An intelligent first arrival picking method and system

Abstract

This invention provides an intelligent first arrival picking method and system, belonging to the field of geophysical exploration technology. The method uses the energy similarity attributes of seismic data and seismic data as dual-channel data in the input layer, and utilizes a network model to pick the first arrival. The method includes: (1) shot gather data preprocessing: reading seismic data and performing seismic trace completion processing; (2) preparing network input data: dynamic correction processing, seismic data normalization, and energy similarity attribute calculation; (3) preparing label data: first arrival travel time interpolation processing at bad and missing traces, and mapping the first arrival travel time to the three-dimensional data volume; (4) network model training: training the 3DUnet network to obtain a trained network model; (5) first arrival picking: inputting the preprocessed seismic data and using the trained network model to obtain the first arrival travel time. The method of this invention has high first arrival picking efficiency, reduces manual intervention, and improves the efficiency and accuracy of first arrival picking.

Description

TECHNICAL FIELD

[0001] The present application belongs to the technical field of geophysical exploration, and particularly relates to an intelligent first arrival picking method and system, which realizes first arrival picking through first arrival picking oriented shot gather data preprocessing technology and a double-channel 3DUnet network and can be used for first arrival picking processing in the preprocessing process of petroleum geophysical exploration. BACKGROUND

[0002] First arrival picking is an important link in seismic data processing, and the precision of first arrival picking directly affects the quality of static correction and imaging effect. Traditional first arrival picking methods mainly include energy ratio method, seismic trace instantaneous intensity ratio method, wavelet amplitude ratio method and the like, which identify first arrival by using the characteristics of waveforms in energy, frequency and the like, but the application effect is poor on seismic data with medium and low signal-to-noise ratio, and most of them are picked up channel by channel, the correlation between multiple channels is used to a low degree, and the lateral continuity is poor. In addition, for different exploration areas, different first arrival picking methods suitable for the exploration area need to be selected according to the actual situation of the exploration area, and manual parameter adjustment is required, manual intervention is high, and it is time-consuming and laborious to pick up first arrival on massive data, and the efficiency is low.

[0003] In recent years, deep learning technology has developed rapidly and is widely used in the field of seismic data processing, showing great application prospects. If seismic data is regarded as an image, first arrival picking can be regarded as an image segmentation problem. Under this assumption, first arrival picking methods based on full convolution network FCN, residual neural network ResNet and U-shaped full convolution network Unet have been proposed, among which the first arrival picking based on Unet is the most studied. Unet splices shallow features and deep features through skip layer connection, so that the output of the network integrates global features and local texture features of the image, and has the characteristics of simplicity, efficiency and small demand for training samples. Although there are many studies on first arrival picking based on Unet, there are few application studies on it in actual data processing. SUMMARY

[0004] The present application aims to solve the problems existing in the prior art, and provides an intelligent first arrival picking method and system, which realizes intelligent first arrival picking on shot gather data by using 3DUnet, reduces the degree of manual intervention in first arrival picking, improves the efficiency and precision of first arrival picking, and better serves the static correction in seismic data processing, thereby providing strong technical support for shortening the production cycle and accelerating the seismic exploration process.

[0005] The present application is realized by the following technical solutions:

[0006] In a first aspect, the present application provides an intelligent first arrival picking method, which takes the energy similarity attribute of seismic data and seismic data as double-channel data of the input layer, and picks up first arrival by using a network model.

[0007] Preferably, the method comprises:

[0008] (1) Gun set data preprocessing: reading seismic data and seismic trace padding processing;

[0009] (2) Prepare network input data: moveout correction processing, seismic data normalization and energy similarity attribute calculation;

[0010] (3) Prepare label data: first arrival travel time interpolation processing at bad and missing channels, and mapping of first arrival travel time to three-dimensional data volume;

[0011] (4) Network model training: training 3DUnet network to obtain trained network model;

[0012] (5) First arrival picking: input preprocessed seismic data, and obtain first arrival travel time using the trained network model.

[0013] Preferably, the operation of step (1) comprises:

[0014] (11) Reading seismic data;

[0015] (12) According to shot number and receiver line number, the data is divided into multiple three-dimensional gun set data volumes;

[0016] (13) According to the set maximum number of each arrangement N, seismic trace padding processing is performed on each three-dimensional gun set data volume.

