the number of hidden layer nodes and the number of output layer nodes by using a genetic algorithm to obtain the weight and threshold with the best fitness.

[0183] The solution unit is used to calculate the optimal solution of the single pile bearing capacity by using a local optimization algorithm based on the weight and threshold of the optimal fitness.

[0184] The further functional description of each of the above modules and units is the same as that of the above corresponding embodiments and will not be repeated here.

[0185] The prediction device for the bearing capacity of a single offshore wind power pile in this embodiment is presented in the form of a functional unit, where the unit refers to an ASIC (Application Specific Integrated Circuit) circuit, a processor and memory that executes one or more software or fixed programs, and / or other devices that can provide the above functions.

[0186] The embodiment of the present invention also provides a computer device having the above Figure 6 The prediction device for the bearing capacity of a single offshore wind power pile is shown.

[0187] See also Figure 7 , Figure 7 is a schematic diagram of the structure of a computer device provided by an optional embodiment of the present invention, such as Figure 7As shown, the computer device includes: one or more processors 10, a memory 20, and interfaces for connecting various components, including high-speed interfaces and low-speed interfaces. Various components are connected to each other using different buses for communication, and can be installed on a common mainboard or installed in other ways as needed. The processor can process the instructions executed in the computer device, including instructions stored in or on the memory to display the graphical information of the GUI on an external input / output device (such as, a display device coupled to the interface). In some optional embodiments, if necessary, multiple processors and / or multiple buses can be used together with multiple memories and multiple memories. Similarly, multiple computer devices can be connected, and each device provides some necessary operations (for example, as a server array, a group of blade servers, or a multi-processor system). Figure 7 A processor 10 is taken as an example.

[0188] The processor 10 may be a central processing unit, a network processor or a combination thereof. The processor 10 may further include a hardware chip. The hardware chip may be a dedicated integrated circuit, a programmable logic device or a combination thereof. The programmable logic device may be a complex programmable logic device, a field programmable gate array, a general purpose array logic or any combination thereof.

[0189] The memory 20 stores instructions executable by at least one processor 10, so that the at least one processor 10 executes the method shown in the above embodiment.

[0190] The memory 20 may include a program storage area and a data storage area, wherein the program storage area may store an operating system, an application required for at least one function; the data storage area may store data created according to the use of the computer device, etc. In addition, the memory 20 may include a high-speed random access memory, and may also include a non-transient memory, such as at least one disk storage device, a flash memory device, or other non-transient solid-state storage device. In some optional embodiments, the memory 20 may optionally include a memory remotely arranged relative to the processor 10, and these remote memories may be connected to the computer device via a network. Examples of the above-mentioned network include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network, and combinations thereof.

[0191] The memory 20 may include a volatile memory, such as a random access memory; the memory may also include a non-volatile memory, such as a flash memory, a hard disk or a solid state drive; the memory 20 may also include a combination of the above types of memory.

[0192] The computer device also includes an input device 30 and an output device 40. The processor 10, the memory 20, the input device 30 and the output device 40 may be connected via a bus or other means. Figure 7 The example of connecting through bus is taken in the following.

[0193] The input device 30 can receive input digital or character information, and generate key signal input related to the user settings and function control of the computer device, such as a touch screen, a keypad, a mouse, a track pad, a touch pad, an indicator bar, one or more mouse buttons, a trackball, a joystick, etc. The output device 40 may include a display device, an auxiliary lighting device (e.g., an LED) and a tactile feedback device (e.g., a vibration motor), etc. The above-mentioned display device includes but is not limited to a liquid crystal display, a light emitting diode, a display and a plasma display. In some optional embodiments, the display device can be a touch screen.

[0194] The embodiment of the present invention also provides a computer-readable storage medium. The method according to the embodiment of the present invention can be implemented in hardware, firmware, or can be implemented as a computer code that can be recorded in a storage medium, or can be implemented as a computer code that is originally stored in a remote storage medium or a non-temporary machine-readable storage medium and will be stored in a local storage medium through a network download, so that the method described herein can be stored in such software processing on a storage medium using a general-purpose computer, a dedicated processor, or programmable or dedicated hardware. Among them, the storage medium can be a magnetic disk, an optical disk, a read-only storage memory, a random access memory, a flash memory, a hard disk or a solid-state hard disk, etc.; further, the storage medium can also include a combination of the above types of memories. It can be understood that a computer, a processor, a microprocessor controller, or programmable hardware includes a storage component that can store or receive software or computer code. When the software or computer code is accessed and executed by a computer, a processor, or hardware, the method shown in the above embodiment is implemented.

[0195] Although the embodiments of the present invention have been described in conjunction with the accompanying drawings, those skilled in the art may make various modifications and variations without departing from the spirit and scope of the present invention, and such modifications and variations are all within the scope defined by the appended claims.

