New energy power data processing method and device

Abstract

The application is suitable for the technical field of data processing, and provides a new energy power data processing method and device. The method comprises the following steps: acquiring new energy power data; extracting first keywords from each text of the power data and second keywords from each image, converting the first keywords of each text and the second keywords of each image into corresponding preset format files and feature vectors respectively; calculating the distance between the feature vectors of each text and the feature vectors of the images, and determining the associated text and image according to the distance; fusing the feature vectors of the associated text and image to obtain a fused feature vector, and processing the new energy power data based on the preset format file and the fused feature vector. The application can associate the text and image in the new energy power data, reduce the complexity of new energy power data management, and reduce the time required for processing the new energy power data.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of data processing, and particularly relates to a new energy power data processing method and device. BACKGROUND

[0002] New energy refers to the use of resources that are constantly updated in nature, such as solar energy, wind energy, and clean and renewable energy generated through biomass energy, geothermal energy, etc. Compared with traditional thermal power generation, new energy power generation has the advantages of environmental protection and sustainability, and is an important direction for future power development.

[0003] New energy power generation involves many system platforms, and the data information generated by each system platform is inconsistent in format. Moreover, the data information between different system platforms is isolated from each other, and the data between different system platforms cannot be accessed by each other, which leads to difficulties in integrating new energy data and cannot effectively support management and decision-making. Therefore, how to realize the data integration of new energy power to improve the efficiency of management and decision-making has become a problem to be solved. SUMMARY

[0004] Therefore, the embodiments of the present application provide a new energy power data processing method and device, which can

[0005] To achieve the above purpose, the technical scheme is as follows:

[0006] In a first aspect, the embodiments of the present application provide a new energy power data processing method, comprising:

[0007] Obtaining new energy power data, the power data comprising a plurality of texts and a plurality of images, the power data being derived from a plurality of power system platforms;

[0008] Extracting first keywords from each text, converting the first keywords extracted from each text into a first preset format file, and converting the first keywords of each text into a first feature vector;

[0009] Extracting second keywords from each image, converting the second keywords extracted from each image into a second preset format file, and converting the second keywords of each image into a second feature vector;

[0010] Calculating the distance between each first feature vector and each second feature vector, and determining the associated text and image according to the distance;

[0011] Fusing the first feature vector and the second feature vector corresponding to the associated text and image to obtain a fusion feature vector, and processing the new energy power data based on the first preset format file, the second preset format file and the fusion feature vector.

[0012] In some embodiments based on the first aspect, the extracting the second keyword from each image comprises:

[0013] performing filtering and noise reduction on the image;

[0014] inputting the image sample containing the label information into the image segmentation model to train the image segmentation model; wherein the label information comprises an image category and a segmentation position of an image region;

[0015] segmenting the image subjected to the filtering and noise reduction using the trained image segmentation model to obtain at least one image region;

[0016] identifying text from the image region and extracting the second keyword from the identified text.

[0017] In some embodiments based on the first aspect, the image category is multiple, and each category of image corresponds to a segmentation position of an image region, the image region containing text or a picture, and the segmentation position of the image region containing text is different from the segmentation position of the image region containing the picture.

[0018] In some embodiments based on the first aspect, the image category is determined based on text in a preset region in the image.

[0019] The segmenting the image subjected to the filtering and noise reduction using the trained image segmentation model to obtain at least one image region comprises:

[0020] inputting the image subjected to the filtering and noise reduction into the trained image segmentation model;

[0021] the image segmentation model extracts text from a preset region in the image, determines the image category according to the text, and calls the segmentation position of the image region corresponding to the image category;

[0022] the image segmentation model segments the image based on the segmentation position of the image region to obtain at least one image region.

[0023] In some embodiments based on the first aspect, the extracting the first keyword from each text comprises:

[0024] performing word segmentation on the text to determine a first frequency and a position of each word in the text;

[0025] determining a first span of each word in the text according to a distance between a first position and a second position of the word in the text; wherein the first position is a position where the word first appears in the text, the second position is a position where the word last appears in the text, and the distance between the first position and the second position is a number of words between the first position and the second position;

[0026] determine a target word with the first frequency greater than the threshold value, and determine a first preset number of words with the largest first span in the target word as candidate keywords of the text, the number of the candidate keywords being the first preset number;

[0027] determine a second frequency and a second span of each candidate keyword in other texts;

[0028] calculate a product of the first frequency and the first span, and a product of the second frequency and the second span, and determine a first keyword of the text as a candidate keyword with the largest difference value of the products, the number of the first keyword being a second preset number, the second preset number being less than the first preset number.

[0029] In some embodiments based on the first aspect, the calculating the distance between each first feature vector and each second feature vector, and determining the associated text and image according to the distance comprises:

[0030] for any one first feature vector, calculating a cosine similarity between the first feature vector and each second feature vector;

[0031] associating the first feature vector with the second feature vector with the smallest cosine similarity, to obtain the associated text and image.

