CN118069331B - Intelligent acquisition task scheduling method and device based on digital twinning - Google Patents
Intelligent acquisition task scheduling method and device based on digital twinning Download PDFInfo
- Publication number
- CN118069331B CN118069331B CN202410494247.7A CN202410494247A CN118069331B CN 118069331 B CN118069331 B CN 118069331B CN 202410494247 A CN202410494247 A CN 202410494247A CN 118069331 B CN118069331 B CN 118069331B
- Authority
- CN
- China
- Prior art keywords
- data acquisition
- semantic
- feature vectors
- semantic coding
- coding feature
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
- 238000000034 method Methods 0.000 title claims abstract description 39
- 238000005457 optimization Methods 0.000 claims abstract description 8
- 238000012545 processing Methods 0.000 claims abstract description 7
- 239000013598 vector Substances 0.000 claims description 162
- 238000012549 training Methods 0.000 claims description 71
- 239000011159 matrix material Substances 0.000 claims description 14
- 235000004257 Cordia myxa Nutrition 0.000 claims description 3
- 244000157795 Cordia myxa Species 0.000 claims description 3
- 238000005259 measurement Methods 0.000 claims description 3
- 230000011218 segmentation Effects 0.000 claims description 2
- 238000005516 engineering process Methods 0.000 abstract description 7
- 230000007547 defect Effects 0.000 abstract description 5
- 238000004458 analytical method Methods 0.000 abstract description 4
- 238000012163 sequencing technique Methods 0.000 abstract description 4
- 238000007796 conventional method Methods 0.000 description 4
- 238000013480 data collection Methods 0.000 description 4
- 238000010586 diagram Methods 0.000 description 4
- 230000008569 process Effects 0.000 description 3
- 230000008859 change Effects 0.000 description 2
- 230000005611 electricity Effects 0.000 description 2
- 238000005265 energy consumption Methods 0.000 description 2
- 238000012544 monitoring process Methods 0.000 description 2
- 238000012913 prioritisation Methods 0.000 description 2
- 230000003068 static effect Effects 0.000 description 2
- 238000004364 calculation method Methods 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 230000003993 interaction Effects 0.000 description 1
- 230000014759 maintenance of location Effects 0.000 description 1
- 238000013507 mapping Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000003062 neural network model Methods 0.000 description 1
- 230000004044 response Effects 0.000 description 1
- 230000026676 system process Effects 0.000 description 1
Landscapes
- Machine Translation (AREA)
Abstract
The application discloses a digital twinning-based intelligent acquisition task scheduling method and device, which are used for carrying out semantic analysis and understanding of task texts and device capability texts by acquiring a plurality of task text descriptions and data acquisition device capability descriptions of data acquisition, introducing a data processing and text semantic understanding algorithm at the rear end, and carrying out sequencing and optimization of each data acquisition task based on semantic relativity between the task text description and the device capability description. Therefore, the acquisition task can be automatically optimized and scheduled according to the text description of the acquisition task and the equipment capability, the real-time and dynamic representation of the equipment and the task can be realized by utilizing a digital twin technology, the defects of the traditional method can be overcome, and the accuracy, the efficiency and the flexibility of the data acquisition task scheduling are improved.
Description
Technical Field
The application relates to the field of intelligent scheduling, in particular to an intelligent acquisition task scheduling method and device based on digital twinning.
Background
In power marketing business, data collection tasks are critical, which involve collecting data from various power devices and systems, such as electricity meter readings, energy consumption, device status, etc., as well as collecting customer information and marketing data, such as information about customer demographics, preferences, and behavior. These data are critical to monitoring energy usage, optimizing energy distribution, making load predictions, and developing effective marketing strategies and making accurate business decisions.