[0017] Preferably, the operation of step (2) comprises:

[0018] (21) Using offset and moveout correction velocity, moveout correction processing is performed on each three-dimensional gun set data volume to obtain moveout corrected gun set data;

[0019] (22) On the moveout corrected gun set data, the seismic data between zero time and the latest arrival time of the first arrival wave is intercepted as the to-be-processed data, and the to-be-processed data is normalized to [-1, 1] as the first channel data of the input layer;

[0020] (23) Calculate the energy similarity attribute of the to-be-processed data as the second channel data of the input layer.

[0021] Preferably, the operation of step (3) comprises:

[0022] Reading first arrival travel time, and using the first arrival travel time of adjacent seismic traces to perform interpolation processing at bad and missing channels;

[0023] Label data is generated based on the first arrival travel time of each seismic trace: the following processing is performed on each seismic trace in sequence: data with time less than the first arrival travel time are assigned a value of 0, and data with time greater than or equal to the first arrival travel time are assigned a value of 1; after processing all seismic traces, the first arrival travel time of all seismic traces is mapped into a three-dimensional data volume that is either 0 or 1, i.e., label data.

[0024] Preferably, step (4) includes: training the 3DUnet network using the training sample label set to obtain a trained network model; the training sample label set includes first channel data, second channel data and label data.

[0025] Preferably, step (5) includes:

[0026] (51) Process the seismic data to be processed using the same method as in steps (1) and (2) to obtain the first channel data and the second channel data;

[0027] (52) Input the first channel data and the second channel data into the trained network model, and the trained network model outputs the predicted three-dimensional data volume;

[0028] (53) Map the predicted three-dimensional data volume back to the initial arrival time data.

[0029] Preferably, step (53) includes the following operations:

[0030] The following processing is performed sequentially on each seismic trace in the predicted 3D data volume:

[0031] (531) The first time when the value on the seismic trace is greater than 0.5 is taken as the first arrival time;

[0032] (532) The initial arrival time of the seismic trace is obtained by adding the dynamic correction amount to the initial arrival time of the picked-up first arrival time.

[0033] A second aspect of the present invention provides an intelligent initial arrival pickup system, the system comprising:

[0034] The preprocessing unit is used to read seismic data and perform seismic trace completion processing;

[0035] The input data preparation unit, connected to the preprocessing unit, is used for dynamic correction processing, seismic data normalization, and energy similarity attribute calculation.

[0036] The tag data preparation unit is connected to the preprocessing unit and the input data preparation unit respectively. It is used for first arrival time interpolation processing at bad and missing paths and mapping of first arrival time to three-dimensional data volume.

[0037] The network model training unit is connected to the input data preparation unit and the label data preparation unit respectively, and is used to train the 3DUnet network to obtain the trained network model.

[0038] The first arrival picking unit is connected to the preprocessing unit, the input data preparation unit, and the network model training unit, respectively. It is used to input the preprocessed seismic data and obtain the first arrival travel time using the trained network model.

[0039] A third aspect of the present invention provides a computer-readable storage medium storing at least one computer-executable program, which, when executed by the computer, causes the computer to perform steps in an intelligent arrival picking method of the present invention.

[0040] Compared with the prior art, the beneficial effects of the present invention are: the method of the present invention has high initial arrival picking efficiency, reduces the degree of manual intervention, improves the initial arrival picking efficiency and accuracy, has good stability, and shows good application effect under moderately complex geological conditions. Attached Figure Description

[0041] Figure 1 A flowchart illustrating the steps of the method of this invention;

[0042] Figure 2 This is a schematic diagram of the network structure used in this invention;

[0043] Figure 3a The results of 3DUnet initial arrival picking in a work area in western China using the method of this invention;

[0044] Figure 3b The initial arrival and pickup results were obtained using commercial software in a work area in western China.

[0045] Figure 3c This is a partial comparison of the results obtained by the two methods. Detailed Implementation

[0046] The present invention will now be described in further detail with reference to the accompanying drawings:

[0047] To address the problem of first-arrival picking in real-world data, this invention proposes a fast and efficient intelligent first-arrival picking process using 3DUnet. This invention ensures data continuity through first-arrival interpolation on bad and missing channels, extracts the energy similarity attributes of seismic data and uses them, along with the seismic data, as dual-channel input data for the layer. It then utilizes a network model to pick first-arrivals, improving the accuracy of first-arrival picking, and reduces computational load by applying linear dynamic correction to the seismic data.