[0032] In some embodiments based on the first aspect, the converting the first keyword of each text into a first feature vector comprises: the first keyword is a plurality, each row of elements in the first feature vector corresponds to a first keyword, and the value of the element corresponding to each first keyword is calculated according to the frequency of the first keyword in the target text multiplied by the span and a preset weight, to obtain the first feature vector;

[0033] the converting the second keyword of each image into a second feature vector comprises: the second keyword is a plurality, each row of elements in the second feature vector corresponds to a second keyword, and the value of the element corresponding to each second keyword is calculated according to the frequency of the second keyword in the target image multiplied by the span and a preset weight, to obtain the second feature vector.

[0034] In some embodiments based on the first aspect, the fusing the first feature vector and the second feature vector corresponding to the associated text and image to obtain a fused feature vector comprises:

[0035] calculating the sum of two elements at the same position of the first feature vector and the second feature vector, the two elements respectively corresponding to two weight values, and the weight value of the element of the first feature vector being greater than the weight value of the element of the second feature vector;

[0036] taking the sum of the two elements as the value of the element at the corresponding position of the fused feature vector.

[0037] The sum of the two elements of each position is calculated in sequence to obtain the value of the element of each position of the fusion feature vector.

[0038] Based on the first aspect, in some embodiments, the processing of the new energy power data based on the first preset format file, the second preset format file and the fusion feature vector comprises:

[0039] The first preset format file, the second preset format file and the fusion feature vector are associated, and preset feature identification information is set, wherein the preset feature identification information is set based on a preset rule;

[0040] A new energy power data query request input by a user is obtained, and the new energy power data query request is packaged with feature identification;

[0041] In response to the new energy power data query request, target preset feature identification information consistent with the feature identification is determined;

[0042] The first preset format file, the second preset format file and the fusion feature vector corresponding to the target preset feature identification information are displayed to the user.

[0043] The second aspect, the embodiment of the application provides a new energy power data processing device, comprising:

[0044] Compared with the prior art, the embodiment of the application has the beneficial effects including:

[0045] In the embodiment of the application, the new energy power data of a plurality of power systems can be obtained in advance, then the keywords are extracted from the new energy power data and converted into feature vectors, and the new energy power data is converted into a preset format file, then the associated text and image in the new energy power data are determined according to the distance between the feature vectors, finally the feature vectors of the associated image and text are fused, and the new energy power data is processed according to the fusion feature vector and the preset format file. When it is necessary to process the new energy power data of some sites or some businesses, the corresponding new energy power data can be called to process, so that the text and image in the new energy power data can be associated, the complexity of new energy power data management is reduced, and the new energy power data can be processed in advance, so that the time required for processing the new energy power data is reduced. BRIEF DESCRIPTION OF DRAWINGS

[0046] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings needed to be used in the embodiments or prior art description will be briefly introduced. Obviously, the drawings in the following description are only some embodiments of the present application, and other drawings can be obtained by those skilled in the art without any creative labor.

[0047] Figure 1 is a flowchart of a new energy power data processing method provided by the embodiments of the present application;

[0048] Figure 2 is a structural schematic diagram of a new energy power data processing device provided by the embodiments of the present application;

[0049] Figure 3 is a structural schematic diagram of a terminal device provided by the embodiments of the present application. DETAILED DESCRIPTION

[0050] The present application will be described more clearly in combination with specific embodiments. The following embodiments will help those skilled in the art to further understand the role of the present application, but do not limit the present application in any form. It should be pointed out that, for those skilled in the art, without departing from the concept of the present application, a number of modifications and improvements can be made. These all belong to the protection scope of the present application.

[0051] In order to make the purpose, technical solutions and advantages of the present application more clear, the following will be described by specific embodiments in combination with the drawings.

[0052] Referring to Figure 1 The new energy power data processing method provided by the embodiments of the present application can include the following steps:

[0053] Step 101, acquiring new energy power data.

[0054] Among them, the power data includes a plurality of texts and a plurality of images, and the power data comes from a plurality of power system platforms.

[0055] In some embodiments, step 101 can specifically include: sending a new energy power data acquisition request to each system platform associated with the new energy power business, the request containing screening information such as business identification or site identification; receiving new energy power data sent by each system platform associated with the new energy power business, the new energy power data being obtained by each system platform in response to the new energy power data acquisition request to retrieve power data corresponding to the screening information. The screening information can be one or more. For the case of multiple business identifications or site identifications, after obtaining the new energy power data, the screening information of the new energy power data is determined, and each new energy power data containing the same screening information is combined into a new energy power data set.

[0056] In yet other embodiments, step 101 can specifically include: sending a new energy power data acquisition request to each system platform associated with the new energy power business, the request containing a time period and business identification or site identification; receiving new energy power data sent by each system platform associated with the new energy power business, the new energy power data being obtained by each system platform in response to the new energy power data acquisition request to retrieve new energy power data within the time period; extracting the business identification or site identification in each new energy power data; for the case of multiple business identifications or site identifications, combining each new energy power data containing the same business identification or site identification into a new energy power data set to obtain multiple new energy power data sets.