However, the conventional data acquisition task scheduling method generally depends on a static rule or a simple optimization algorithm, and cannot flexibly adapt to the real-time change condition of data, so that the task scheduling efficiency is low. Moreover, conventional methods often lack deep understanding of task and equipment capabilities, and cannot make accurate task prioritization and scheduling decisions, resulting in low accuracy of data acquisition. That is, conventional methods often lack comprehensive consideration of device capabilities and tasks, and cannot flexibly adapt to changing service demands and device states, resulting in a lack of flexibility and adaptability of a scheduling scheme.
Accordingly, an optimized intelligent acquisition task scheduling scheme is desired.
Disclosure of Invention
The present application has been made to solve the above-mentioned technical problems. The embodiment of the application provides an intelligent acquisition task scheduling method and device based on digital twinning, which are used for carrying out semantic analysis and understanding of task texts and device capability texts by acquiring a plurality of task text descriptions and data acquisition device capability descriptions of data acquisition, introducing a data processing and text semantic understanding algorithm at the rear end, and carrying out sequencing and optimization of each data acquisition task based on semantic relativity between the task text and the device capability text. Therefore, the acquisition task can be automatically optimized and scheduled according to the text description of the acquisition task and the equipment capability, the real-time and dynamic representation of the equipment and the task can be realized by utilizing a digital twin technology, the defects of the traditional method can be overcome, and the accuracy, the efficiency and the flexibility of the data acquisition task scheduling are improved.
According to one aspect of the application, there is provided a digital twinning-based intelligent acquisition task scheduling method, comprising:
Acquiring text descriptions of a plurality of data acquisition tasks;
acquiring capability descriptions of a plurality of data acquisition devices;
Semantic coding is carried out on the text descriptions of the plurality of data acquisition tasks to obtain semantic coding feature vectors of the plurality of data acquisition tasks;
carrying out semantic coding on the capability descriptions of the plurality of data acquisition devices to obtain capability semantic coding feature vectors of the plurality of data acquisition devices;
Performing global semantic association coding on the capacity semantic coding feature vectors of the plurality of data acquisition devices to obtain capacity semantic coding feature vectors of the plurality of context data acquisition devices;
Calculating a global unidirectional semantic attention weight value of the feature distribution of each data acquisition task semantic coding feature vector in the plurality of data acquisition task semantic coding feature vectors relative to the plurality of context data acquisition equipment capacity semantic coding feature vectors to obtain a plurality of unidirectional semantic attention weight values;
and determining priorities of the plurality of data acquisition tasks based on the plurality of unidirectional semantic attention weight values, and displaying text descriptions of the plurality of data acquisition tasks, capability descriptions of the plurality of data acquisition devices and priorities of the plurality of data acquisition tasks.
According to another aspect of the present application, there is provided a digital twinning-based intelligent acquisition task scheduling apparatus, comprising:
the task text description acquisition module is used for acquiring text descriptions of a plurality of data acquisition tasks;
the device capability description module is used for acquiring capability descriptions of a plurality of data acquisition devices;
the task semantic coding module is used for carrying out semantic coding on the text descriptions of the plurality of data acquisition tasks to obtain semantic coding feature vectors of the plurality of data acquisition tasks;
the device capability semantic coding module is used for carrying out semantic coding on the capability descriptions of the plurality of data acquisition devices to obtain capability semantic coding feature vectors of the plurality of data acquisition devices;
The global semantic association coding module is used for carrying out global semantic association coding on the capacity semantic coding feature vectors of the plurality of data acquisition equipment to obtain capacity semantic coding feature vectors of the plurality of context data acquisition equipment;
The attention weight measurement module is used for calculating unidirectional semantic attention weight values of the feature distribution global of each data acquisition task semantic coding feature vector in the plurality of data acquisition task semantic coding feature vectors relative to the plurality of context data acquisition equipment capacity semantic coding feature vectors so as to obtain a plurality of unidirectional semantic attention weight values;
The result generation module is used for determining the priorities of the data acquisition tasks based on the unidirectional semantic attention weight values and displaying text descriptions of the data acquisition tasks, capability descriptions of the data acquisition devices and the priorities of the data acquisition tasks.
