Energy exploration method and system based on distributed optical fiber sensing, device, and medium

Abstract

An energy exploration method, system and apparatus based on distributed optical fiber sensing, a device, and a medium. The method comprises the following steps: receiving a preliminary analysis result sent by a signal processing module, wherein the preliminary analysis result is obtained by the signal processing module performing signal conversion and signal analysis on physical quantity data related to energy in an energy exploration area; and inputting the preliminary analysis result into an optimized energy exploration prediction model for prediction, so as to obtain an energy exploration prediction result. By collecting and processing distributed optical fiber data, and using an energy exploration prediction model to perform energy exploration prediction, variation characteristics of energy activities at different time and spatial scales can be comprehensively captured, and energy distribution and reserves can be predicted more accurately, thereby improving the overall efficiency of energy exploration, and reducing exploration errors.

Description

Energy exploration methods, systems, equipment, and media based on distributed optical fiber sensing Technical Field

[0001] This invention relates to the field of energy exploration technology, specifically to an energy exploration method, system, device, equipment, and medium based on distributed optical fiber sensing. Background Technology

[0002] Distributed fiber optic sensing technology utilizes backscattered light (Rayleigh scattering, Brillouin scattering, and Raman scattering) and forward-propagating light in optical fibers for sensing and measurement, enabling continuous distributed measurement of external parameters along the fiber optic line. In energy exploration, this technology can monitor key parameters such as formation temperature and pressure in real time, providing strong support for accurately understanding the distribution and dynamic changes of underground energy resources. Currently, this technology has achieved significant results in oil and gas exploration and coal mine geological monitoring. For example, in oil and gas exploration, it can monitor formation pressure and temperature in real time, providing crucial information for the location and exploitation of oil and gas resources; in coal mine geological monitoring, it can monitor formation deformation in goaf areas, ensuring safe coal mine production. However, distributed fiber optic sensing technology still faces many challenges in energy exploration applications. For instance, its technical complexity leads to high installation and maintenance costs; and the measurement process may be affected by external environmental factors, impacting measurement accuracy. Therefore, improving the measurement accuracy, expanding the application scope, and reducing costs of distributed fiber optic sensing technology in energy exploration are urgent problems to be solved. Furthermore, the data collected by distributed optical fibers is massive and complex, containing various noise and interference. How to effectively process this data to accurately obtain energy-related information is a significant challenge. Traditional data processing methods often fail to fully extract the spatiotemporal characteristics of distributed optical fiber energy exploration data, leading to low accuracy in energy prediction and assessment. Summary of the Invention

[0003] The purpose of this invention is to provide an energy exploration method, system, device, equipment, and medium based on distributed optical fiber sensing, which can effectively improve the accuracy and efficiency of energy exploration.

[0004] This invention is achieved through the following technical solution:

[0005] In a first aspect, the first embodiment of the present invention provides an energy exploration method based on distributed optical fiber sensing, comprising the following steps:

[0006] The system receives preliminary analysis results sent by the signal processing module, which are obtained by the signal processing module through signal conversion and signal analysis of energy-related physical quantity data in the energy exploration area.

[0007] The preliminary analysis results are input into the optimized energy exploration prediction model to obtain the energy exploration prediction results.

[0008] Furthermore, the energy exploration prediction model is constructed using an Elman artificial neural network. The energy exploration prediction model includes an input layer, a hidden layer, an output layer, and a correlation unit. The input layer is used to receive the processed feature dataset. The hidden layer uses an activation function to perform nonlinear exchange on the input data. The output layer outputs prediction results related to energy. The correlation unit is used to store the information output by the hidden layer neurons at the previous time step and feeds the output information back to the input layer.

[0009] Furthermore, the method also includes: training the energy exploration prediction model after it has been constructed, specifically including:

[0010] Obtain the raw data measured by the measurement module;

[0011] The collected raw data is preprocessed to obtain preprocessed data;

[0012] Based on the patterns of energy activity and the characteristics of geological structure, the collection time is divided into multiple characteristic intervals, each interval corresponding to a different stage of energy activity. Combined with spatial location information, key feature data within each time-space unit are extracted to form a feature dataset.

[0013] The feature dataset is input into the energy exploration prediction model for training, so that the energy exploration prediction model learns the temporal and spatial variation patterns of energy-related physical quantities and the relationship between energy distribution and reserves, thus obtaining the trained energy exploration prediction model.

[0014] Furthermore, the method also includes: evaluating the trained energy exploration model using two evaluation metrics, root mean square error and mean absolute percentage error, specifically including:

[0015] Based on the accuracy requirements of energy exploration, the root mean square error threshold and the mean absolute percentage error threshold are set.

[0016] The root mean square error and mean absolute percentage error of the trained energy exploration prediction model were calculated respectively to obtain the root mean square error value and the mean absolute percentage error value.

[0017] If the obtained root mean square error and mean absolute percentage error are lower than the root mean square error threshold and the mean absolute percentage error threshold, then the trained energy exploration prediction model is deemed to meet the standard.

[0018] If the obtained root mean square error (RMSE) and mean absolute percentage error (MAS) values ​​are higher than the RMSE threshold and MAS threshold respectively, then the trained energy exploration prediction model will be optimized and adjusted.

[0019] Furthermore, the optimization and adjustment of the trained energy exploration prediction model employs the Zebra Optimization Algorithm to optimize the energy exploration prediction model. Specific methods include:

[0020] Encoding and Initialization: The training parameters in the energy exploration prediction model are encoded as the location information of individual zebras, the location of the zebra herd is randomly initialized, and the parameters of the algorithm are set.