[0057] In the embodiments of the present application, the terminal device can send a new energy power data acquisition request to each system platform, the request containing business identification or site identification, at this time each system platform can send new energy power data corresponding to the business identification or site identification to the terminal device. Alternatively, the terminal device can send a new energy power data acquisition request to each system platform, the new energy power data acquisition request containing a time period, the new energy power data in each system platform not directly containing business identification or site identification, the key words of the business identification or site identification in each new energy power data can be extracted first, the business identification or site identification of each new energy power data is determined according to the key words of the business identification or site identification, and each system platform then sends the new energy power data within the time period to the terminal device.

[0058] For example, the embodiment of the present application can send new energy power data in certain time period to the terminal device in advance, or send new energy power data of certain service to the terminal device, or send new energy power data of certain site to the terminal device, and the terminal device processes the new energy power data to realize subsequent processing of the new energy power data. When the new energy power data of a certain time period or certain service or certain site needs to be processed, the new energy power data obtained by subsequent processing can be called to process.

[0059] Step 102, extracting first keywords from each text, converting the first keywords extracted from each text into a first preset format file, and converting the first keywords of each text into a first feature vector.

[0060] In some embodiments, the process of extracting first keywords from each text can include: segmenting the text, determining the first frequency and position of each word in the text; determining the first span of each word in the text according to the distance between the first position and the second position of each word in the text; wherein the first position is the position of the first appearance in the text, the second position is the position of the last appearance in the text, and the distance between the first position and the second position is the number of words between the first position and the second position; determining the target word whose first frequency is greater than the threshold value, and taking the first preset number of words with the largest first span in the target word as the candidate keywords of the text, the number of candidate keywords being the first preset number; determining the second frequency and the second span of each candidate keyword in other texts; calculating the product of the first frequency and the first span, and the product of the second frequency and the second span, and taking the candidate keyword with the largest product difference as the first keyword of the text, the number of first keywords being the second preset number, the second preset number being less than the first preset number.

[0061] For example, the first frequency is the number of times a word appears in a text, the second frequency is the number of times a candidate keyword appears in other texts, and the span is the number of words between the first position and the second position.

[0062] Step 103, extracting second keywords from each image, converting the second keywords extracted from each image into a second preset format file, and converting the second keywords of each image into a second feature vector.

[0063] In some embodiments, the process of extracting the second keyword from each of the images can include: performing filtering and noise reduction on the images; inputting image samples containing label information into an image segmentation model, and training the image segmentation model; wherein the label information contains an image category and a segmentation position of an image region; segmenting the images that have undergone filtering and noise reduction using the trained image segmentation model to obtain at least one image region; identifying text from the image region, and extracting the second keyword from the identified text.

[0064] For example, the filtering and noise reduction on the images can be implemented based on a convolutional neural network. Specifically, to train a filtering and noise reduction model with excellent performance, a large number of noisy images need to be prepared as a training set. At the same time, corresponding noise-free images need to be prepared as labels. The image denoising model based on the convolutional neural network is composed of multiple convolutional layers, pooling layers, and fully connected layers. Among them, the convolutional layer is responsible for learning features from the input image, the pooling layer is responsible for reducing the feature dimension, and the fully connected layer is responsible for mapping the learned features to the labels. During the training process, the input image is convolved by the filter learned by the convolutional layer to obtain a denoised image; by comparing the difference between the denoised image and the label, the loss function is calculated and the filter parameters are updated by back propagation; the above process is repeated until the model performance meets the expected requirements.

[0065] In this embodiment, the image category can be multiple, and each category of image corresponds to a segmentation position of an image region, which contains text or pictures. The segmentation position of the image region containing text is different from that of the image region containing pictures.

[0066] For example, the image category can be determined according to the site, image content, etc. For example, the image categories of the images of new energy power data generated by different sites are different; or the image categories are different corresponding to different image contents in the images. The image content can be the model of the equipment in the image, or the heading title of the table in the image, etc., which is not limited.

[0067] In some embodiments, the image category is determined based on the text in the preset region of the image, for example, based on the heading title text of the table in the image. The segmentation of the image that has undergone filtering and noise reduction using the trained image segmentation model to obtain at least one image region can include: inputting the image that has undergone filtering and noise reduction into the trained image segmentation model; the image segmentation model extracts text from the preset region of the image, determines the image category according to the text, and calls the segmentation position of the image region corresponding to the image category; the image segmentation model segments the image based on the segmentation position of the image region to obtain at least one image region.

[0068] In this embodiment, for a specific image category, the position of the contained text in the image is fixed, so the segmentation position of the image region can be set in advance according to the image category. Specifically, the segmentation position of the image region can be manually set by a person, or can be learned by a neural network model to set the segmentation position of the image region.

[0069] In this step, the image can be an image of a certain format file. For the image of the format file, the image region where the required content is located can be extracted from the image first, and then keyword extraction is performed in the extracted image region.