Compared with the prior art, the intelligent acquisition task scheduling method and device based on digital twinning provided by the application have the advantages that the task text descriptions and the data acquisition equipment capability descriptions are acquired, the data processing and text semantic understanding algorithm is introduced at the rear end to perform semantic analysis and understanding of the task text and the equipment capability text, and the sequencing and optimization of each data acquisition task are performed based on the semantic relativity between the task text and the equipment capability text. Therefore, the acquisition task can be automatically optimized and scheduled according to the text description of the acquisition task and the equipment capability, the real-time and dynamic representation of the equipment and the task can be realized by utilizing a digital twin technology, the defects of the traditional method can be overcome, and the accuracy, the efficiency and the flexibility of the data acquisition task scheduling are improved.
Drawings
The above and other objects, features and advantages of the present application will become more apparent by describing embodiments of the present application in more detail with reference to the attached drawings. The accompanying drawings are included to provide a further understanding of embodiments of the application and are incorporated in and constitute a part of this specification, illustrate the application and together with the embodiments of the application, and not constitute a limitation to the application. In the drawings, like reference numerals generally refer to like parts or steps.
FIG. 1 is a flow chart of a digital twinning-based intelligent acquisition task scheduling method according to an embodiment of the present application;
FIG. 2 is a system architecture diagram of a digital twinning-based intelligent acquisition task scheduling method according to an embodiment of the present application;
FIG. 3 is a flow chart of a training phase of a digital twinning-based intelligent acquisition task scheduling method according to an embodiment of the present application;
fig. 4 is a block diagram of an intelligent acquisition task scheduler based on digital twinning according to an embodiment of the present application.
Detailed Description
Hereinafter, exemplary embodiments according to the present application will be described in detail with reference to the accompanying drawings. It should be apparent that the described embodiments are only some embodiments of the present application and not all embodiments of the present application, and it should be understood that the present application is not limited by the example embodiments described herein.
As used in the specification and in the claims, the terms "a," "an," "the," and/or "the" are not specific to a singular, but may include a plurality, unless the context clearly dictates otherwise. In general, the terms "comprises" and "comprising" merely indicate that the steps and elements are explicitly identified, and they do not constitute an exclusive list, as other steps or elements may be included in a method or apparatus.
Although the present application makes various references to certain modules in a system according to embodiments of the present application, any number of different modules may be used and run on a user terminal and/or server. The modules are merely illustrative, and different aspects of the systems and methods may use different modules.
A flowchart is used in the present application to describe the operations performed by a system according to embodiments of the present application. It should be understood that the preceding or following operations are not necessarily performed in order precisely. Rather, the various steps may be processed in reverse order or simultaneously, as desired. Also, other operations may be added to or removed from these processes.
Hereinafter, exemplary embodiments according to the present application will be described in detail with reference to the accompanying drawings. It should be apparent that the described embodiments are only some embodiments of the present application and not all embodiments of the present application, and it should be understood that the present application is not limited by the example embodiments described herein.
The traditional data acquisition task scheduling method generally depends on static rules or simple optimization algorithms, and cannot flexibly adapt to the real-time change condition of data, so that task scheduling efficiency is low. Moreover, conventional methods often lack deep understanding of task and equipment capabilities, and cannot make accurate task prioritization and scheduling decisions, resulting in low accuracy of data acquisition. That is, conventional methods often lack comprehensive consideration of device capabilities and tasks, and cannot flexibly adapt to changing service demands and device states, resulting in a lack of flexibility and adaptability of a scheduling scheme. Accordingly, an optimized intelligent acquisition task scheduling scheme is desired.
Digital twinning refers to a virtual representation of an entity or process based on a digitized model that can simulate, predict, and optimize the behavior of the entity or process. In intelligent data acquisition task scheduling, a digital twin technology can be used for constructing a virtual model of data acquisition tasks and equipment so as to better understand the relationship between the tasks and the equipment and optimize task scheduling strategies.