[0021] Fitness assessment: Based on the location information of the encoded zebra individuals, a corresponding energy exploration prediction model is constructed, and the root mean square error or mean absolute percentage error of the model on the validation set is calculated. The root mean square error or mean absolute percentage error is used as the fitness value of the individual.

[0022] Leader selection and group update: The zebra with the lowest fitness value is selected as the leader, and the positions of the remaining individuals are updated according to the leader's position and the update rules in the algorithm;

[0023] Iteration and convergence judgment: Repeat the fitness evaluation, leader selection and group update steps until the maximum number of iterations is reached or the convergence condition is met, and obtain the optimal zebra individual position as the optimized model parameters;

[0024] Evaluation and application of the optimized model: The energy exploration prediction model is reconstructed using the optimized model parameters. The root mean square error and mean absolute percentage error are calculated on the test set to evaluate whether the model performance has been improved. If the expected effect is not achieved, the Zebra optimization algorithm parameters are adjusted and optimized again to obtain the optimized energy exploration prediction model.

[0025] Secondly, another embodiment of the present invention provides an energy exploration system based on distributed optical fiber sensing, comprising: a receiving module and a prediction module;

[0026] The receiving module is configured to receive the preliminary analysis results sent by the signal processing module. The preliminary analysis results are obtained by the signal processing module through signal conversion and signal analysis of energy-related physical quantity data in the energy exploration area.

[0027] The prediction module is configured to input the preliminary analysis results into the optimized energy exploration prediction model to obtain the energy exploration prediction results.

[0028] Furthermore, it also includes a model building module, which uses an Elman artificial neural network to build an energy exploration prediction model. The energy exploration prediction model includes an input layer, a hidden layer, an output layer, and a correlation unit. The input layer is used to receive the processed feature dataset. The hidden layer uses an activation function to perform nonlinear exchange on the input data. The output layer outputs prediction results related to energy. The correlation unit is used to store the information output by the hidden layer neurons at the previous time step and feeds the output information back to the input layer.

[0029] The system also includes a model training module, which is configured to acquire raw data measured by the measurement module, preprocess the acquired raw data to obtain preprocessed data, divide the acquisition time into multiple feature intervals according to the energy activity pattern and geological structure characteristics, each interval corresponding to a different stage of energy activity, and extract key feature data within each time-space unit by combining spatial location information to form a feature dataset. The feature dataset is then input into the energy exploration prediction model for training, enabling the energy exploration prediction model to learn the temporal-spatial variation patterns of energy-related physical quantities and the relationship between energy distribution and reserves, thus obtaining the trained energy exploration prediction model.

[0030] The system also includes a model optimization module, which is configured to optimize the trained energy exploration prediction model using the Zebra optimization algorithm. The model optimization module includes an encoding and initialization unit, a fitness evaluation unit, an update unit, an iteration unit, and an evaluation unit.

[0031] The encoding and initialization unit is configured to encode the training parameters in the energy exploration prediction model into the location information of individual zebras, randomly initialize the location of the zebra herd, and set the parameters of the algorithm.

[0032] The fitness evaluation unit is configured to construct a corresponding energy exploration prediction model based on the encoded zebra individual location information, calculate the root mean square error or mean absolute percentage error of the model on the validation set, and use the root mean square error or mean absolute percentage error as the fitness value of the individual.

[0033] The update unit is configured to select the zebra individual with the smallest fitness value as the leader, and the remaining individuals update their positions according to the leader's position and the update rules in the algorithm.

[0034] The iterative unit is configured to repeat the fitness evaluation and leader selection and group update steps until the maximum number of iterations is reached or the convergence condition is met, so as to obtain the optimal zebra individual position as the optimized model parameters.

[0035] The evaluation unit is configured to reconstruct the energy exploration prediction model using the optimized model parameters, calculate the root mean square error and mean absolute percentage error on the test set, evaluate whether the model performance has been improved, and if the expected effect is not achieved, then the Zebra optimization algorithm parameters are adjusted and optimized again.

[0036] Thirdly, another embodiment of the present invention provides an energy exploration device based on distributed optical fiber sensing, comprising: a light source module, a measurement module, a signal processing module, and a processor, wherein the light source module includes a light source generator and an optical isolator, and the light source generator is used to generate a broadband light source;

[0037] The optical isolator is connected to the light source generator and is used to filter broadband light sources;

[0038] The measurement module includes an optical coupler, a polarization controller, a phase modulator, and a distributed fiber optic sensor. The optical coupler is used to split the input target light source into two paths according to a preset ratio, and output them to the polarization controller and the phase modulator respectively.

[0039] The polarization controller is used to compensate for the polarization state of the optical signal;

[0040] The phase modulator is used for phase modulation of the optical signal;

[0041] The distributed fiber optic sensors are deployed underground to measure energy-related physical quantities.

[0042] The signal processing module includes a photoelectric converter and a spectrum analyzer. The photoelectric converter is used to convert the received optical signal into a voltage signal.

[0043] The spectrum analyzer receives voltage signals from the photoelectric converter, analyzes the voltage signals to obtain frequency, amplitude and power spectral density, and sends the preliminary analysis results to the processor.

[0044] The processor executes the energy exploration method based on distributed optical fiber sensing described in the above embodiments.

[0045] Fourthly, another embodiment of the present invention provides an electronic device including a processor, an input device, an output device, and a memory. The processor is connected to the input device, the output device, and the memory, respectively. The memory is used to store a computer program, the computer program including program instructions, and the processor is configured to call the program instructions to execute the method described in the above embodiments.

[0046] Fifthly, another embodiment of the present invention provides a computer-readable storage medium storing a computer program, the computer program including program instructions, which, when executed by a processor, cause the processor to perform the methods described in the above embodiments.