[0070] Specifically, the implementation process of the above "identifying text from the image region, and extracting a second keyword from the identified text" can include: converting the image region into a to-be-recognized text image, performing connected region extraction on the to-be-recognized text image to obtain a plurality of connected regions; determining the contour of each character in each connected region, and based on the contour, segmenting to obtain a plurality of characters and recording the positional relationship of each character in the connected region; recognizing each character, and generating a to-be-extracted text according to the positional relationship of each character in the connected region; and extracting a second keyword from the to-be-extracted text. Wherein, converting the image region into a to-be-recognized text image can be converting the image region into a gray image to obtain the to-be-recognized text image.

[0071] For example, a format file usually contains a file name and file specific content, so in this step, the second keyword can be extracted from the text corresponding to the file name part, or the second keyword can be extracted from the text corresponding to the file name and the file specific content. The process of extracting the second keyword can refer to the related content in step 102, which will not be repeated here.

[0072] For example, performing connected region extraction on the to-be-recognized text image to obtain a plurality of connected regions can include: determining the gray value of each pixel point in the to-be-recognized text image, and determining the pixel group in each row of pixel points in the to-be-recognized text image according to the gray value of each pixel point, wherein each pixel point in the pixel group is connected and the gray value is a preset gray value; determining the pixel group connected between different rows of pixel points according to the position of each pixel group to obtain a connected region; wherein the pixel groups of adjacent rows in the same connected region have pixel points located in the same column.

[0073] Specifically, in the identified text image, the gray values of the pixel points corresponding to the text characters are the same. Therefore, the gray values of the pixel points in each row of pixel points of the to-be-identified text image can be compared with the preset gray value, the pixel points with the same gray value as the preset gray value are found out, and the pixel points of the text characters are found out. If these pixel points are connected, a pixel group is formed. Thus, the pixel groups in each row of pixel points are found out, and the connected pixel groups form a plurality of connected regions, i.e., the regions where the text characters are located.

[0074] In step 104, the distance between each first feature vector and each second feature vector is calculated, and the associated text and image are determined according to the distance.

[0075] In some embodiments, the implementation process of step 104 can include: for any one first feature vector, calculating the cosine similarity between the first feature vector and each second feature vector; associating the first feature vector with the second feature vector with the smallest cosine similarity, to obtain the associated text and image.

[0076] The cosine similarity is the cosine value of the included angle between two vectors, and is used to measure the consistency of the directions of the two vectors. The greater the cosine similarity, the smaller the included angle between the two vectors, and the higher the direction consistency. The smaller the cosine similarity, the greater the included angle between the two vectors, and the lower the direction consistency.

[0077] In some embodiments, the process of converting each text first keyword into a first feature vector can include: the first keyword is multiple, each row of elements in the first feature vector corresponds to a first keyword, and the value of the element corresponding to each first keyword is calculated according to the frequency of each first keyword appearing in the target text multiplied by the span and the preset weight, to obtain the first feature vector.

[0078] The process of converting each image second keyword into a second feature vector can include: the second keyword is multiple, each row of elements in the second feature vector corresponds to a second keyword, and the value of the element corresponding to each second keyword is calculated according to the frequency of each second keyword appearing in the target image multiplied by the span and the preset weight, to obtain the second feature vector.

[0079] In step 105, the first feature vector and the second feature vector corresponding to the associated text and image are fused to obtain a fusion feature vector, and new energy power data is processed based on the first preset format file, the second preset format file, and the fusion feature vector.

[0080] In some embodiments, the fusing the first feature vector and the second feature vector corresponding to the associated text and image to obtain a fused feature vector can include: calculating a sum of two elements at the same position of the first feature vector and the second feature vector, the two elements corresponding to two weight values respectively, and the weight value of the element of the first feature vector being greater than the weight value of the element of the second feature vector; taking the sum of the two elements as the value of the element at the corresponding position of the fused feature vector; and sequentially calculating the sum of the two elements at each position to obtain the value of the element at each position of the fused feature vector.

[0081] Specifically, the two elements at the same position of the first feature vector and the second feature vector are multiplied by the respective corresponding weight values, and then the two products are added to obtain the value of the element at the corresponding position of the fused feature vector.

[0082] In addition, the values of the elements at the corresponding positions of the first feature vector and the second feature vector can also be synthesized into a new value. For example, the value of an element of the first feature vector is ab, and the value of the element at the corresponding position of the second feature vector is cd, and the synthesized new value is abcd, where a, b, c and d are numerical values.

[0083] In some embodiments, the processing the new energy power data based on the first preset format file, the second preset format file and the fused feature vector can include: associating the first preset format file, the second preset format file and the fused feature vector, and setting preset feature identification information; wherein the preset feature identification information is set based on a preset rule; obtaining a new energy power data query request input by a user, the new energy power data query request including feature identification; in response to the new energy power data query request, determining target preset feature identification information consistent with the feature identification; and displaying the first preset format file, the second preset format file and the fused feature vector corresponding to the target preset feature identification information to the user.