In the technical scheme of the application, an intelligent acquisition task scheduling method based on digital twinning is provided. Fig. 1 is a flowchart of an intelligent acquisition task scheduling method based on digital twinning according to an embodiment of the present application. Fig. 2 is a system architecture diagram of an intelligent acquisition task scheduling method based on digital twinning according to an embodiment of the present application. As shown in fig. 1 and 2, the intelligent acquisition task scheduling method based on digital twin according to the embodiment of the application comprises the following steps: s1, acquiring text descriptions of a plurality of data acquisition tasks; s2, acquiring capability descriptions of a plurality of data acquisition devices; s3, carrying out semantic coding on the text descriptions of the data acquisition tasks to obtain semantic coding feature vectors of the data acquisition tasks; s4, carrying out semantic coding on the capability descriptions of the plurality of data acquisition devices to obtain capability semantic coding feature vectors of the plurality of data acquisition devices; s5, performing global semantic association coding on the semantic coding feature vectors of the capacity of the plurality of data acquisition devices to obtain semantic coding feature vectors of the capacity of the plurality of context data acquisition devices; s6, calculating a global unidirectional semantic attention weight value of the feature distribution of each data acquisition task semantic coding feature vector in the plurality of data acquisition task semantic coding feature vectors relative to the plurality of context data acquisition equipment capacity semantic coding feature vectors to obtain a plurality of unidirectional semantic attention weight values; and S7, determining the priorities of the data acquisition tasks based on the unidirectional semantic attention weight values, and displaying text descriptions of the data acquisition tasks, capability descriptions of the data acquisition devices and the priorities of the data acquisition tasks.
In particular, the step S1 and the step S2 are used for acquiring text descriptions of a plurality of data acquisition tasks; and acquiring capability descriptions of the plurality of data acquisition devices. Among other things, these tasks involve collecting data from various power devices and systems, such as electricity meter readings, energy consumption, device status, etc., as well as collecting customer information and marketing data, such as information regarding customer demographics, preferences, and behavior. These data are critical to monitoring energy usage, optimizing energy distribution, making load predictions, and developing effective marketing strategies and making accurate business decisions.
In particular, the step S3 performs semantic coding on the text descriptions of the plurality of data acquisition tasks to obtain semantic coding feature vectors of the plurality of data acquisition tasks. Considering that related text description semantics exist in the text description of each data acquisition task and the capability description of the data acquisition device, the text descriptions often contain rich semantic information, such as the emergency degree of the task, the performance index of the device and the like. Based on the above, in order to enable semantic understanding of text descriptions and equipment capability descriptions of the tasks and provide a basis for subsequent task scheduling decisions, in the technical scheme of the application, the text descriptions of the plurality of data acquisition tasks need to be subjected to semantic encoding so as to extract semantic encoding features in the text descriptions of the data acquisition tasks respectively, thereby obtaining semantic encoding feature vectors of the plurality of data acquisition tasks. Specifically, performing semantic coding on text descriptions of the plurality of data acquisition tasks to obtain semantic coding feature vectors of the plurality of data acquisition tasks, including: word segmentation processing is carried out on the text descriptions of the plurality of data acquisition tasks so as to convert the text descriptions of the plurality of data acquisition tasks into word sequences composed of a plurality of words; using a word embedding layer of a data acquisition task semantic understanding device comprising the word embedding layer to map each word in the word sequence into a word embedding vector so as to obtain a sequence of the word embedding vector; using a converter of a data acquisition task semantic comprehener comprising a word embedding layer to carry out global context semantic coding on a sequence of the word embedding vectors based on a converter thought so as to obtain a plurality of global context semantic feature vectors; and cascading the plurality of global context semantic feature vectors to obtain the plurality of data acquisition task semantic coding feature vectors.