[0047] Compared with the prior art, the present invention has the following advantages and beneficial effects:

[0048] The energy exploration method, system, device, equipment, and medium based on distributed optical fiber sensing provided in this invention have the following advantages:

[0049] (1) Improve the accuracy of energy exploration: By using the measurement module to collect data, the signal processing module to process data, and the energy exploration prediction model to predict energy exploration, the distribution and reserves of energy can be predicted more accurately, and exploration errors can be reduced.

[0050] (2) Fully explore spatiotemporal features: Time-space feature extraction and processing can comprehensively capture the changing features of energy activities at different time and space scales, providing rich information for accurate energy assessment.

[0051] (3) Strong adaptability: The constructed energy exploration prediction model can adapt to the exploration needs of different types of energy and complex geological structures, and can be self-optimized and updated according to the actual situation, which improves the versatility and practicality of the method.

[0052] (4) Improve exploration efficiency: Accurate prediction results help optimize exploration plans, reduce unnecessary exploration workload, and thus improve the overall efficiency of energy exploration. Attached Figure Description

[0053] To more clearly illustrate the technical solutions of the exemplary embodiments of the present invention, the accompanying drawings used in the embodiments will be briefly described below. It should be understood that the following drawings only show some embodiments of the present invention and should not be considered as a limitation of the scope. For those skilled in the art, other related drawings can be obtained based on these drawings without creative effort. In the drawings:

[0054] Figure 1 is a structural block diagram of an energy exploration device based on distributed optical fiber sensing provided in the first embodiment of the present invention;

[0055] Figure 2 is a flowchart of an energy exploration method based on distributed optical fiber sensing provided in the second embodiment of the present invention;

[0056] Figure 3 is a structural block diagram of an energy exploration system based on distributed optical fiber sensing provided in the third embodiment of the present invention. Detailed Implementation

[0057] To make the objectives, technical solutions, and advantages of the present invention clearer, the present invention will be further described in detail below with reference to the embodiments and accompanying drawings. The illustrative embodiments and descriptions of the present invention are only used to explain the present invention and are not intended to limit the present invention.

[0058] Example 1

[0059] As shown in Figure 1, the first embodiment of the present invention provides a structural block diagram of an energy exploration device based on distributed optical fiber sensing. The device includes a light source module, a measurement module, a signal processing module, and a processor.

[0060] The light source module includes a light source generator and an optical isolator. The light source generator is used to generate a broadband light source. The optical isolator is connected to the light source generator and is used to filter the broadband light source, filter out noise and reflected light, and output the target light source.

[0061] The light source generator employs a broadband spontaneous emission (ASE) light source generated by a high-power erbium-doped fiber amplifier (EDFA). This broadband light source uses a semiconductor laser as a pump source to excite erbium ions from a low energy level to a high energy level, causing a population inversion within the erbium-doped fiber. Subsequently, as the high-energy particles spontaneously radiate, they transition back to a low energy level. The released energy can then stimulate the high-energy particles to radiate back to a low energy level, releasing the energy of two photons. This photon energy is then used to continue stimulated emission, resulting in a high-power light source output. When a linearly polarized semiconductor laser is used as the light source, if a portion of the output laser experiences feedback, the operation of active optical devices such as the laser or optical amplifier will become unstable, its spectrum will become cluttered, and noise will be generated in the communication system due to interference. To avoid this light feedback phenomenon, this embodiment uses an optical isolator to prevent light return, ensuring that the light propagates in a single direction.

[0062] The measurement module includes an optical coupler, a polarization controller, a phase modulator, and a distributed fiber optic sensor. The optical coupler is used to split the input target light source into two paths according to a preset ratio, and output them to the polarization controller and the phase modulator respectively. The polarization controller is used to compensate for the polarization state of the optical signal. The phase modulator is used to modulate the phase of the optical signal. The distributed fiber optic sensor, which is deployed underground, is used to measure energy-related physical quantities.

[0063] An optical coupler functions as a beam splitter to separate optical power in a desired ratio, or as a mixer to combine optical signals from different paths. An optical coupler with two output ports, where the coupling ratio can be 1:1 or 1:n, is called a beam splitter or T-coupler. For beam-splitting optical couplers, the input power is distributed between the two output ports according to the coupling ratio. In this embodiment, the optical coupler is a 2×2 optical coupler, which splits the input optical signal according to a preset ratio, outputting two optical signals.

[0064] A polarization controller is used to improve signal fading caused by polarization in propagating optical signals. When an optical signal propagates in a single-mode optical fiber, its polarization state is easily affected by the birefringence of the fiber itself or by disturbances in the external environment, leading to changes in the polarization state of the guided light. The intensity of the interference signal varies with the change in the polarization state of the light wave, thus affecting the observation of the sensing results. In severe cases, the interference signal may even be completely unmeasurable, resulting in signal fading. Therefore, this embodiment uses a polarization controller to overcome the above problems and improve measurement accuracy.

[0065] The distributed fiber optic sensor in this embodiment can be a coated fiber optic sensor. A coated fiber optic sensor has a thin film coated on an optical fiber, coupling pressure waves into the fiber, causing the fiber phase to deform as the pressure changes. Coated fiber optic sensors are suitable for distributed sensors.

[0066] The signal processing module includes a photoelectric converter and a spectrum analyzer. The photoelectric converter is used to convert the received optical signal into a voltage signal. The spectrum analyzer receives the voltage signal from the photoelectric converter, analyzes the voltage signal to obtain the frequency, amplitude and power spectral density, and sends the analysis results to the processor.

