A numerical analysis method of formation vibration model based on big data analysis

Abstract

The present invention discloses a numerical analysis method for a stratum vibrothelastic model based on big data analysis, which relates to the field of vibrator data analysis. The method includes monitoring the construction progress of the vibrator, simulating historical data, selecting the numerical range of current data during the retention phase, constructing a three-dimensional visualization model of voltage, frequency, and predicted core loss, and reflecting the stability parameters of the vibrator through the predicted core loss data; analyzing whether the current, voltage, and frequency data under the frequency conversion strategy are within the core loss numerical range. The core loss numerical range is divided based on the training data of the historical execution of the pull-up and then insert strategy. If it is within the core loss numerical range, the pull-up and then insert strategy is executed. When the vibrator construction is obstructed, the stratum vibrothelastic model output is used to adjust the construction parameters of the vibrator and connect to the vibrator control system. The present invention uses data models to analyze and quantify the stability parameters of the vibrator, rationally implement the operation strategy, prevent machine damage, and ensure work efficiency.

Description

Technical Field

[0001] The present invention relates to the field of vibrator data analysis, and in particular to a numerical analysis method of a stratum vibratory model based on big data analysis. Background Art

[0002] When the vibrator is running at a high current when encountering a hard or dense layer, it is easy to cause an accident of burning the motor due to current overload;

[0003] The existing technology adopts current limiting and frequency reduction to reduce the operating frequency and current of the vibratory impactor to ensure the stability of the equipment, but it affects the working efficiency of the vibratory impactor.

[0004] Therefore, the present invention considers a numerical analysis method of a formation vibration model based on big data analysis. Summary of the Invention

[0005] The present invention provides a numerical analysis method of a formation vibration model based on big data analysis to solve the problems of the prior art, comprising the following steps:

[0006] Obtain the current, frequency, voltage and tension data of the vibrator;

[0007] Monitor the vibratory impactor's progress and connect real-time current, frequency, voltage, and tension data to the formation vibratory impact model for numerical analysis.

[0008] The formation vibrothelastic model uses historical current, frequency, voltage, and tension data to train two response strategies: a vibrator lift-and-insert strategy and a frequency conversion strategy.

[0009] By simulating historical data, when the fluctuation range of the monitored tension data does not exceed the preset threshold value that affects the verticality, the numerical range of the current data in the vibration retention stage is selected, the frequency and voltage data to be trained are read in, and a three-dimensional visualization model of voltage, frequency and predicted core loss is constructed. The stability parameters of the vibrator are reflected by the predicted core loss data; the Steinmetz empirical formula including frequency and flux density data is loaded to estimate the core loss, and the flux density data is calculated through the physical model of the vibrator. The trained formation vibration model is connected to the real-time construction data for numerical analysis, and the current, voltage and frequency data of the variable frequency strategy are calculated. It is analyzed whether the current, voltage and frequency data under the variable frequency strategy are within the core loss value range. The core loss value range is divided according to the training data of the historical execution of the lift-up and then insertion strategy. If it is within the core loss value range, the lift-up and then insertion strategy is executed. If not, the variable frequency strategy is executed.

[0010] When it is detected that the vibratory impactor construction is obstructed, the output of the stratum vibratory impact model is used to adjust the construction parameters of the vibratory impactor and is connected to the vibratory impactor control system.

[0011] Furthermore, when training the formation vibration model, current data is introduced to correct the empirical formula, and the current data is added to the empirical formula to correct the accuracy of the empirical formula;

[0012] By correlating the changes in current values ​​in the non-retention stage with the changes in corresponding frequency and voltage values, the data is loaded into the formation vibration model to calculate the predicted core loss value. The core loss value range divided by strategy training is inserted into the original historical execution after the improvement to filter the predicted core loss value calculated in the non-retention stage.

[0013] Furthermore, the current values ​​of multiple retention stages are grouped, and the voltage and frequency values ​​corresponding to the current fluctuations in different retention stages are input into the formation vibration model to train an expanded range of core loss values.

[0014] Furthermore, during the vibration retention phase, the monitored vibrator current data is divided into a starting current period between 225A and 300A as the vibration retention phase.

[0015] Furthermore, when the vibrator executes the lift-and-insert strategy, during the vibration retention stage, the hole-making depth is monitored to ensure that it meets the preset vibration retention depth. After the current is controlled to be between 225A-300A for a preset vibration retention period, the lift-and-vibration retention operation is performed.