In particular, the step S4 is to perform semantic coding on the capability descriptions of the plurality of data acquisition devices to obtain a plurality of semantic coding feature vectors of the capability of the data acquisition devices. That is, in the technical scheme of the application, the capability descriptions of the plurality of data acquisition devices are subjected to semantic coding so as to extract semantic coding features of the capability descriptions of the data acquisition devices respectively, thereby obtaining capability semantic coding feature vectors of the plurality of data acquisition devices. That is, semantic coding may convert task text descriptions and device capability descriptions into high-dimensional feature vectors that contain key features of the task and device. Such a characteristic representation facilitates a better understanding of the relationship between tasks and devices by subsequent task scheduling algorithms, thereby making scheduling decisions more accurately. It is contemplated that because of the critical semantics of the task that exist in each data acquisition device, and that the capabilities and characteristics between different devices may interact in the scheduling of data acquisition tasks, e.g., there may be a cooperative or competing relationship between certain devices. Thus, there is associated semantic information between the capability semantics of these data collection devices. Based on the above, in the technical scheme of the application, the capacity semantic coding feature vectors of the plurality of data acquisition devices are further coded in a context coder based on a converter module, so as to extract context associated feature information among capacity semantic coding features of each data acquisition device, thereby obtaining the capacity semantic coding feature vectors of the plurality of context data acquisition devices. Through the context encoder, the relation among the devices can be taken into consideration, so that the capability of each device is more comprehensively understood, and a decision basis is provided for the priority scheduling of the subsequent data acquisition task. Specifically, in one specific example of the present application, the plurality of data acquisition device capability semantic coding feature vectors are arranged in one dimension to obtain a global data acquisition device capability semantic coding feature vector; calculating the product between the global data acquisition equipment capacity semantic coding feature vector and the transpose vector of each data acquisition equipment capacity semantic coding feature vector in the plurality of data acquisition equipment capacity semantic coding feature vectors to obtain a plurality of self-attention association matrixes; respectively carrying out standardization processing on each self-attention correlation matrix in the plurality of self-attention correlation matrices to obtain a plurality of standardized self-attention correlation matrices; obtaining a plurality of probability values by using a Softmax classification function through each normalized self-attention correlation matrix in the normalized self-attention correlation matrices; and weighting each data acquisition equipment capacity semantic coding feature vector in the plurality of data acquisition equipment capacity semantic coding feature vectors by taking each probability value in the plurality of probability values as a weight so as to obtain the plurality of context semantic data acquisition equipment capacity semantic coding feature vectors.
In particular, the step S5 is to perform global semantic association encoding on the plurality of semantic encoding feature vectors of the data acquisition device capability to obtain a plurality of semantic encoding feature vectors of the context data acquisition device capability. It should be appreciated that different data acquisition devices may provide different types of feature vectors, which may be integrated together by global semantic association coding to form a unified feature representation. This helps to integrate the information collected by the different devices.
In particular, the step S6 is to calculate a global unidirectional semantic attention weight of each data acquisition task semantic coding feature vector of the plurality of data acquisition task semantic coding feature vectors with respect to the feature distribution of the plurality of context data acquisition device capability semantic coding feature vectors to obtain a plurality of unidirectional semantic attention weight values. It should be appreciated that for each data acquisition task, it needs to take into account global information about the capabilities of the plurality of acquisition devices, and thus, attention weighting of the respective task based on the context semantics of the respective data acquisition device is required to make a scheduling decision for the task. Based on the above, in order to determine the importance of each data acquisition task when considering the device capability, so as to better understand the relationship between the task and the device, so as to more accurately perform task scheduling decision, in the technical scheme of the application, the feature distribution global unidirectional semantic attention weight of each data acquisition task semantic coding feature vector in the multiple data acquisition task semantic coding feature vectors relative to the multiple context data acquisition device capability semantic coding feature vectors is further calculated so as to obtain multiple unidirectional semantic attention weight values. That is, the degree of dependence of different data acquisition tasks on the capabilities of each device may be different, and by calculating the unidirectional semantic attention weight value, the attention weight of each task relative to the global device capability semantics can be analyzed, so that the system can more effectively schedule appropriate devices according to the task characteristics, thereby improving the overall efficiency of the system. Specifically, calculating a unidirectional semantic attention weight value of a feature distribution global of each data acquisition task semantic coding feature vector in the plurality of data acquisition task semantic coding feature vectors relative to the plurality of context data acquisition device capability semantic coding feature vectors to obtain a plurality of unidirectional semantic attention weight values, including: calculating a global unidirectional semantic attention weight value of the feature distribution of each data acquisition task semantic coding feature vector in the plurality of data acquisition task semantic coding feature vectors relative to the plurality of context data acquisition device capability semantic coding feature vectors according to the following attention weight formula to obtain a plurality of unidirectional semantic attention weight values; wherein, the attention weight formula is:
wherein, Representing the semantically encoded feature vectors of the respective data acquisition tasks,Representation ofIs a matrix of the (c) in the matrix,Equal to the dimension of the semantically encoded feature vector of each of the data acquisition tasks,Is thatIs a matrix of the (c) in the matrix,Equal to the number of feature vectors in the plurality of context data acquisition device capability semantic coding feature vectors,The representation selu activates the function,The variables that represent the input are represented by,Which represents a predetermined parameter that is to be used,Meaning that if it is,Representing each of the plurality of context data acquisition device capability semantic coding feature vectors,Representing a scale of each of the plurality of context data acquisition device capability semantic coding feature vectors,Representing the respective one-way semantic attention weight values, anRepresenting a predetermined weight parameter.
In particular, the step S7 of determining priorities of the plurality of data acquisition tasks based on the plurality of unidirectional semantic attention weight values, displaying text descriptions of the plurality of data acquisition tasks, capability descriptions of the plurality of data acquisition devices, and ordering based on the plurality of unidirectional semantic attention weight values, and determining priorities of the plurality of data acquisition tasks; displaying text descriptions of the plurality of data acquisition tasks, capability descriptions of the plurality of data acquisition devices, and priorities of the plurality of data acquisition tasks. The priorities of the plurality of data acquisition tasks. In particular, in one specific example of the present application, it should be appreciated that tasks of higher importance may be prioritized in front of the execution order based on the ordering of the unidirectional semantic attention weighting values. The method is helpful for optimizing task scheduling, ensures that the system processes the most critical tasks first, and improves the overall efficiency of the system. Meanwhile, tasks with higher importance are placed at positions with higher priorities, so that the tasks can be ensured to obtain more resources and attention, and the efficiency and quality of task completion are improved. In addition, the digital twinning technique can integrate information of different data sources together to visually demonstrate text descriptions of multiple data acquisition tasks, capability descriptions of data acquisition devices, and priorities of tasks. Such comprehensive presentation helps the user to fully understand the relationship between tasks, devices and priorities.
It should be appreciated that training of the data acquisition task semantic understander and the converter module based context encoder is required before the inference is made using the neural network model described above. That is, in the intelligent acquisition task scheduling method based on digital twin of the present application, a training phase is further included for training the data acquisition task semantic understanding device and the context encoder based on the converter module.
FIG. 3 is a flow chart of a training phase of a digital twinning-based intelligent acquisition task scheduling method according to an embodiment of the present application. As shown in fig. 3, the intelligent acquisition task scheduling method based on digital twin according to the embodiment of the application includes: a training phase comprising: s110, training data is acquired, wherein the training data comprises training text descriptions of the plurality of data acquisition tasks; and, training capability descriptions of the plurality of data acquisition devices; s120, carrying out semantic coding on training text descriptions of the plurality of data acquisition tasks to obtain semantic coding feature vectors of the plurality of training data acquisition tasks; s130, optimizing the semantic coding feature vectors of the training data acquisition tasks to obtain semantic coding feature vectors of the training data acquisition tasks; s140, carrying out semantic coding on the training capacity descriptions of the plurality of data acquisition devices to obtain a plurality of training data acquisition device capacity semantic coding feature vectors; s150, enabling the plurality of training data acquisition equipment capacity semantic coding feature vectors to pass through a context encoder based on a converter module to obtain the plurality of training context data acquisition equipment capacity semantic coding feature vectors; s160, calculating a global unidirectional semantic attention weight value of the feature distribution of each optimized training data acquisition task semantic coding feature vector in the plurality of optimized training data acquisition task semantic coding feature vectors relative to the plurality of training context data acquisition equipment capacity semantic coding feature vectors to obtain a plurality of training unidirectional semantic attention weight values; s170, calculating a difference loss function value between the training unidirectional semantic attention weight values and the real training unidirectional semantic attention weight values; s180, training the data acquisition task semantic understanding device and the context encoder based on the converter module based on the difference loss function value.