[0067] A photoelectric converter is used to receive optical signals and convert them into electrical signals. It typically employs semiconductor-based photodiodes, such as PN junction diodes, PIN diodes, or avalanche diodes. A transimpedance amplifier and a limiting amplifier are used to process the converted photocurrent. The transimpedance amplifier and limiting amplifier convert the photocurrent into a voltage signal with a small amplitude, which is then converted into a digital signal by a comparator circuit at the back end. For high-speed fiber optic communication systems, signal attenuation is often relatively severe. To avoid severe distortion of the output signal, a clock recovery circuit (CDR) and a phase-locked loop (PLL) are usually added at the back end of the receiver circuit to process the signal before output.

[0068] The main function of a spectrum analyzer is to analyze the frequency and intensity of a signal, used for signal display and analysis in measurement systems. Its input signal is a voltage signal output from a photoelectric converter; the instrument allows observation of the power spectral density of this input signal at various frequencies.

[0069] In this embodiment, the processor can be a host computer connected to a spectrum analyzer. The host computer receives preliminary analysis results from the spectrum analyzer, inputs these results into the energy exploration prediction model, outputs the energy exploration prediction results, and uploads them to the cloud management platform. The main function of the host computer is to receive the preliminary analysis results from the spectrum analyzer and call the energy exploration prediction model for in-depth analysis. This energy exploration prediction model is trained by the cloud management platform using cloud computing resources based on historical energy exploration prediction results in a database. The cloud management platform receives the energy exploration prediction results from the host computer, stores them in the database, updates the database, and synchronizes the updated model to the host computer. Alternatively, the host computer can forward the preliminary analysis results to the cloud management platform for in-depth analysis, thus reducing the computational burden on the host computer and enabling a lightweight implementation.

[0070] A method for energy exploration based on a distributed optical fiber sensing energy exploration device includes the following steps:

[0071] Step 1: After receiving the measurement command sent by the host computer, the light source generator generates a broadband light source and sends the broadband light source signal to the optical isolator via optical fiber. The optical isolator filters the broadband light source signal to remove noise and reflected light, and obtains the target light source signal.

[0072] Step 2: The optical coupler receives the target light source signal from the optical isolator and splits the target light source signal into two optical signals according to a preset ratio. The first optical signal is input to the polarization controller, and the second optical signal is input to the phase modulator.

[0073] Step 3: The polarization controller compensates for the polarization state of the first input optical signal, and the phase modulator modulates the phase of the second input optical signal, and then outputs the two optical signals to the distributed optical fiber sensor respectively.

[0074] Step 4: Based on the two optical signals mentioned above, the distributed optical fiber sensor measures and obtains energy-related physical quantity data, including a first optical measurement signal corresponding to the first optical signal and a second optical measurement signal corresponding to the second optical signal. The first optical measurement signal is input to the phase modulator, and the second optical measurement signal is input to the polarization controller. The signal is then sent to the photoelectric converter via the coupler.

[0075] Step 5: The photoelectric converter converts the received optical signal into a voltage signal and sends the converted voltage signal to the spectrum analyzer; the spectrum analyzer analyzes the voltage signal to obtain the frequency, amplitude, and power spectral density, and sends the frequency, amplitude, and power spectral density as preliminary analysis results to the host computer;

[0076] Step 6: The host computer receives the preliminary analysis results from the spectrum analyzer, inputs the preliminary analysis results into the energy exploration prediction model, outputs the energy exploration prediction results as the final analysis results, and uploads them to the cloud management platform;

[0077] Step 7: The cloud management platform receives the energy exploration prediction results from the host computer and stores them in the database; it updates the energy exploration prediction model and synchronizes the updated model to the host computer.

[0078] This embodiment presents an energy exploration device based on distributed fiber optic sensing. A light source module generates a target light source, a measurement module measures energy-related physical quantities based on the target light source, a signal processing module converts the optical signal into an electrical signal and analyzes it, and a host computer obtains the final analysis results through an energy exploration prediction model. This device achieves accurate energy exploration based on the acquisition and analysis of underground sound waves, improving efficiency and accuracy. Furthermore, this embodiment uses two optical signals for measurement, and the measurement data from both signals is used as analysis data, further improving the accuracy of the analysis results and thus enhancing the accuracy of energy exploration.

[0079] Example 2

[0080] As shown in Figure 2, the second embodiment of the present invention provides an energy exploration method based on distributed optical fiber sensing. The execution subject of this method is a processor, and it includes the following steps:

[0081] The system receives preliminary analysis results sent by the signal processing module, which are obtained by the signal processing module through signal conversion and signal analysis of energy-related physical quantity data in the energy exploration area.

[0082] The preliminary analysis results are input into the optimized energy exploration prediction model to obtain the energy exploration prediction results.

[0083] The energy exploration prediction model is constructed using an Elman artificial neural network. This Elman artificial neural network has an input layer, a hidden layer, an output layer, and a correlation unit. The correlation unit is used to store the information output by the hidden layer neurons at the previous moment and feed it back into the network, enabling the network to learn the differences caused by the timing of system inputs. The connection weights between the input layer and the hidden layer and between the hidden layer and the output layer are set. When the network is passed forward, the output of each layer is calculated. The model is trained using a feature dataset, enabling the model to learn the temporal and spatial variation patterns of energy-related physical quantities and the relationship between energy distribution and reserves.

