CN-121997690-A - Evaluation method and system for composite impact multi-tooth rock breaking test

CN121997690ACN 121997690 ACN121997690 ACN 121997690ACN-121997690-A

Abstract

The application provides an evaluation method and system for a composite impact multi-tooth rock breaking test, wherein the evaluation method comprises the following steps: determining main control factors which cause the rock damage state to be different as variable parameters through a composite impact test, acquiring test results of rock damage and corresponding variable parameters as sample data, changing the variable parameters to perform the composite impact test, acquiring the sample data, establishing a data set, establishing a neural network structure and an activation function, training a neural network to form a neural network model, inputting specific variable parameters to be tested into the neural network model, outputting corresponding test prediction results, and completing test evaluation. According to the evaluation method and system, the main control factors influencing the test are determined, the comprehensiveness of the input parameters and the data set is ensured, and the characteristics are automatically learned from multiple data through the neural network model, so that the complex nonlinear relation data is processed, and the efficient and accurate prediction evaluation is maintained in the face of complex geological environment.

Inventors

YANG TANGBIN
LUO MIN
CHEN PENGYU
LIU BIN
CHEN RENJIN
XING YUZHONG
GONG XINGLIN
ZHANG WENQI

Assignees

中国石油天然气股份有限公司
中国石油国际勘探开发有限公司
中石油阿姆河天然气勘探开发(北京)有限公司
中国石油集团科学技术研究院有限公司

Dates

Publication Date: 20260508
Application Date: 20241106

Claims (10)

1. The evaluation method of the composite impact multi-tooth rock breaking test is characterized by comprising the following steps of: Selecting and determining a main control factor which causes the rock damage state to be different as a variable parameter through a composite impact test; Based on the composite impact test, obtaining a rock damage test result and corresponding variable parameters as sample data, changing the variable parameters to perform the next group of composite impact test until all sample data under different parameters are obtained, and establishing a data set; establishing a neural network structure and an activation function according to the data set, and training the neural network to form a neural network model; Inputting the specific variable parameters to be tested into the neural network model, outputting the corresponding test prediction results, and completing test evaluation.
2. The method of claim 1, wherein the variable parameters include a sleeve impact speed, a sleeve rotational speed, a tooth shape, a tooth rake angle, a tooth diameter, a rock type, a formation poisson's ratio, an elastic modulus, and a confining pressure.
3. The method for evaluating the composite impact multi-tooth rock breaking test according to claim 2, wherein the test result comprises penetration depth, injury area and breaking volume.
4. The method for evaluating the composite impact multi-tooth rock breaking test according to claim 1, wherein in the process of establishing the neural network structure according to the data set, the data set is distributed into a training set, a verification set and a test set, wherein the training set is a test set, the verification set=70-80%: 10-15%; the training set is used for training a neural network model; the verification set is used for adjusting and optimizing model parameters; the test set is then used to verify the generalization ability of the model.
5. The method of evaluating a composite impact multi-tooth rock breaking test according to claim 4, wherein in establishing the neural network structure from the data set; the neural network structure comprises an input layer, an output layer and a hidden layer, wherein the number of neurons of the input layer and the output layer is consistent with the variable parameters and the types of the test results, and the number of the hidden layers is at least one.
6. The method of claim 5, wherein the activation function is a non-saturated activation function.
7. The method for evaluating a composite impact multi-tooth rock breaking test according to claim 6, wherein the neural network model comprises a first layer of integrated learning framework and a second layer of integrated learning framework; The first layer of integrated learning framework comprises a BP neural network, an RF model and a XGBoost model, wherein the input parameters of the first layer of integrated learning framework are variable parameters, and the output parameters are the calculation result of the BP neural network, the calculation result of the RF model and the calculation result of the XGBoost model; The second layer of integrated learning framework is used for fusing the calculation result of the BP neural network, the calculation result of the RF model and the calculation result of the XGBoost model and outputting a test prediction result.
8. An evaluation system of a composite impact multi-tooth rock breaking test device, which is characterized by comprising: the parameter selection module is used for selecting and determining main control factors which cause the rock damage state to be different as variable parameters through a composite impact test; the data set establishing module is used for acquiring a rock damage test result and corresponding variable parameters as sample data based on the composite impact test, changing the variable parameters to perform the next group of composite impact test until all sample data under different parameters are acquired, and establishing a data set; The model building module is used for building a neural network structure and an activation function according to the data set and training the neural network, wherein the neural network is used for learning and predicting the damage form of the rock after the impact test from the database; And the model operation module inputs the specific variable parameters to be tested to the neural network model, outputs the corresponding test prediction result and completes test evaluation.
9. The evaluation system of the composite impact multi-tooth rock breaking test device according to claim 8, wherein the variable parameters comprise sleeve impact speed, sleeve rotation speed, drilling tooth shape, drilling tooth inclination angle, drilling tooth diameter, rock type, stratum poisson ratio, elastic modulus and confining pressure; The test results comprise penetration depth, injury area and crushing volume.
10. The evaluation system of a composite impact multi-tooth rock breaking test device according to claim 9, wherein the neural network model comprises a first layer of integrated learning framework and a second layer of integrated learning framework; The first layer of integrated learning framework comprises a BP neural network, an RF model and a XGBoost model, wherein the input parameters of the first layer of integrated learning framework are variable parameters, and the output parameters are the calculation result of the BP neural network, the calculation result of the RF model and the calculation result of the XGBoost model; The second layer of integrated learning framework is used for fusing the calculation result of the BP neural network, the calculation result of the RF model and the calculation result of the XGBoost model and outputting a test prediction result.