In particular, in the technical scheme of the application, the plurality of training data acquisition task semantic coding feature vectors and the plurality of training data acquisition device capability semantic coding feature vectors respectively express the training text description of the plurality of data acquisition tasks and the coding text semantic features of the training capability description of the plurality of data acquisition devices, so that after the plurality of training data acquisition device capability semantic coding feature vectors pass through a context encoder based on a converter module, single sample inter-domain distribution associated features based on single sample domain coding semantic feature distribution context relevance in a multi-sample domain can be extracted, and thus, when each training data acquisition task semantic coding feature vector in the plurality of training data acquisition task semantic coding feature vectors is compared with a unidirectional semantic attention weight value of feature distribution overall of the plurality of training context data acquisition device capability semantic coding feature vectors, the local feature distribution based on the training data acquisition device capability semantic coding feature vectors in the multi-sample domain semantic feature vectors can be calculated, and the local feature distribution associated with the single training data acquisition task semantic coding feature vectors in the single training data acquisition device capability semantic feature vectors can be considered to be used as the unidirectional semantic attention weight value of the overall training data of the training data acquisition device capability semantic feature, and the local feature distribution associated with the single training data acquisition device semantic feature vectors can be mapped to the unidirectional semantic feature vectors, and the local feature vectors can be used as the overall training feature data overall feature data with the overall feature data feature distribution associated with the single text feature. Therefore, the applicant optimizes the plurality of training data acquisition task semantic coding feature vectors at each model iteration, i.e. based on the back propagation of the difference loss function between the predicted plurality of training unidirectional semantic attention weight values and the real plurality of training unidirectional semantic attention weight values, by:
Wherein the method comprises the steps of AndRespectively cascading feature vectors obtained by cascading the semantic coding feature vectors of the training data acquisition tasksIs the first of (2)And (d)The characteristic value of the location is used to determine,The multiplication operation is represented by a number of steps,Representing addition by location. That is, by introducing a training cascade feature vector of the plurality of training data acquisition task semantic coding feature vector cascadesIs used as an external information source for carrying out the retrieval enhancement of the feature vectors so as to avoid the training cascade feature vectors caused by the local overflow information distribution based on the local statistics intensive information structuringTo obtain the training cascade feature vectorInformation-bearing reasoning based on local distribution group dimension retention, i.e. obtaining the training cascade feature vectorAnd the credible distribution mapping response in the text semantic feature space of the local-global sample domain of the semantic coding feature vectors is generated based on the local feature distribution, so that the calculation accuracy of the unidirectional semantic attention weight values based on the overall correlation is improved. Therefore, the collection task can be automatically optimized and scheduled according to the text description semantics of a plurality of tasks and equipment capability, the real-time and dynamic representation of the equipment and the tasks can be realized by utilizing a digital twin technology, the defects of the traditional method can be overcome, and the accuracy, the efficiency and the flexibility of the data collection task scheduling are improved.
In summary, the intelligent acquisition task scheduling method based on digital twinning according to the embodiment of the application is explained, which performs semantic analysis and understanding of task texts and device capability texts by acquiring a plurality of task text descriptions and data acquisition device capability descriptions of data acquisition, introducing a data processing and text semantic understanding algorithm at the back end, and performing sequencing and optimization of each data acquisition task based on semantic relativity between the task text description and the data acquisition device capability description. Therefore, the acquisition task can be automatically optimized and scheduled according to the text description of the acquisition task and the equipment capability, the real-time and dynamic representation of the equipment and the task can be realized by utilizing a digital twin technology, the defects of the traditional method can be overcome, and the accuracy, the efficiency and the flexibility of the data acquisition task scheduling are improved.