[0084] The constructed energy exploration prediction model comprises an input layer, a hidden layer, an output layer, and a correlation unit. The input layer receives temporal and spatial feature data after feature extraction. The hidden layer uses an appropriate activation function (such as the sigmoid or tanh function) to perform a nonlinear transformation on the input data. The output layer outputs energy-related prediction results, such as energy reserve estimates and energy exploitation potential assessments. The correlation unit stores information from the previous time step of the hidden layer neurons and feeds it back to the input layer, enabling the network to learn the differences caused by the timing of system inputs, thereby better capturing the dynamic characteristics of energy activity. The model is trained using a feature dataset. By adjusting the connection weight matrix and other neural network parameters (such as the learning rate and number of iterations), the model learns the temporal and spatial variation patterns of energy-related physical quantities and the relationship between energy distribution and reserves, minimizing the error between the predicted results and actual energy parameters.

[0085] The method also includes: training the energy exploration prediction model after it has been built, specifically including:

[0086] The system acquires raw data measured by the measurement module. Specifically, it uses distributed fiber optic sensors to collect energy-related physical quantity data from multiple locations within the energy exploration area at different times. The acquisition time range covers at least one complete energy activity cycle, ensuring that the collected data encompasses different geological structures and energy states. Multiple measurement points are deployed within the energy exploration area using distributed fiber optic sensors according to a predetermined spatial layout to ensure comprehensive coverage of the entire exploration area. The distributed fiber optic sensors continuously collect energy-related physical quantity data, such as temperature, strain, and vibration. The acquisition frequency is set according to the dynamic characteristics of energy activity to ensure the capture of subtle changes in energy activity. The acquisition time span covers at least one complete energy activity cycle, such as a complete oil and gas extraction cycle or a geothermal energy heat exchange cycle, to obtain data characteristics of energy states at different stages.

[0087] The collected raw data undergoes preprocessing to obtain preprocessed data. Specifically, the raw data is cleaned to remove outliers and noise interference, and then normalized to ensure that different physical quantities have uniform dimensions and numerical ranges, facilitating subsequent analysis. Data cleaning: Outliers are identified and removed using methods based on statistical analysis and physical model constraints. For example, for temperature data, if the temperature value at a certain measurement point deviates from the temperature values ​​at surrounding measurement points by more than a certain threshold and does not conform to the temperature variation range corresponding to energy activity patterns and geological structure characteristics, it is identified as an outlier and corrected or deleted. Simultaneously, signal processing algorithms are used to remove noise interference, such as wavelet transform, to separate and filter out noise signals from the raw data. Normalization: Different physical quantities are normalized to ensure they have uniform dimensions and numerical ranges. For example, temperature data may be between 0 and 100 degrees Celsius, and strain data may be between -1 and 1. Normalization maps them to a specific interval, such as the 0-1 interval, for subsequent data analysis and model training.

[0088] The time-space feature extraction steps are as follows: Based on the energy activity patterns and geological structure characteristics, the data collection time is divided into multiple feature intervals, each corresponding to a different stage of energy activity. Combined with spatial location information, key feature data within each time-space unit is extracted to form a feature dataset. Based on energy type (e.g., oil and gas, geothermal, coal) and the geological structure characteristics of the exploration area (e.g., stratigraphic structure, rock type), the data collection time is divided into multiple physically meaningful feature intervals. For example, in oil and gas exploration, these intervals can be divided into reservoir pressure rise period, stable production period, and depletion period; in geothermal energy exploration, they can be divided into reservoir heating period, stable heat exchange period, and cooling period. Combined with spatial location information, for each time-space unit (i.e., a specific spatial region within a specific time interval), key feature data reflecting the energy state and geological characteristics are extracted. For example, the average, variance, gradient, and other statistical characteristics of physical quantity data within the spatial region, as well as the correlation characteristics between different physical quantities, are calculated to form a feature dataset. This dataset can comprehensively describe the distribution and changes of energy in time and space. The number and range of the defined characteristic intervals are dynamically adjusted based on energy type and geological structure characteristics in order to accurately capture the temporal and spatial variation characteristics of energy activities.

[0089] The feature dataset is input into the energy exploration prediction model for training, so that the energy exploration prediction model learns the temporal and spatial variation patterns of energy-related physical quantities and the relationship between energy distribution and reserves, thus obtaining the trained energy exploration prediction model.

[0090] This method also includes evaluating the trained energy exploration model using two metrics: root mean square error (RMSE) and mean absolute percentage error (MAPE). The RMSE calculation formula is as follows:

[0091] Among them, a i p represents the actual energy parameter value. i To predict energy parameter values, q represents the number of prediction samples.

[0092] The formula for calculating MAPE is:

[0093] Thresholds for RMSE and MAPE are set according to the accuracy requirements of energy exploration. When the model evaluation index is below the threshold, the trained energy exploration model is considered to have met the requirements. If the model evaluation index is above the threshold, the trained energy exploration prediction model is optimized and adjusted. Optimization measures include increasing the amount of training data, such as collecting data from more time points or spatial locations; adjusting neural network parameters, such as reducing the learning rate to improve model convergence stability, and increasing the number of hidden layer neurons to enhance the model's expressive power, until the model meets the requirements. In this embodiment, the Zebra Optimization Algorithm (ZOA) is used to optimize the trained energy exploration prediction model. Once the model evaluation meets the requirements, it is applied to energy prediction within a distributed optical fiber energy exploration area.

[0094] The Zebra Optimization Algorithm (ZOA) is used to optimize the trained energy exploration prediction model. The specific steps include:

[0095] Encoding and Initialization: The training parameters in the energy exploration prediction model are encoded into the location information of individual zebras. The training parameters include at least the connection weights between the input layer and the hidden layer, the connection weights between the hidden layer and the output layer, and the number of neurons in the hidden layer in the Elman artificial neural network. The location of the zebra population is randomly initialized, and the parameters of the algorithm are set, including the population size and the maximum number of iterations.