[0016] Furthermore, when monitoring historical data, after the starting current period of the vibration retention stage, the current of the vibrator drops below 225A after a preset vibration retention period, then the current period below 225A is divided into the same vibration retention stage until the current exceeds 225A and the division is ended.

[0017] Furthermore, after obtaining the current, frequency, voltage and tension data of the vibrator, the historical current data is classified, and the relevant data of the vibrator's lift-up and then insertion strategy in the historical data are grouped into the corresponding current data. The vibrators with different specifications are combined with the physical model to generate a core loss group. By outputting the data content of the core loss group that affects the stability of the vibrator, the critical core loss value of the core loss group under different physical models is calculated and analyzed. The critical core loss value is obtained by calculating the current data corresponding to the implementation of the lift-up and then insertion strategy in the physical model of each group of vibrators with different specifications.

[0018] Furthermore, in the execution of the frequency conversion strategy, historical frequency conversion data is monitored, and the DNN output function is loaded to fit multiple sets of frequency change data. The dynamic model of the vibrator is constructed based on the physical model of the corresponding frequency conversion data. The parameter changes of the vibrator in the implementation of historical frequency conversion operations are measured, and the frequency conversion data parameters in the dynamic model are statistically analyzed. The frequency data is synthesized as a whole with the variance as the parameter for neural network output, and an objective function is set. The objective function is used to converge multiple sets of frequency conversion data from low frequency to high frequency in turn, and the step size of the objective function is adjusted. Multiple global optimal solutions in a small neighborhood of the historical frequency conversion data that is initialized and loaded are output. The discrete multiple global optimal solutions are fitted to the original input frequency conversion data, and the global optimal solution is interpolated to supplement the data volume and recorded in the operating parameters. The frequency conversion is executed according to the operating parameters.

[0019] Furthermore, it also includes a verification process of the global optimal solution. For the DNN output function during the training process, the global optimal solution realizes the first-order Taylor expansion near the historical frequency-varying data loaded at the initialization at any time.

[0020] Furthermore, the predicted core loss is used to determine the loss caused by the alternating magnetic field caused by the current fluctuation at the boundary point between the retention strategy and the lift-and-insert strategy.

[0021] The present invention provides a numerical analysis method for a stratum vibration model based on big data analysis. The method uses a data model to analyze and quantify the stability parameters of the vibrator, conducts statistical analysis on historical data, and rationally implements operation strategies to prevent machine damage and ensure work efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] The drawings described herein are used to provide a further understanding of the embodiments of the present invention, constitute a part of the present invention, and do not constitute a limitation of the embodiments of the present invention. In the drawings:

[0023] Figure 1 A flow chart of a numerical analysis method of a formation vibration impact model based on big data analysis is provided as an exemplary embodiment of the present invention. DETAILED DESCRIPTION

[0024] Exemplary embodiments will be described in detail herein, examples of which are illustrated in the accompanying drawings. In the following description, when referring to the drawings, like numbers in different figures represent like or similar elements unless otherwise indicated. The embodiments described in the following exemplary embodiments are not intended to represent all possible embodiments consistent with the present invention. Rather, they are merely examples of apparatus and methods consistent with certain aspects of the present invention, as detailed in the appended claims.

[0025] The common Steinmetz equation often uses a fixed-frequency design to analyze the effect of temperature on core loss. In the industrial application of vibratory impactors, the hole is first drawn and then the hole is made in the engineering stage. In order to assist in determining whether the vibratory impactor should maintain vibration or be pulled and then reinserted to make the hole, the vibratory impactor operator performs manual operation based on machine feedback data and operating experience. The analysis of the geological and geographical conditions and the internal structure of the vibratory impactor head is unclear. The present invention provides a numerical analysis method of a stratum vibratory impact model based on big data analysis, which aims to solve the above technical problems of the prior art.

[0026] The present invention contemplates access to historical data for deep learning, and quantitatively analyzes the operating limits of vibrators of different specifications under different circumstances, that is, when to use retention vibration and when to use lifting and then insertion to create holes, because the two have different losses on the machine, and the current and frequency data are recorded in the internal chip during the machine operation. The following specific embodiments describe in detail the technical solution of the present invention and how the technical solution of the present invention solves the above-mentioned technical problems. The following specific embodiments can be combined with each other, and the same or similar concepts or processes may not be repeated in some embodiments. The embodiments of the present invention will be described below in conjunction with the accompanying drawings.