[0084] For example, the first keywords extracted from each text are used to generate a first preset format file, so that each text is converted into a first preset format file with unified format, which facilitates subsequent processing, searching and displaying. Similarly, the second keywords extracted from each image are used to generate a second preset format file, so that each image is converted into a second preset format file with unified format, which facilitates subsequent processing, searching and displaying.

[0085] In this embodiment, after obtaining the fusion feature vector of the associated text and image, the first preset format file of the associated text and the second preset format file of the image are associated with the fusion feature vector, and then preset feature identification information is set. The preset feature identification information can be a feature identification for facilitating the user to search the new energy power data, for example, a feature identification containing site information or a feature identification containing business information.

[0086] When the user needs to search the new energy power data related to a certain site or a certain business, the user terminal inputs a new energy power data query request containing a feature identification. The terminal device can respond to the new energy power data query request and determine the target preset feature identification information consistent with the feature identification in the request. Then, the first preset format file, the second preset format file and the fusion feature vector corresponding to the target preset feature identification information are sent to the user terminal of the user for display.

[0087] In addition, the user can apply a preset operation on the display interface of the user terminal. The user terminal can send an original data viewing request to the terminal device. The terminal device responds to the original data viewing request and sends the text corresponding to the first preset format file and the image corresponding to the second preset format file to the user terminal for the user to view.

[0088] Further, the above new energy power data processing method can further include: setting access permissions for each group of associated first preset format files, second preset format files and fusion feature vectors based on preset feature identification information. Specifically, the MAC (Media Access Control Address) address of the user terminal with access permission corresponding to each preset feature identification information and the MAC address of the user terminal without access permission can be set. When the user wants to access the new energy power data related to a certain site or a certain business through the user terminal, the user terminal inputs a new energy power data query request containing a feature identification. The terminal device responds to the new energy power data query request and determines the target preset feature identification information consistent with the feature identification. Then, the terminal device determines whether the user terminal has the permission to access the new energy power data corresponding to the target preset feature identification information according to the MAC address of the user terminal. If the user terminal has the permission, the first preset format file, the second preset format file and the fusion feature vector corresponding to the target preset feature identification information are sent to the user terminal for display; otherwise, the first preset format file, the second preset format file and the fusion feature vector corresponding to the target preset feature identification information are not sent to the user terminal.

[0089] The new energy power data processing method can pre-acquire new energy power data of multiple power systems, then extract keywords from the new energy power data and convert the keywords into feature vectors, convert the new energy power data into a preset format file, then determine associated text and images in the new energy power data according to distances between the feature vectors, and finally fuse feature vectors of the associated images and text, and process the new energy power data according to the fused feature vectors and the preset format file. When it is necessary to process new energy power data of some sites or some businesses, the corresponding new energy power data can be called to process, so that the text and images in the new energy power data can be associated, the complexity of new energy power data management is reduced, and the new energy power data can be pre-processed, so that the time required for processing the new energy power data is reduced.

[0090] It should be understood that the size of the serial number of each step in the above embodiment does not mean the order of execution, and the execution order of each process should be determined according to its function and inherent logic, and should not constitute any limitation on the implementation process of the embodiments of the present application.

[0091] According to the new energy power data processing method described in the above embodiment, Figure 2 In addition to the structure block diagram of the new energy power data processing device provided by the embodiments of the present application, only the parts related to the embodiments of the present application are shown for the convenience of description.

[0092] Referring to Figure 2 The embodiments of the present application provide a new energy power data processing device, which comprises a data acquisition module 201, a first vector conversion module 202, a second vector conversion module 203, an association module 204 and a fusion module 205.

[0093] Specifically, the data acquisition module 201 is configured to acquire new energy power data, wherein the power data comprises multiple texts and multiple images, and the power data is from multiple power system platforms.

[0094] The first vector conversion module 202 is configured to extract first keywords from each text, convert the first keywords extracted from each text into a first preset format file, and convert the first keywords of each text into a first feature vector.

[0095] The second vector conversion module 203 is configured to extract second keywords from each image, convert the second keywords extracted from each image into a second preset format file, and convert the second keywords of each image into a second feature vector.

[0096] The association module 204 is configured to calculate distances between each first feature vector and each second feature vector, and determine associated text and images according to the distances.

[0097] The fusion module 205 is configured to fuse the first feature vector and the second feature vector corresponding to the associated text and image to obtain a fusion feature vector, and process new energy power data based on the first preset format file, the second preset format file and the fusion feature vector.

[0098] Optionally, the second vector conversion module 203 is specifically configured to: perform filtering and noise reduction processing on the image; input an image sample containing label information into an image segmentation model, and train the image segmentation model; wherein the label information contains an image category and a segmentation position of an image region; segment the image subjected to the filtering and noise reduction processing by using the trained image segmentation model to obtain at least one image region; identify text from the image region, and extract a second keyword from the identified text.