[0096] Fitness assessment: Based on the location of the encoded zebra individuals, a corresponding energy exploration prediction model is constructed. The RMSE or MAPE of the model on the validation set is calculated as the fitness value of the individual. The smaller the fitness value, the better the model performance.

[0097] Leader selection and group update: The zebra individual with the smallest fitness value is selected as the leader. The remaining individuals are updated in position according to the leader's position and the update rules in the algorithm. The update rules include foraging and defense strategies, exploring new parameter spaces, and ensuring that the new position meets the constraints. The constraints include at least the thresholds of the root mean square error (RMSE) and the mean absolute percentage error (MAPE).

[0098] Iteration and convergence judgment: Repeat the fitness evaluation, leader selection and group update steps above until the maximum number of iterations is reached or the convergence condition is met. The optimal zebra individual position obtained at this time is the optimized model parameter.

[0099] Evaluation and application of the optimized model: The energy exploration prediction model is reconstructed using the optimized model parameters. RMSE and MAPE are calculated on the test set to evaluate whether the model performance has been improved. If the expected effect is not achieved, the ZOA algorithm parameters are adjusted and optimized again to obtain the optimized energy exploration prediction model.

[0100] Once the model meets the evaluation requirements, it is applied to energy prediction at unknown locations within the distributed fiber optic energy exploration area. During application, the model can self-optimize and update based on newly acquired real-time data. For example, when the geological structure within the exploration area changes (such as stratigraphic fractures, rock deformation, etc.) or the energy state undergoes dynamic changes (such as changes in reservoir pressure due to oil and gas extraction, changes in heat exchange efficiency due to geothermal energy extraction, etc.), the model can automatically adjust its internal parameters to adapt to the new situation and continuously provide accurate energy exploration prediction results.

[0101] This invention provides an energy exploration method based on distributed optical fiber sensing. By acquiring and processing distributed optical fiber data and using an energy exploration prediction model to predict energy exploration, it can comprehensively capture the changing characteristics of energy activity at different temporal and spatial scales. This enables more accurate prediction of energy distribution and reserves, improving the overall efficiency of energy exploration and reducing exploration errors. The energy exploration prediction model can adapt to the exploration needs of different types of energy and complex geological structures, and can self-optimize and update according to actual conditions.

[0102] Example 3:

[0103] As shown in Figure 3, the third embodiment of the present invention provides an energy exploration system based on distributed optical fiber sensing, used to implement an energy exploration method based on distributed optical fiber sensing described in the second embodiment. The system includes a receiving module and a prediction module. The receiving module is configured to receive preliminary analysis results sent by the signal processing module. The preliminary analysis results are obtained by the signal processing module through signal conversion and signal analysis of energy-related physical quantity data in the energy exploration area. The prediction module is configured to input the preliminary analysis results into an optimized energy exploration prediction model to predict and obtain energy exploration prediction results.

[0104] The energy exploration system based on distributed optical fiber sensing also includes a model building module. The model building module uses an Elman artificial neural network to construct an energy exploration prediction model. The energy exploration prediction model includes an input layer, a hidden layer, an output layer, and a correlation unit. The input layer is used to receive the processed feature dataset. The hidden layer uses an activation function to perform nonlinear exchange on the input data. The output layer outputs prediction results related to energy. The correlation unit is used to store the information output by the hidden layer neurons at the previous time step and feeds the output information back to the input layer.

[0105] The system also includes a model training module, which is configured to acquire the raw data measured by the measurement module, preprocess the acquired raw data to obtain preprocessed data, divide the acquisition time into multiple feature intervals according to the energy activity patterns and geological structure characteristics, with each interval corresponding to a different stage of energy activity, and extract key feature data within each time-space unit in combination with spatial location information to form a feature dataset. The feature dataset is then input into the energy exploration prediction model for training, enabling the energy exploration prediction model to learn the temporal-spatial variation patterns of energy-related physical quantities and the relationship between energy distribution and reserves, thus obtaining the trained energy exploration prediction model.

[0106] The system also includes a model optimization module, which is configured to optimize the trained energy exploration prediction model using the Zebra optimization algorithm. The model optimization module includes an encoding and initialization unit, a fitness evaluation unit, an update unit, an iteration unit, and an evaluation unit.

[0107] The encoding and initialization unit is configured to encode the training parameters in the energy exploration prediction model into the location information of individual zebras, randomly initialize the location of the zebra herd, and set the parameters of the algorithm.

[0108] The fitness evaluation unit is configured to construct a corresponding energy exploration prediction model based on the encoded zebra individual location information, calculate the root mean square error or mean absolute percentage error of the model on the validation set, and use the root mean square error or mean absolute percentage error as the fitness value of the individual.

[0109] The update unit is configured to select the zebra individual with the smallest fitness value as the leader, and the remaining individuals update their positions according to the leader's position and the update rules in the algorithm.

[0110] The iterative unit is configured to repeat the fitness evaluation and leader selection and group update steps until the maximum number of iterations is reached or the convergence condition is met, and the optimal zebra individual position is obtained as the optimized model parameters.

[0111] The evaluation unit is configured to reconstruct the energy exploration prediction model using the optimized model parameters, calculate the root mean square error and mean absolute percentage error on the test set, evaluate whether the model performance has been improved, and if the expected effect is not achieved, the Zebra optimization algorithm parameters are adjusted and optimized again.