[0048] like Figure 2As shown, the 3DUnet network used for initial pickup in this invention (see reference: 3D U-Net: Learning Dense Volumetric Segmentation from Sparse Annotation) consists of an encoder and a decoder. The basic network structure of the encoder includes convolutional layers, batch normalization, and max pooling layers. The decoder includes upper convolutional layers, batch normalization, and convolutional layers. The convolutional kernel size is 3*3*3. Except for the output layer, which uses sigmoid as the activation function, all other layers use ReLU as the activation function. There are skip connections between the encoder and decoder to achieve the concatenation of shallow and deep features.

[0049] In actual data processing, the specific network structure (number of convolutional layers and number of channels in each convolutional layer, etc.) can be adjusted according to the shot data.

[0050] The loss function used in this invention is the Dice function, and its formula is as follows:

[0051]

[0052] Among them, Y pred Y is the network output value. label Let be the label, and ε be the smoothing coefficient, which is set to 1.0 in this invention. The Adam algorithm is used to continuously update the network model by minimizing the loss function until a set number of iterations is reached or the loss function value stabilizes, resulting in the final trained 3DUnet network.

[0053] When applied to real-world data, data preprocessing and the preparation of training sample label sets are crucial components of the first-arrival picking process in the 3DUnet network. The training sample label set consists of two parts: input and labels. The input includes two channels: seismic data and energy similarity attributes. The labels are three-dimensional data volumes mapped from the first-arrival travel times.

[0054] To reduce computational load, linear dynamic correction is performed on the shot gather data. The formula for calculating the dynamic correction is as follows:

[0055]

[0056] Where x is the shot-receiver distance; v0 is the dynamic correction speed, which is a constant value.

[0057] Each seismic trace in the shot gather is shifted by Δt along the time axis in the direction of decreasing time (shifting the seismic trace in the direction of decreasing time changes the time corresponding to each sampling point on the seismic trace), resulting in dynamically corrected shot gather data. From the dynamically corrected shot gather data, seismic data between time zero and the latest arrival time of the first arrival wave is extracted (a complete shot gather dataset contains not only the first arrival wave but also reflected waves, etc. Theoretically, the first arrival wave is the first wave to jump in the seismic trace; the goal of this invention is to identify the first arrival, therefore, the seismic data corresponding to the first arrival waveform is extracted as network input), and this data is normalized to the range [-1, 1], serving as the first channel data of the input layer. Simultaneously, the energy similarity attribute of this data segment is calculated, serving as the second channel data of the input layer. The label size is consistent with the input layer size ("size" refers to the number of sampling points in the x, y, and z directions of the data volume), with 0 above the first arrival and 1 below the first arrival.

[0058] To ensure consistent data size in the training sample label set, the maximum number of traces in each permutation was set to N. For single permutations with an actual number of traces less than N, the data was padded using seismic traces from right to left. Furthermore, to ensure the continuity of first arrival travel times and eliminate outliers, interpolation was performed on the first arrival data at bad and empty traces.

[0059] The method of this invention includes a series of techniques such as shot gather data preprocessing for first arrival picking, preparation of training sample label set, network model training and inference, to achieve intelligent picking of first arrivals in shot gather data.

[0060] like Figure 1 As shown, the method of the present invention includes the following steps:

[0061] (1) Preprocessing of shot gather data: reading seismic data and completing seismic traces;

[0062] (2) Prepare network input data: dynamic correction processing, seismic data normalization and energy similarity attribute calculation;

[0063] (3) Prepare label data: interpolation of first arrival times at bad and missing paths, and mapping of first arrival times to three-dimensional data volumes;

[0064] (4) Network model training: Train the 3DUnet network using the sample label set to obtain the trained network model;

[0065] (5) First arrival picking: Input the preprocessed seismic data and use the trained network model to obtain the first arrival travel time.

[0066] Compared with existing technologies, the present invention has the advantages of high initial arrival picking efficiency, low degree of human intervention, and good stability.