[0099] Optionally, the image category can be multiple, and the image of each category corresponds to a segmentation position of an image region, the image region contains text or a picture, and the segmentation position of the image region containing the text is different from the segmentation position of the image region containing the picture.

[0100] Optionally, the image category is determined based on text in a preset region in the image; and the segmentation of the image subjected to the filtering and noise reduction processing by using the trained image segmentation model to obtain at least one image region can include: inputting the image subjected to the filtering and noise reduction processing into the trained image segmentation model; extracting text from the preset region in the image by using the image segmentation model, determining an image category according to the text, and calling a segmentation position of an image region corresponding to the image category; segmenting the image based on the segmentation position of the image region by using the image segmentation model to obtain at least one image region.

[0101] Optionally, the first vector conversion module 202 can be specifically configured to: perform word segmentation on the text to determine a first frequency and a position of each word in the text; determine a first span of each word in the text according to a distance between a first position and a second position of each word in the text; wherein the first position is a position at which the word first appears in the text, the second position is a position at which the word last appears in the text, and the distance between the first position and the second position is a number of words between the first position and the second position; determine a target word with a first frequency greater than a threshold value, and take a first preset number of words with the largest first span in the target word as candidate keywords of the text, a number of the candidate keywords being the first preset number; determine a second frequency and a second span of each candidate keyword in other texts; calculate a product of the first frequency and the first span, and a product of the second frequency and the second span, and take a candidate keyword with the largest product difference as a first keyword of the text, a number of the first keyword being a second preset number, the second preset number being less than the first preset number.

[0102] Optionally, the association module 204 can be specifically configured to: for any one first feature vector, calculate the cosine similarity between the first feature vector and each second feature vector; and associate the first feature vector with the second feature vector having the minimum cosine similarity, to obtain the associated text and image.

[0103] Optionally, the process of converting the first keyword of each text into the first feature vector by the first vector conversion module 202 can include: the first keyword is multiple, each row element in the first feature vector corresponds to a first keyword, and the value of the element corresponding to each first keyword is calculated according to the frequency of the first keyword appearing in the target text multiplied by the span and the preset weight, to obtain the first feature vector.

[0104] The process of converting the second keyword of each image into the second feature vector by the second vector conversion module 203 can include: the second keyword is multiple, each row element in the second feature vector corresponds to a second keyword, and the value of the element corresponding to each second keyword is calculated according to the frequency of the second keyword appearing in the target image multiplied by the span and the preset weight, to obtain the second feature vector.

[0105] Optionally, the fusion module 205 can be specifically configured to: calculate the sum of two elements at the same position of the first feature vector and the second feature vector, the two elements respectively corresponding to two weight values, and the weight value of the element of the first feature vector being greater than the weight value of the element of the second feature vector; take the sum of the two elements as the value of the element at the corresponding position of the fusion feature vector; and sequentially calculate the sum of the two elements at each position to obtain the value of the element at each position of the fusion feature vector.

[0106] Optionally, the fusion module 205 can be specifically configured to: associate the first preset format file, the second preset format file and the fusion feature vector, and set preset feature identification information; wherein the preset feature identification information is set based on a preset rule; obtain a new energy power data query request input by a user, the new energy power data query request being packaged with feature identification; in response to the new energy power data query request, determine target preset feature identification information consistent with the feature identification; and display the first preset format file, the second preset format file and the fusion feature vector corresponding to the target preset feature identification information to the user.

[0107] Figure 3 is a schematic diagram of a terminal device provided by an embodiment of the present application. Figure 3As shown, the terminal device 300 of this embodiment includes a processor 310 and a memory 320, and the memory 320 stores a computer program, for example, a new energy power data processing program, which can be run on the processor 310. The processor 310 implements the steps in the above new energy power data processing method embodiments when executing the computer program, for example Figure 1 As shown, the processor 310 implements the functions of the modules / units in the above device embodiments when executing the computer program, for example Figure 2 As shown, the processor 310 implements the functions of the data acquisition module 201 to the fusion module 205.

[0108] The terminal device 300 can be a desktop computer, a notebook computer, a palm computer, a cloud terminal device, and the like. The terminal device can include, but is not limited to, the processor 310 and the memory 320. Those skilled in the art can understand that Figure 3 The terminal device 300 is only an example and does not constitute a limitation on the terminal device 300, and can include more or fewer components than shown, or combine certain components, or different components, for example, the terminal device can also include an input / output device, a network access device, a bus, and the like.

[0109] The processor 320 can be a central processing unit (CPU), and can also be other general-purpose processors, digital signal processors (DSP), application specific integrated circuits (ASIC), field-programmable gate arrays (FPGA) or other programmable logic devices, discrete gates or transistor logic components, discrete hardware components, or the like. The general-purpose processor can be a microprocessor or the processor can also be any conventional processor.