[0027] The present invention adopts the following solution: Figure 1 As shown:

[0028] S1. Obtain the current, frequency, voltage and tension data of the vibrator;

[0029] S2, monitor the construction progress of the vibratory impactor and connect the real-time current, frequency, voltage and tension data to the formation vibratory impact model for numerical analysis;

[0030] The formation vibrothelastic model uses historical current, frequency, voltage, and tension data to train two response strategies: a vibrator lift-and-insert strategy and a frequency conversion strategy.

[0031] S3. By simulating historical data, when the fluctuation range of the monitored tension data does not exceed the preset threshold value affecting the verticality, the value range of the current data in the vibration retention stage is selected;

[0032] S4. Read the frequency and voltage data to be trained, build a three-dimensional visualization model of voltage, frequency and predicted core loss, and reflect the stability parameters of the vibrator through the predicted core loss data;

[0033] A 3D model is constructed for the selected parameters and displayed inside the display for visual feedback to the vibrator operator.

[0034] S5. Load the Steinmetz empirical formula including frequency and magnetic flux density data for estimating core loss. The magnetic flux density data is calculated using the physical model of the vibrator. The trained formation vibratory model is connected to the real-time construction data for numerical analysis. The current, voltage, and frequency data for executing the variable frequency strategy are calculated. The current, voltage, and frequency data under the variable frequency strategy are analyzed to determine whether they are within the core loss value range. The core loss value range is divided based on the training data of the historical execution of the pull-up and then insert strategy. If the core loss value range is within the core loss value range, the pull-up and then insert strategy is executed. If not, the variable frequency strategy is executed.

[0035] The formation vibration model is a deep learning model that accesses current and frequency data to the physical model of the vibrator to obtain core loss. The specific training process is as follows: after obtaining the current, frequency, voltage and tension data of the vibrator, for the historical current data, the current data is classified, and the relevant data of the vibrator's lifting and insertion strategy in the historical data are grouped into the corresponding current data. The vibrators with different specifications are matched with the physical model to generate a core loss group. By outputting the data content of the core loss group that affects the stability of the vibrator, the critical core loss value of the core loss group under different physical models is calculated and analyzed. The critical core loss value is obtained by calculating the current data corresponding to the lifting and insertion strategy in the physical model of each group of vibrators with different specifications; the corresponding frequency and voltage are associated with the change of the current value in the non-retention stage. The numerical changes of the data are loaded into the formation vibration model to calculate the predicted core loss value, and the predicted core loss value calculated in the non-retention vibration stage is screened out in the core loss value range divided by the original historical execution of the lift-up and then insertion strategy training; the current values ​​of multiple retention vibration stages are grouped, and according to the voltage and frequency values ​​corresponding to the current fluctuations in different retention vibration stages, they are input into the formation vibration model for training to expand the core loss value range; wherein, for the retention vibration stage, the monitoring vibrator current data between 225A-300A is divided into the starting current period of the retention vibration stage, and when the vibrator executes the lift-up and then insertion strategy, in the retention vibration stage, the monitoring hole-making depth meets the preset retention vibration depth, and the control current is continuously between 225A-300A for a preset retention period before the lifting and retention vibration operation is performed. Among them, when monitoring historical data, after the starting current period of the vibration retention stage, if the current of the vibrator drops below 225A after a preset vibration retention period, the current period below 225A will be divided into the same vibration retention stage until the current exceeds 225A and the division is ended.

[0036] To execute the frequency conversion strategy, historical frequency conversion data is monitored, a DNN output function is loaded to fit multiple sets of frequency change data, a dynamic model of the vibrator is constructed based on the physical model of the corresponding frequency conversion data, parameter changes of the vibrator during the historical frequency conversion operation are measured, statistical regularity analysis is performed on the frequency conversion data parameters in the dynamic model, the frequency data is synthesized as a whole using the variance as a parameter for neural network output, an objective function is set to converge multiple sets of frequency conversion data from low frequency to high frequency, the step size of the objective function is adjusted, and multiple global optimal solutions are output within a small neighborhood of the initially loaded historical frequency conversion data. The multiple discrete global optimal solutions are fitted to the original input frequency conversion data, the global optimal solution is interpolated to supplement the data volume, and the data volume is recorded in the operating parameters, and the frequency conversion is executed according to the operating parameters. The global optimal solution verification process is also included. During the training process of the DNN output function, the global optimal solution is realized as a first-order Taylor expansion near the initially loaded historical frequency conversion data at any time.