[0112] This invention provides an energy exploration system based on distributed optical fiber sensing. By acquiring, processing, and utilizing distributed optical fiber data and employing an energy exploration prediction model, it can comprehensively capture the changing characteristics of energy activity at different temporal and spatial scales. This enables more accurate prediction of energy distribution and reserves, improving the overall efficiency of energy exploration and reducing exploration errors. The energy exploration prediction model can adapt to the exploration needs of different types of energy and complex geological structures, and can self-optimize and update according to actual conditions.

[0113] Example 4

[0114] The fourth embodiment of the present invention provides an electronic device, which includes a processor, an input device, an output device, and a memory. The processor is connected to the input device, the output device, and the memory. The memory is used to store a computer program, which includes program instructions. The processor is configured to call the program instructions to execute the method described in the second embodiment above.

[0115] It should be understood that, in the embodiments of the present invention, the processor may be a Central Processing Unit (CPU), but it may also be other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. A general-purpose processor may be a microprocessor or any conventional processor.

[0116] Input devices may include touchpads, fingerprint sensors (for collecting the user's fingerprint information and fingerprint orientation information), microphones, etc., while output devices may include displays (LCDs, etc.), speakers, etc.

[0117] The memory may include read-only memory and random access memory, and provides instructions and data to the processor. A portion of the memory may also include non-volatile random access memory. For example, the memory may also store information about the device type.

[0118] In specific implementations, the processor, input device, and output device described in the embodiments of the present invention can execute the implementation methods described in the method embodiments of the present invention, or they can execute the implementation methods described in the system embodiments of the present invention, which will not be repeated here.

[0119] Example 5

[0120] The fifth embodiment of the present invention also provides an embodiment of a computer-readable storage medium storing a computer program, the computer program including program instructions, which, when executed by a processor, cause the processor to perform the method described in the second embodiment above.

[0121] The computer-readable storage medium can be an internal storage unit of the terminal described in the foregoing embodiments, such as the terminal's hard drive or memory. The computer-readable storage medium can also be an external storage device of the terminal, such as a plug-in hard drive, Smart Media Card (SMC), Secure Digital (SD) card, or Flash Card equipped on the terminal. Further, the computer-readable storage medium can include both internal storage units and external storage devices of the terminal. The computer-readable storage medium is used to store the computer program and other programs and data required by the terminal. The computer-readable storage medium can also be used to temporarily store data that has been output or will be output.

[0122] Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, computer software, or a combination of both. To clearly illustrate the interchangeability of hardware and software, the components and steps of the various examples have been generally described in terms of functionality in the foregoing description. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementations should not be considered beyond the scope of this invention.

[0123] Those skilled in the art will understand that, for the sake of convenience and brevity, the specific working processes of the terminals and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0124] In the several embodiments provided in this application, it should be understood that the disclosed systems and methods can be implemented in other ways. For example, the device embodiments described above are merely illustrative. For instance, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. In addition, the mutual coupling or direct coupling or communication connection shown or discussed may be indirect coupling or communication connection through some interfaces, devices or units, or may be electrical, mechanical or other forms of connection.

[0125] The specific embodiments described above further illustrate the purpose, technical solution, and beneficial effects of the present invention. It should be understood that the above description is only a specific embodiment of the present invention and is not intended to limit the scope of protection of the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. An energy exploration method based on distributed optical fiber sensing, characterized in that, Includes the following steps: The system receives preliminary analysis results sent by the signal processing module, which are obtained by the signal processing module through signal conversion and analysis of energy-related physical quantity data in the exploration area. The preliminary analysis results are input into the optimized energy exploration prediction model to obtain the energy exploration prediction results.

2. The energy exploration method based on distributed optical fiber sensing according to claim 1, characterized in that, The energy exploration prediction model is constructed using an Elman artificial neural network. The energy exploration prediction model includes an input layer, a hidden layer, an output layer, and a correlation unit. The input layer is used to receive the processed feature dataset. The hidden layer uses an activation function to perform nonlinear exchange on the input data. The output layer outputs prediction results related to energy. The correlation unit is used to store the information output by the hidden layer neurons at the previous time step and feeds the output information back to the input layer.

3. The energy exploration method based on distributed optical fiber sensing according to claim 2, characterized in that, The method further includes: training the energy exploration prediction model after it has been constructed, specifically including: Obtain the raw data measured by the measurement module; The collected raw data is preprocessed to obtain preprocessed data; Based on the patterns of energy activity and the characteristics of geological structure, the collection time is divided into multiple characteristic intervals, each interval corresponding to a different stage of energy activity. Combined with spatial location information, key feature data within each time-space unit are extracted to form a feature dataset. The feature dataset is input into the energy exploration prediction model for training, so that the energy exploration prediction model learns the temporal and spatial variation patterns of energy-related physical quantities and the relationship between energy distribution and reserves, thus obtaining the trained energy exploration prediction model.

4. The energy exploration method based on distributed optical fiber sensing according to claim 3, characterized in that, The method further includes evaluating the trained energy exploration model using two evaluation metrics: root mean square error and mean absolute percentage error. Specifically, this includes: Based on the accuracy requirements of energy exploration, the root mean square error threshold and the mean absolute percentage error threshold are set. The root mean square error and mean absolute percentage error of the trained energy exploration prediction model were calculated respectively to obtain the root mean square error value and the mean absolute percentage error value. If the obtained root mean square error and mean absolute percentage error are lower than the root mean square error threshold and the mean absolute percentage error threshold, then the trained energy exploration prediction model is deemed to meet the standard. If the obtained root mean square error (RMSE) and mean absolute percentage error (MAS) values ​​are higher than the RMSE threshold and MAS threshold respectively, then the trained energy exploration prediction model will be optimized and adjusted.