[0110] The memory 320 can be an internal storage unit of the terminal device 300, for example, a hard disk or a memory of the terminal device 300. The memory 320 can also be an external storage device of the terminal device 300, for example, a plug-in hard disk, a smart media card (SMC), a secure digital (SD) card, a flash card, etc. Further, the memory 320 can also include both the internal storage unit and the external storage device of the terminal device 300. The memory 320 is used to store the computer program and other programs and data required by the terminal device. The memory 320 can also be used to temporarily store data that has been output or will be output.

[0111] In the above embodiments, the description of each embodiment has its own focus, and the parts not described or recorded in a certain embodiment can be referred to the relevant description of other embodiments.

[0112] The above-described embodiments are only used to illustrate the technical solutions of the present application, rather than limit them; although the present application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that they can still modify the technical solutions recorded in the foregoing embodiments, or make equivalent replacements for some technical features; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present application, and should be included in the protection scope of the present application.

Claims

1. A new energy power data processing method, characterized in that, The method comprises the following steps: acquiring new energy power data, the power data comprising a plurality of texts and a plurality of images, the power data being derived from a plurality of power system platforms; extracting first keywords from each text, converting the first keywords extracted from each text into a first preset format file, and converting the first keywords of each text into a first feature vector; extracting second keywords from each image, converting the second keywords extracted from each image into a second preset format file, and converting the second keywords of each image into a second feature vector; calculating the distance between each first feature vector and each second feature vector, and determining the associated text and image according to the distance; fusing the first feature vector and the second feature vector corresponding to the associated text and image to obtain a fused feature vector, and processing the new energy power data based on the first preset format file, the second preset format file, and the fused feature vector; the step of extracting second keywords from each image comprises: performing filtering and noise reduction processing on the image; inputting an image sample containing label information into an image segmentation model, and training the image segmentation model; wherein the label information contains image categories and segmentation positions of image regions; segmenting the image subjected to the filtering and noise reduction processing by using the trained image segmentation model to obtain at least one image region; identifying text from the image region, and extracting second keywords from the identified text; the image categories are a plurality of categories, each category of image corresponds to a segmentation position of an image region, the image region contains text or a picture, the segmentation position of the image region containing the text is different from the segmentation position of the image region containing the picture; the image categories are determined according to sites and image contents, the image categories of the images of the new energy power data generated by different sites are different, the image categories corresponding to different image contents in the images are different, and the image contents are the models of devices in the images or the headline titles of tables in the images; the step of identifying text from the image region and extracting second keywords from the identified text comprises: converting the image region into a to-be-recognized text image, performing connected region extraction on the to-be-recognized text image to obtain a plurality of connected regions; determining the contour of each character in each connected region, and segmenting and recording the position relationship of each character in the connected region based on the contour; recognizing each character, and generating a to-be-extracted text according to the position relationship of each character in the connected region; and extracting second keywords from the to-be-extracted text; wherein the conversion of the image region into the to-be-recognized text image is the conversion of the image region into a gray image to obtain the to-be-recognized text image. The connected region extraction on the text image to be recognized comprises: determining the gray value of each pixel point in the text image to be recognized, determining the pixel cluster in each row of pixel points in the text image to be recognized according to the gray value of each pixel point, wherein each pixel point in the pixel cluster is connected and the gray value of each pixel point is a preset gray value; determining the pixel cluster connected between different rows of pixel points according to the position of each pixel cluster to obtain the connected region; wherein the pixel clusters of adjacent rows in the same connected region are connected through the pixel points in the same column; The first keyword extraction from each text comprises: performing word segmentation on the text to determine the first frequency and position of each word in the text; determining the first span of each word in the text according to the distance between the first position and the second position of each word in the text; wherein the first position is the position of the first occurrence in the text, the second position is the position of the last occurrence in the text, and the distance between the first position and the second position is the number of words between the first position and the second position; determining a target word with a first frequency greater than a threshold value, and taking the first preset number of words with the largest first span in the target word as candidate keywords of the text, wherein the number of candidate keywords is the first preset number; determining the second frequency and second span of each candidate keyword in other texts; calculating the product of the first frequency and the first span, and the product of the second frequency and the second span, and taking the candidate keyword with the largest product difference as the first keyword of the text, wherein the number of first keywords is the second preset number, and the second preset number is less than the first preset number.

2. The new energy power data processing method according to claim 1, characterized in that, The image category is determined based on the text in the preset region of the image. The image segmentation model is trained based on the image and the text in the preset region of the image. The image segmentation model is trained based on the image and the text in the preset region of the image. The image segmentation model is trained based on the image and the text in the preset region of the image. The distance between each first feature vector and each second feature vector is calculated, and the associated text and image are determined according to the distance.

3. The new energy power data processing method according to any one of claims 1-2, characterized in that, For any first feature vector, the cosine similarity between the first feature vector and each second feature vector is calculated. The first feature vector is associated with the second feature vector with the smallest cosine similarity to obtain the associated text and image. The first keyword of each text is converted into a first feature vector, wherein the first keyword is a plurality of first keywords, each row of elements in the first feature vector corresponds to a first keyword, the value of the element corresponding to each first keyword is calculated according to the frequency, span and preset weight of each first keyword in the target text, and the first feature vector is obtained.