Claims

1. A method for predicting the bearing capacity of a single offshore wind power pile, characterized in that: The method comprises: Obtain measured data of different strata where the single pile is located based on pore pressure static penetration test; Determine the side friction resistance and unit pile end resistance of different strata where the single pile is located based on the measured data; The BP neural network model is trained based on the preset single pile attribute data, side friction resistance and single pile end resistance to obtain a single pile bearing capacity prediction model; Optimizing the single pile bearing capacity prediction model based on a genetic algorithm to obtain a data processing type single pile bearing capacity prediction model with gradually increasing fitness; The data processing type single pile bearing capacity prediction model with gradually increasing fitness is solved to obtain the optimal solution for the single pile bearing capacity.

2. The method according to claim 1, characterized in that The measured data include cone tip resistance, side wall friction resistance and pore water pressure; Determining the side friction resistance and unit pile end resistance of different strata of a single pile based on the measured data includes: determining a preliminary corrected cone tip resistance based on the cone tip resistance and the pore water pressure; determining an effective cone tip resistance based on the preliminary corrected cone tip resistance and the pore water pressure; The side friction resistance is determined based on the side wall friction resistance, and the single pile end resistance is determined based on the effective cone tip resistance.

3. The method according to claim 1, characterized in that The preset single pile attribute data includes single pile diameter and single pile length; The BP neural network model is trained based on the preset single pile attribute data, side friction resistance and single pile end resistance, and the single pile bearing capacity prediction model is obtained, including: The single pile diameter, single pile length, side friction resistance and single pile end resistance are used as a data sample set; The BP neural network model is trained based on the data sample set and the gradient descent method is adopted to obtain a single pile bearing capacity prediction model.

4. The method according to claim 3, characterized in that Based on the data sample set and using the gradient descent method to train the BP neural network model, a single pile bearing capacity prediction model is obtained, including: Dividing the data sample set into a training sample set and a verification sample set; Based on the training sample set, the BP neural network model is trained by using a gradient descent method to obtain a training result; Inputting the verification sample set into the BP neural network model for verification to obtain a verification result; The error between the training result and the verification result is calculated. When the error does not decrease or increase after a preset number of consecutive times, the training of the BP neural network model is stopped, and the BP neural network model obtained after the last training is used as a single pile bearing capacity prediction model.

5. The method according to claim 3 or 4, characterized in that: The single pile bearing capacity prediction model is expressed by the following formula: Q ult =Q s +Q b =Σf pi TO si +q b TO b Among them, Q ult is the ultimate bearing capacity of a single pile, Q s is the lateral bearing capacity of a single pile, Q b is the pile end bearing capacity, f pi is the unit lateral friction of the ith layer of the stratum where the single pile is located, A si is the lateral area of ​​a single pile in the i-th stratum, q b is the single pile end resistance, A b is the pile tip area, which is calculated from the single pile diameter and single pile length.

6. The method according to claim 4, characterized in that The single pile bearing capacity prediction model is optimized based on a genetic algorithm, and a data processing type single pile bearing capacity prediction model with gradually increasing fitness is obtained, including: Evaluating the applicability of the single pile bearing capacity prediction model based on the error; The fitness is used to optimize the symbol string population in the evolution process of the genetic algorithm, so as to obtain a data processing type single pile bearing capacity prediction model with gradually increasing fitness.

7. The method according to claim 1, characterized in that The data processing type single pile bearing capacity prediction model with gradually increasing fitness is solved to obtain the optimal solution of the single pile bearing capacity, including: Determining the number of input layer nodes, the number of hidden layer nodes, and the number of output layer nodes of the data processing type single pile bearing capacity prediction model with gradually increasing fitness; Based on the number of input layer nodes, the number of hidden layer nodes and the number of output layer nodes, a genetic algorithm is used to perform a global search on a data processing type single pile bearing capacity prediction model with gradually increasing fitness, so as to obtain a weight and a threshold with the best fitness; Based on the weight and threshold of the optimal fitness, a local optimization algorithm is used to calculate the optimal solution for the single pile bearing capacity.

8. A device for predicting the bearing capacity of a single offshore wind power pile, characterized in that: The device comprises: A measured data acquisition module is used to obtain measured data of different strata where the single pile is located based on the pore pressure static penetration test; A resistance calculation module, used to determine the side friction resistance and unit pile end resistance of different strata where the single pile is located based on the measured data; A model training module is used to train the BP neural network model based on preset single pile attribute data, side friction resistance and single pile end resistance to obtain a single pile bearing capacity prediction model; A model optimization module, used for optimizing the single pile bearing capacity prediction model based on a genetic algorithm to obtain a data processing type single pile bearing capacity prediction model with gradually increasing fitness; The model solving module is used to solve the data processing type single pile bearing capacity prediction model with gradually increasing fitness to obtain the optimal solution of the single pile bearing capacity.

9. A computer device, characterized in that: include: A memory and a processor, wherein the memory and the processor are communicatively connected to each other, the memory stores computer instructions, and the processor executes the method for predicting the bearing capacity of a single offshore wind power pile according to any one of claims 1 to 7 by executing the computer instructions.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores computer instructions, and the computer instructions are used to enable a computer to execute the method for predicting the bearing capacity of an offshore wind power single pile according to any one of claims 1 to 7.

Images