5. The energy exploration method based on distributed optical fiber sensing according to claim 4, characterized in that, The optimization and adjustment of the trained energy exploration prediction model employs the Zebra optimization algorithm, and the specific methods include: Encoding and Initialization: The training parameters in the energy exploration prediction model are encoded as the location information of individual zebras, the location of the zebra herd is randomly initialized, and the parameters of the algorithm are set. Fitness assessment: Based on the location information of the encoded zebra individuals, a corresponding energy exploration prediction model is constructed, and the root mean square error or mean absolute percentage error of the model on the validation set is calculated. The root mean square error or mean absolute percentage error is used as the fitness value of the individual. Leader selection and group update: The zebra with the lowest fitness value is selected as the leader, and the positions of the remaining individuals are updated according to the leader's position and the update rules in the algorithm; Iteration and convergence judgment: Repeat the fitness evaluation, leader selection and group update steps until the maximum number of iterations is reached or the convergence condition is met, and obtain the optimal zebra individual position as the optimized model parameters; Evaluation and application of the optimized model: The energy exploration prediction model is reconstructed using the optimized model parameters. The root mean square error and mean absolute percentage error are calculated on the test set to evaluate whether the model performance has been improved. If the expected effect is not achieved, the Zebra optimization algorithm parameters are adjusted and optimized again to obtain the optimized energy exploration prediction model.

6. An energy exploration system based on distributed optical fiber sensing, characterized in that, include: Receive module and prediction module; The receiving module is configured to receive the preliminary analysis results sent by the signal processing module. The preliminary analysis results are obtained by the signal processing module through signal conversion and signal analysis of energy-related physical quantity data in the energy exploration area. The prediction module is configured to input the preliminary analysis results into the optimized energy exploration prediction model to obtain the energy exploration prediction results.

7. The energy exploration system based on distributed optical fiber sensing according to claim 6, characterized in that, The system also includes a model building module, which uses an Elman artificial neural network to build an energy exploration prediction model. The energy exploration prediction model includes an input layer, a hidden layer, an output layer, and a correlation unit. The input layer is used to receive the processed feature dataset. The hidden layer uses an activation function to perform nonlinear exchange on the input data. The output layer outputs prediction results related to energy. The correlation unit is used to store the information output by the hidden layer neurons at the previous moment and feed the output information back to the input layer. The system also includes a model training module, which is configured to acquire raw data measured by the measurement module, preprocess the acquired raw data to obtain preprocessed data, divide the acquisition time into multiple feature intervals according to the energy activity pattern and geological structure characteristics, each interval corresponding to a different stage of energy activity, and extract key feature data within each time-space unit by combining spatial location information to form a feature dataset. The feature dataset is then input into the energy exploration prediction model for training, enabling the energy exploration prediction model to learn the temporal-spatial variation patterns of energy-related physical quantities and the relationship between energy distribution and reserves, thus obtaining the trained energy exploration prediction model. The system also includes a model optimization module, which is configured to optimize the trained energy exploration prediction model using the Zebra optimization algorithm. The model optimization module includes an encoding and initialization unit, a fitness evaluation unit, an update unit, an iteration unit, and an evaluation unit. The encoding and initialization unit is configured to encode the training parameters in the energy exploration prediction model into the location information of individual zebras, randomly initialize the location of the zebra herd, and set the parameters of the algorithm. The fitness evaluation unit is configured to construct a corresponding energy exploration prediction model based on the encoded zebra individual location information, calculate the root mean square error or mean absolute percentage error of the model on the validation set, and use the root mean square error or mean absolute percentage error as the fitness value of the individual. The update unit is configured to select the zebra individual with the smallest fitness value as the leader, and the remaining individuals update their positions according to the leader's position and the update rules in the algorithm. The iterative unit is configured to repeat the fitness evaluation and leader selection and group update steps until the maximum number of iterations is reached or the convergence condition is met, so as to obtain the optimal zebra individual position as the optimized model parameters. The evaluation unit is configured to reconstruct the energy exploration prediction model using the optimized model parameters, calculate the root mean square error and mean absolute percentage error on the test set, evaluate whether the model performance has been improved, and if the expected effect is not achieved, then the Zebra optimization algorithm parameters are adjusted and optimized again.

8. An energy exploration device based on distributed optical fiber sensing, characterized in that, include: The system includes a light source module, a measurement module, a signal processing module, and a processor. The light source module includes a light source generator and an optical isolator. The light source generator is used to generate a broadband light source. The optical isolator is connected to the light source generator and is used to filter broadband light sources; The measurement module includes an optical coupler, a polarization controller, a phase modulator, and a distributed fiber optic sensor. The optical coupler is used to split the input target light source into two paths according to a preset ratio, and output them to the polarization controller and the phase modulator respectively. The polarization controller is used to compensate for the polarization state of the optical signal; The phase modulator is used for phase modulation of the optical signal; The distributed fiber optic sensors are deployed underground to measure energy-related physical quantities. The signal processing module includes a photoelectric converter and a spectrum analyzer. The photoelectric converter is used to convert the received optical signal into a voltage signal. The spectrum analyzer receives voltage signals from the photoelectric converter, analyzes the voltage signals to obtain frequency, amplitude and power spectral density, and sends the preliminary analysis results to the processor. The processor executes the energy exploration method based on distributed optical fiber sensing as described in any one of claims 1-5.

9. An electronic device comprising a processor, an input device, an output device, and a memory, wherein the processor is connected to the input device, the output device, and the memory, and the memory is used to store a computer program, the computer program comprising program instructions, characterized in that, The processor is configured to invoke the program instructions to perform the method as described in any one of claims 1-5.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program, the computer program including program instructions that, when executed by a processor, cause the processor to perform the method as described in any one of claims 1-5.

Images