4. The new energy power data processing method according to claim 3, characterized in that, ​ The converting of the second keywords of each image into a second feature vector comprises: the second keywords are multiple, each row element in the second feature vector corresponds to a second keyword, and the value of the element corresponding to each second keyword is calculated according to the frequency of the appearance of each second keyword in the target image, the span and a preset weight, to obtain the second feature vector.

5. The new energy power data processing method according to claim 4, characterized in that, The fusing of the first feature vector and the second feature vector corresponding to the associated text and image comprises: calculating the sum of two elements at the same position of the first feature vector and the second feature vector, the two elements corresponding to two weight values respectively, and the weight value of the element of the first feature vector being greater than the weight value of the element of the second feature vector; taking the sum of the two elements as the value of the element at the corresponding position of the fusion feature vector; sequentially calculating the sum of the two elements at each position to obtain the value of the element at each position of the fusion feature vector.

6. The new energy power data processing method according to claim 5, characterized in that, The processing of new energy power data based on the first preset format file, the second preset format file and the fusion feature vector comprises: associating the first preset format file, the second preset format file and the fusion feature vector, and setting preset feature identification information; wherein the preset feature identification information is set based on a preset rule; obtaining a new energy power data query request input by a user, the new energy power data query request being packaged with feature identification; in response to the new energy power data query request, determining target preset feature identification information consistent with the feature identification; sending the first preset format file, the second preset format file and the fusion feature vector corresponding to the target preset feature identification information to the user.

7. A new energy electric power data processing device, characterized in that, comprise: a data acquisition module configured to acquire new energy power data, the power data comprising multiple texts and multiple images, and the power data being derived from multiple power system platforms; a first vector conversion module configured to extract first keywords from each text, convert the first keywords extracted from each text into a first preset format file, and convert the first keywords of each text into a first feature vector; a second vector conversion module configured to extract second keywords from each image, convert the second keywords extracted from each image into a second preset format file, and convert the second keywords of each image into a second feature vector; an association module configured to calculate the distance between each first feature vector and each second feature vector, and determine the associated text and image according to the distance; a fusion module configured to fuse the first feature vector and the second feature vector corresponding to the associated text and image to obtain a fusion feature vector, and process new energy power data based on the first preset format file, the second preset format file and the fusion feature vector. The step of extracting the second keyword from each image includes: filtering and denoising the images; inputting image samples containing labeled information into an image segmentation model and training the image segmentation model; wherein, the labeled information includes image category and segmentation position of image region; using the trained image segmentation model to segment the filtered and denoised images to obtain at least one image region; identifying text from the image region and extracting the second keyword from the identified text; there are multiple image categories, and each category of images corresponds to a segmentation position of an image region, wherein the image region contains text or images, and the segmentation position of the image region containing text is different from the segmentation position of the image region containing images; the image category is determined according to the station and image content, and the image categories of images of new energy power data generated by different stations are different, and the image categories corresponding to different image content in the image are different, wherein the image content is the model of the equipment in the image, or the header title of the table in the image; The step of identifying text from an image region and extracting a second keyword from the identified text includes: converting the image region into a text image to be identified; performing connected component extraction on the text image to be identified to obtain multiple connected components; determining the contour of each character in each connected component; segmenting multiple characters based on the contour and recording the positional relationship of each character in the connected components; identifying each character and generating text to be extracted based on the positional relationship of each character in the connected components; and extracting the second keyword from the text to be extracted. Specifically, converting the image region into a text image to be identified involves converting the image region into a grayscale image to obtain the text image to be identified. The process of extracting connected components from the text image to be recognized to obtain multiple connected components includes: determining the grayscale value of each pixel in the text image to be recognized; determining pixel clusters in each row of pixels in the text image to be recognized based on the grayscale value of each pixel, wherein each pixel in the pixel cluster is connected and the grayscale value is a preset grayscale value; determining pixel clusters that are interconnected between pixels in different rows based on the position of each pixel cluster, thereby obtaining connected components; wherein, in the same connected component, there are pixels in the same column between pixel clusters in adjacent rows. The first keyword extraction from each text comprises: segmenting the text, determining the first frequency and position of each word in the text; determining the first span of each word in the text according to the distance between the first position and the second position of each word in the text; wherein the first position is the first occurrence position in the text, the second position is the last occurrence position in the text, and the distance between the first position and the second position is the number of words between the first position and the second position; determining the target word with the first frequency greater than the threshold value, and taking the first preset number of words with the maximum first span in the target word as the candidate keywords of the text, the number of candidate keywords being the first preset number; determining the second frequency and the second span of each candidate keyword in other texts; calculating the product of the first frequency and the first span, and the product of the second frequency and the second span, and taking the candidate keyword with the maximum product difference as the first keyword of the text, the number of first keywords being the second preset number, and the second preset number being less than the first preset number.

Images