[0067] The embodiments of the method of the present invention are as follows:

[0068] Example 1:

[0069] The operations in step (1) include:

[0070] (11) Read seismic data;

[0071] (12) According to the shot number and geophone line number, split the data into multiple 3D shot gather data volumes;

[0072] (13) According to the maximum number of traces N set for each array, perform seismic trace padding processing on each 3D shot gather data volume.

[0073] Among them, the operations in step (12) include:

[0074] Select all seismic traces with the same shot number to form a 3D shot gather data volume, where the seismic traces with the same geophone line number are placed in the same row to form an array; each seismic trace in each array corresponds to a different trace number, and a 3D shot gather data volume is composed of multiple arrays arranged side by side. The three dimensions of this 3D shot gather data volume are geophone line number, trace number, and time respectively.

[0075] Among them, the operations in step (13) include:

[0076] Assume that the single-array data has a total of M traces. If M < N, then select N - M seismic traces from right to left in this array in sequence to perform padding processing on this array, so that the total number of seismic traces in the processed single-array data is N.

[0077] Example 2:

[0078] The operations in step (2) include:

[0079] (21) Use the shot-receiver offset and NMO velocity to perform NMO processing on each 3D shot gather data volume to obtain the NMO-processed shot gather data, specifically including:

[0080] (211) Use the following formula to obtain the NMO correction amount Δt for each seismic trace:

[0081]

[0082] Among them, x is the shot-receiver offset; v0 is the NMO velocity, which is a constant value and can be estimated by obtaining the slope of the first arrival wave on the single-shot record;

[0083] (212) Sequentially move each seismic trace in the shot gather along the time axis in the direction of decreasing time by the NMO correction amount Δt corresponding to this seismic trace to obtain the NMO-processed shot gather data.

[0084] (22) On the shot gather data after dynamic correction, the seismic data between the zero time and the latest arrival time of the first arrival wave are extracted as the data to be processed, and the data to be processed is normalized to the range [-1,1] as the first channel data of the input layer, as follows:

[0085] The operation of normalizing the data to be processed to the range [-1, 1] includes:

[0086] The data to be processed is normalized to the range [-1, 1] using the following formula:

[0087]

[0088] Where X represents the original seismic data, ||X|| ∞ Let X be the infinite norm of the seismic data. norm This is the normalized seismic data.

[0089] (23) Calculate the energy similarity attribute of the data to be processed as the second channel data of the input layer, thus forming dual-channel data used as network input, as follows:

[0090] The energy similarity property is obtained by calculating using the following formula:

[0091]

[0092] Where X(i,j,l) is the seismic data at location (i,j,l), W is the size of the smoothing window, and Y(i,j,k) is the energy similarity attribute at location (i,j,k).

[0093] The energy similarity attribute describes the energy intensity variation relationship of seismic data, and the first arrival waveform corresponds to the location of strong energy. Introducing this attribute is more conducive to picking the first arrival than using seismic data alone, and can improve the accuracy of first arrival picking.

[0094] The network input used for first-arrival picking only needs to contain the seismic data segment with the first arrival wave. After linear dynamic correction of the seismic data, the number of longitudinal sampling points of the seismic data containing the first arrival wave is reduced, which can reduce the number of convolution operations in each convolutional layer of the network model, thereby reducing the amount of computation.

[0095] Example 3:

[0096] Step (3) includes the following operations:

[0097] Read the first arrival time, and interpolate it using the first arrival times of nearby seismic traces at bad and empty traces; generate label data based on the first arrival times of each seismic trace.

[0098] The process of generating label data based on the first arrival travel time of each seismic trace includes:

[0099] The following processing was performed on each seismic trace in sequence:

[0100] Data with times less than the initial arrival time are assigned a value of 0, and data with times greater than or equal to the initial arrival time are assigned a value of 1.

[0101] The initial arrival time here can be obtained through commercial software or manual collection.

[0102] After all seismic traces have been processed, the first arrival travel times of all seismic traces are mapped into a three-dimensional data volume that is either 0 or 1, i.e., label data.

[0103] Example 4:

[0104] Step (4) includes: based on the 3DUnet network, using the training sample label set, setting the learning rate, using the Adam algorithm to train the network, and saving the trained network model. The training sample label set includes the first channel data, the second channel data, and the label data.