[0037] When training the formation vibration model, current data is introduced to correct the empirical formula, and the current data is added to the empirical formula to correct the accuracy of the empirical formula;

[0038] S6. When it is detected that the vibratory impactor construction is obstructed, the stratum vibratory impactor model output is used to adjust the construction parameters of the vibratory impactor and is connected to the vibratory impactor control system.

[0039] At this time, for the obstructed situation, the vibration impact model is connected to analyze the construction data output results at this time in the core loss range, and different strategies are implemented according to different core loss ranges.

[0040] In the several embodiments provided herein, it should be understood that the disclosed systems and methods can be implemented in other ways. For example, the system embodiments described above are merely illustrative. For example, the module division is merely a logical functional division. In actual implementation, other division methods may be used, such as combining or integrating multiple modules or components into another system, or omitting or not implementing certain features.

[0041] The modules described as separate components may or may not be physically separate, and the components shown as modules may or may not be physical modules, that is, they may be located in one place or distributed across multiple network modules. Some or all of the modules may be selected to achieve the purpose of the present embodiment according to actual needs.

[0042] In addition, the functional modules in various embodiments of the present invention may be integrated into a single processing module, each module may exist physically separately, or two or more modules may be integrated into a single module. The aforementioned integrated modules may be implemented in the form of hardware or hardware plus software functional modules.

[0043] Those skilled in the art will appreciate that the embodiments of the present invention may be provided as methods or systems. Thus, the present invention may take the form of an entirely hardware embodiment, an entirely software embodiment, or an embodiment combining software and hardware.

[0044] It should also be noted that the terms "comprises," "includes," or any other variations thereof are intended to encompass non-exclusive inclusion, such that a process, method, commodity, or apparatus that includes a series of elements includes not only those elements but also other elements not explicitly listed, or includes elements inherent to such process, method, commodity, or apparatus. In the absence of further limitations, an element defined by the phrase "comprises a ..." does not exclude the presence of other identical elements in the process, method, commodity, or apparatus that includes the element.

[0045] The above are merely embodiments of the present invention and are not intended to limit the present invention. It will be apparent to those skilled in the art that various modifications and variations of the present invention are possible. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention are intended to be included within the scope of the claims of the present invention.

[0046] Other embodiments of the present invention will readily occur to those skilled in the art upon consideration of the specification and practice of the invention disclosed herein. The present invention is intended to cover any variations, uses, or adaptations of the present invention that follow from the general principles of the invention and include common knowledge or customary techniques in the art not disclosed herein. The description and examples are to be considered as exemplary only, with the true scope and spirit of the invention being indicated by the following claims.

[0047] It should be understood that the present invention is not limited to the exact construction described above and shown in the drawings, and that various modifications and changes may be made without departing from the scope thereof, which is limited only by the appended claims.

Claims

1. A numerical analysis method for formation vibration model based on big data analysis, characterized in that: The steps include: Obtain the current, frequency, voltage and tension data of the vibrator; Monitor the vibratory impactor's progress and connect real-time current, frequency, voltage, and tension data to the formation vibratory impact model for numerical analysis. The formation vibrothelastic model uses historical current, frequency, voltage, and tension data to train two response strategies: a vibrator lift-and-insert strategy and a frequency conversion strategy. By simulating historical data, when the fluctuation range of the monitored tension data does not exceed the preset threshold value that affects the verticality, the numerical range of the current data in the vibration retention stage is selected, the frequency and voltage data to be trained are read in, and a three-dimensional visualization model of voltage, frequency and predicted core loss is constructed. The stability parameters of the vibrator are reflected by the predicted core loss data; the Steinmetz empirical formula including frequency and flux density data is loaded to estimate the core loss, and the flux density data is calculated through the physical model of the vibrator. The trained formation vibration model is connected to the real-time construction data for numerical analysis, and the current, voltage and frequency data of the variable frequency strategy are calculated. It is analyzed whether the current, voltage and frequency data under the variable frequency strategy are within the core loss value range. The core loss value range is divided according to the training data of the historical execution of the lift-up and then insertion strategy. If it is within the core loss value range, the lift-up and then insertion strategy is executed. If not, the variable frequency strategy is executed. When it is detected that the vibratory impactor construction is obstructed, the output of the stratum vibratory impact model is used to adjust the construction parameters of the vibratory impactor and is connected to the vibratory impactor control system.