Description

Evaluation method and system for composite impact multi-tooth rock breaking test Technical Field The invention belongs to the technical field of rock breaking, and particularly relates to an evaluation method and an evaluation system for a composite impact multi-tooth rock breaking test. Background Along with the continuous development of oil and gas exploration and mining technologies, how to effectively improve drilling and mining efficiency and reduce rock damage is a key problem in the current engineering technology, and correspondingly, a rock breaking test is used as an important link in oil and gas exploration and development, and necessary rock mechanical parameters are provided for key engineering links such as drilling, fracturing, well completion and the like, so that the optimization design of yield-increasing measures is facilitated, and the well shaft instability and engineering risks are reduced. However, the conventional rock breaking test depends on a large amount of experimental data and empirical formulas, which is time-consuming and high in cost, and the corresponding evaluation of the rock breaking effect is often based on an indoor test or a numerical simulation method, so that the influence factors of the rock breaking damage are more, and the influence factors have no linear relation, so that the method is difficult to directly apply on an engineering site, and the accuracy of the rock breaking test prediction is affected, and the evaluation conclusion of the rock is affected. In view of this, overcoming the defects in the prior art is a problem to be solved in the art. Disclosure of Invention Aiming at the problems, the invention provides an evaluation method of a composite impact multi-tooth rock breaking test, which comprises the following steps: Selecting and determining a main control factor which causes the rock damage state to be different as a variable parameter through a composite impact test; Based on the composite impact test, obtaining a test result of rock damage and corresponding variable parameters as sample data, changing the variable parameters to perform the next group of composite impact test until all sample data under different parameters are obtained, and establishing a data set; establishing a neural network structure and an activation function according to the data set, and training the neural network to form a neural network model; Inputting the specific variable parameters to be tested into the neural network model, outputting the corresponding test prediction results, and completing test evaluation. Still further, the variable parameters include sleeve impact speed, sleeve rotational speed, tooth shape, tooth rake, tooth diameter, rock type, formation poisson's ratio, modulus of elasticity, and confining pressure. Further, the test results include penetration depth, lesion area, and crush volume. Further, in the process of establishing the neural network structure according to the data set, the data set is distributed into a training set, a verification set and a test set, wherein the training set comprises the verification set=70-80%: 10-15%; The training set is used for training the neural network model; The verification set is used for adjusting and optimizing model parameters; the test set is then used to verify the generalization ability of the model. Further, in establishing the neural network structure from the data set; the neural network structure comprises an input layer, an output layer and a hidden layer, wherein the number of neurons of the input layer and the output layer is consistent with the variable parameters and the types of the test results, and the number of the hidden layers is at least one. Further, the activation function is an unsaturated activation function. Further, the neural network model comprises a first layer of integrated learning framework and a second layer of integrated learning framework; the first layer of integrated learning framework comprises a BP neural network, an RF model and a XGBoost model, wherein the input parameters of the first layer of integrated learning framework are variable parameters, and the output parameters are the calculation result of the BP neural network, the calculation result of the RF model and the calculation result of the XGBoost model; The second layer of integrated learning framework is used for fusing the calculation result of the BP neural network, the calculation result of the RF model and the calculation result of the XGBoost model and outputting a test prediction result. The invention also provides an evaluation system of the composite impact multi-tooth rock breaking test, which comprises: the parameter selection module is used for selecting and determining main control factors which cause the rock damage state to be different as variable parameters through a composite impact test; The data set establishing module is used for acquiring a test result of rock damage and corresponding