[0105] Example 5:

[0106] Step (5) includes the following operations:

[0107] (51) Process the seismic data to be processed using the same method as in steps (1) and (2) to obtain the first channel data and the second channel data;

[0108] (52) Input the first channel data and the second channel data into the trained network model, and the trained network model outputs the predicted three-dimensional data volume;

[0109] (53) Map the predicted three-dimensional data volume back to the initial arrival time data.

[0110] The operation in step (53) includes:

[0111] The following processing is performed sequentially on each seismic trace in the predicted 3D data volume:

[0112] (531) The first time when the value on the seismic trace is greater than 0.5 is taken as the first arrival time;

[0113] (532) The first arrival time of the seismic trace is obtained by adding the dynamic correction amount to the first arrival time of the seismic trace. That is, the first arrival time of the seismic trace obtained by 3DUnet is added to the dynamic correction amount of the seismic trace to obtain the final first arrival time of the seismic trace.

[0114] A second aspect of the present invention provides an intelligent initial arrival pickup system, wherein embodiments of the system are as follows:

[0115] Example 6:

[0116] The system includes:

[0117] The preprocessing unit is used to read seismic data and perform seismic trace completion processing;

[0118] The input data preparation unit, connected to the preprocessing unit, is used for dynamic correction processing, seismic data normalization, and energy similarity attribute calculation.

[0119] The tag data preparation unit is connected to the preprocessing unit and the input data preparation unit respectively. It is used for first arrival time interpolation processing at bad and missing paths and mapping of first arrival time to three-dimensional data volume.

[0120] The network model training unit is connected to the input data preparation unit and the label data preparation unit respectively, and is used to train the 3DUnet network to obtain the trained network model.

[0121] The first arrival picking unit is connected to the preprocessing unit, the input data preparation unit, and the network model training unit, respectively. It is used to input the preprocessed seismic data and obtain the first arrival travel time using the trained network model.

[0122] The intelligent first arrival picking method proposed in this invention was applied to shot gather data from a seismic survey area in western my country. A training sample label set was created using 100 shot gathers from this survey area and their first arrival times. The network structure designed in this invention was trained using this method, with a training cycle of 200. After the network training was completed, other shot gather data from the same survey area were preprocessed (linear dynamic correction, seismic trace completion, seismic data normalization, and energy similarity attribute calculation). The preprocessed 3D data volume and energy similarity attributes were then input into the network model to obtain the picked first arrivals (e.g., ...). Figure 3a (as shown) and the results picked up by commercial software (such as) Figure 3b The results are basically the same as shown, but the first arrival continuity of this invention, based on the 3DUnet network, is better. Figure 3c The left image shows the first arrival time obtained using the method of this invention, and the right image shows the result obtained by commercial software. The black line represents the obtained first arrival time. Figure 3c As shown by the black line, the first arrival travel time of commercial software is suddenly reduced due to noise interference, which does not conform to the propagation law of seismic wave travel time. However, the first arrival picking method based on the 3DUnet network of this invention can make full use of the seismic wave information of adjacent channels, has a certain degree of noise resistance, and the picked first arrival travel time changes more smoothly and has better continuity. The picking results at far offsets are also more accurate, which verifies the effectiveness and accuracy of this invention.

[0123] This invention achieves first-arrival picking through shot gather data preprocessing for first-arrival picking and a dual-channel 3DUnet network, and can be used for first-arrival picking processing in the preprocessing stage of petroleum geophysical exploration.

[0124] In the description of this invention, it should be noted that, unless otherwise explicitly specified and limited, the terms "connected" and "linked" should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral connection; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium. Those skilled in the art can understand the specific meaning of the above terms in this invention according to the specific circumstances.

[0125] In the description of this invention, unless otherwise stated, the terms "upper," "lower," "left," "right," "inner," "outer," etc., indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings. They are used only for the convenience of describing this invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, they should not be construed as limitations on this invention.

[0126] The above technical solution is only one embodiment of the present invention. For those skilled in the art, based on the principles disclosed in the present invention, it is easy to make various types of improvements or modifications, and not limited to the technical solutions described in the specific embodiments of the present invention. Therefore, the foregoing description is only preferred and not restrictive.