2. The method for numerical analysis of formation vibration model based on big data analysis according to claim 1, characterized in that: When training the formation vibration model, current data is introduced to correct the empirical formula, and the current data is added to the empirical formula to correct the accuracy of the empirical formula; By correlating the changes in current values ​​in the non-retention stage with the changes in corresponding frequency and voltage values, the data is loaded into the formation vibration model to calculate the predicted core loss value. The core loss value range divided by strategy training is inserted into the original historical execution after the improvement to filter the predicted core loss value calculated in the non-retention stage.

3. The method for numerical analysis of formation vibration model based on big data analysis according to claim 2, characterized in that: The current values ​​of multiple retention stages are grouped, and the voltage and frequency values ​​corresponding to the current fluctuations in different retention stages are input into the formation vibration model to train and expand the range of core loss values.

4. The method for numerical analysis of formation vibration model based on big data analysis according to claim 3 is characterized in that: During the vibration retention phase, the monitored vibrator current data is divided into a starting current period between 225A and 300A, which is the starting current period of the vibration retention phase.

5. The method for numerical analysis of formation vibration model based on big data analysis according to claim 4 is characterized in that: When the vibrator executes the lift-and-insert strategy, during the vibration retention stage, the hole-making depth is monitored to ensure that it meets the preset vibration retention depth. After the current is controlled between 225A-300A for a preset vibration retention period, the lift-and-vibration retention operation is performed.

6. The method for numerical analysis of formation vibration model based on big data analysis according to claim 5, characterized in that: For example, when monitoring historical data, after the starting current period of the vibration retention stage, if the current of the vibrator drops below 225A after a preset vibration retention period, the current periods below 225A are divided into the same vibration retention stage until the current exceeds 225A and the division is terminated.

7. The method for numerical analysis of formation vibration model based on big data analysis according to claim 6, characterized in that: After obtaining the current, frequency, voltage and tension data of the vibrator, the historical current data is classified, and the relevant data of the vibrator's lift-up and then insertion strategy in the historical data are grouped into the corresponding current data. The vibrators with different specifications are matched with the physical model to generate a core loss group. By outputting the data content of the core loss group that affects the stability of the vibrator, the critical core loss values ​​of the core loss groups under different physical models are calculated and analyzed. The critical core loss values ​​are obtained by calculating the current data corresponding to the lift-up and then insertion strategy in the physical model of each group of vibrators with different specifications.

8. The method for numerical analysis of formation vibration model based on big data analysis according to claim 7, characterized in that: In the execution of the frequency conversion strategy, the historical frequency conversion data is monitored, and the DNN output function is loaded to fit multiple sets of frequency change data. The dynamic model of the vibrator is constructed based on the physical model of the corresponding frequency conversion data. The parameter changes of the vibrator in the implementation of the historical frequency conversion operation are measured, and the frequency conversion data parameters in the dynamic model are statistically analyzed. The frequency data is synthesized as a whole with the variance as the parameter for neural network output, and the objective function is set. The objective function is used to converge multiple sets of frequency conversion data from low frequency to high frequency in sequence, and the step size of the objective function is adjusted. Multiple global optimal solutions in a small neighborhood of the historical frequency conversion data loaded for initialization are output, and multiple discrete global optimal solutions are fitted to the original input frequency conversion data. The global optimal solution is interpolated to supplement the data volume and recorded in the operating parameters, and the frequency conversion is executed according to the operating parameters.

9. The method for numerical analysis of formation vibration model based on big data analysis according to claim 8, characterized in that: It also includes the verification process of the global optimal solution. For the DNN output function during training, the global optimal solution realizes the first-order Taylor expansion near the historical frequency-varying data loaded at the initialization at any time.

10. The method for numerical analysis of formation vibration model based on big data analysis according to claim 1, characterized in that: The predicted core loss is used to determine the loss generated by the alternating magnetic field caused by the current fluctuation at the boundary point between the retention strategy and the lift-and-insert strategy.

Images