Claims

1. An intelligent first-arrival pickup method, characterized in that: The method uses the energy similarity attribute of seismic data and seismic data as dual-channel data in the input layer, and uses a network model to pick the first arrival. The method includes: (1) Preprocessing of shot gather data: reading seismic data and completing seismic traces; (2) Prepare network input data: dynamic correction processing, seismic data normalization and energy similarity attribute calculation; (3) Prepare label data: interpolation of first arrival times at bad and missing paths, and mapping of first arrival times to three-dimensional data volumes; (4) Network model training: Train the 3DUnet network to obtain a trained network model; (5) First arrival picking: Input the preprocessed seismic data and use the trained network model to obtain the first arrival travel time; Step (1) includes the following operations: (11) Read earthquake data; (12) Based on the shot number and receiver line number, the data is divided into multiple three-dimensional shot gather data volumes; (13) Perform seismic trace completion processing on each three-dimensional shot gather data volume according to the maximum number of traces N set for each arrangement; Step (2) includes the following operations: (21) Using the shot-receiver distance and dynamic correction velocity, dynamic correction processing is performed on each three-dimensional shot gather data volume to obtain the dynamically corrected shot gather data; the formula for calculating the dynamic correction amount Δt is as follows: Where x is the shot-receiver distance; v0 is the dynamic correction speed; (22) On the shot gather data after dynamic correction, the seismic data between the zero time and the latest arrival time of the first arrival wave are extracted as the data to be processed, and the data to be processed is normalized to the range [-1,1] as the first channel data of the input layer; (23) Calculate the energy similarity attribute of the data to be processed as the second channel data of the input layer; obtain the energy similarity attribute using the following formula: Where X(i,j,l) is the seismic data at position (i,j,l), W is the size of the smoothing window, and Y(i,j,k) is the energy similarity attribute at position (i,j,k); Step (5) includes the following operations: (51) Process the seismic data to be processed using the same method as in steps (1) and (2) to obtain the first channel data and the second channel data; (52) Input the first channel data and the second channel data into the trained network model, and the trained network model outputs the predicted three-dimensional data volume; (53) Map the predicted 3D data volume back to the initial arrival travel time data; Step (53) includes the following operations: The following processing is performed sequentially on each seismic trace in the predicted 3D data volume: (531) The first time when the value on the seismic trace is greater than 0.5 is taken as the first arrival time; (532) The initial arrival time of the seismic trace is obtained by adding the dynamic correction amount to the initial arrival time of the picked-up first arrival time.

2. The intelligent first-arrival picking method according to claim 1, characterized in that: Step (3) includes the following operations: Read the first arrival time, and in bad and empty tracks, use the first arrival time of the nearby seismic track for interpolation. Label data is generated based on the first arrival travel time of each seismic trace: the following processing is performed on each seismic trace in sequence: data with time less than the first arrival travel time are assigned a value of 0, and data with time greater than or equal to the first arrival travel time are assigned a value of 1; after processing all seismic traces, the first arrival travel time of all seismic traces is mapped into a three-dimensional data volume that is either 0 or 1, i.e., label data.

3. The intelligent first-arrival picking method according to claim 1, characterized in that: Step (4) includes: training the 3DUnet network using the training sample label set to obtain a trained network model; the training sample label set includes the first channel data, the second channel data, and the label data.

4. An intelligent first-arrival pickup system, based on the intelligent first-arrival pickup method as described in any one of claims 1-3, characterized in that: The system includes: The preprocessing unit is used to read seismic data and perform seismic trace completion processing; The input data preparation unit, connected to the preprocessing unit, is used for dynamic correction processing, seismic data normalization, and energy similarity attribute calculation. The tag data preparation unit is connected to the preprocessing unit and the input data preparation unit respectively. It is used for first arrival time interpolation processing at bad and missing paths and mapping of first arrival time to three-dimensional data volume. The network model training unit is connected to the input data preparation unit and the label data preparation unit respectively, and is used to train the 3DUnet network to obtain the trained network model. The first arrival picking unit is connected to the preprocessing unit, the input data preparation unit, and the network model training unit, respectively. It is used to input the preprocessed seismic data and obtain the first arrival travel time using the trained network model.

5. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores at least one computer-executable program, which, when executed by the computer, causes the computer to perform the steps of the intelligent arrival picking method according to any one of claims 1-3.

Images