Further, an intelligent acquisition task scheduling device based on digital twinning is also provided.
Fig. 4 is a block diagram of an intelligent acquisition task scheduler based on digital twinning according to an embodiment of the present application. As shown in fig. 4, the intelligent acquisition task scheduling apparatus 300 based on digital twin according to an embodiment of the present application includes: a task text description collection module 310, configured to obtain text descriptions of a plurality of data collection tasks; a device capability description module 320, configured to obtain capability descriptions of a plurality of data acquisition devices; the task semantic coding module 330 is configured to perform semantic coding on the text descriptions of the plurality of data acquisition tasks to obtain a plurality of data acquisition task semantic coding feature vectors; the device capability semantic coding module 340 is configured to perform semantic coding on capability descriptions of the plurality of data acquisition devices to obtain a plurality of data acquisition device capability semantic coding feature vectors; the global semantic association encoding module 350 is configured to perform global semantic association encoding on the capability semantic encoding feature vectors of the plurality of data acquisition devices to obtain capability semantic encoding feature vectors of the plurality of context data acquisition devices; the attention weight measurement module 360 is configured to calculate a unidirectional semantic attention weight value of a feature distribution global of each data acquisition task semantic coding feature vector in the plurality of data acquisition task semantic coding feature vectors relative to the plurality of context data acquisition device capability semantic coding feature vectors to obtain a plurality of unidirectional semantic attention weight values; the result generating module 370 is configured to determine priorities of the plurality of data acquisition tasks based on the plurality of unidirectional semantic attention weight values, and display text descriptions of the plurality of data acquisition tasks, capability descriptions of the plurality of data acquisition devices, and priorities of the plurality of data acquisition tasks.
As described above, the digital twinning-based intelligent acquisition task scheduling apparatus 300 according to the embodiment of the present application may be implemented in various wireless terminals, such as a server or the like having a digital twinning-based intelligent acquisition task scheduling algorithm. In one possible implementation, the digital twinning-based intelligent acquisition task scheduler 300 according to embodiments of the present application may be integrated into the wireless terminal as a software module and/or hardware module. For example, the digital twinning-based intelligent acquisition task scheduler 300 may be a software module in the operating system of the wireless terminal, or may be an application developed for the wireless terminal; of course, the digital twinning-based intelligent acquisition task scheduler 300 may also be one of the many hardware modules of the wireless terminal.
Alternatively, in another example, the digital twinning-based intelligent acquisition task scheduler 300 and the wireless terminal may be separate devices, and the digital twinning-based intelligent acquisition task scheduler 300 may be connected to the wireless terminal through a wired and/or wireless network and transmit interaction information in a agreed data format.
The foregoing description of the embodiments of the present disclosure has been presented for purposes of illustration and description, and is not intended to be exhaustive or limited to the embodiments disclosed. Many modifications and variations will be apparent to those of ordinary skill in the art without departing from the scope and spirit of the various embodiments described. The terminology used herein was chosen in order to best explain the principles of the embodiments, the practical application, or the improvement of technology in the marketplace, or to enable others of ordinary skill in the art to understand the embodiments disclosed herein.
Claims (7)
1. The intelligent acquisition task scheduling method based on digital twinning is characterized by comprising the following steps of:
Acquiring text descriptions of a plurality of data acquisition tasks;
acquiring capability descriptions of a plurality of data acquisition devices;
Semantic coding is carried out on the text descriptions of the plurality of data acquisition tasks to obtain semantic coding feature vectors of the plurality of data acquisition tasks;
carrying out semantic coding on the capability descriptions of the plurality of data acquisition devices to obtain capability semantic coding feature vectors of the plurality of data acquisition devices;
Performing global semantic association coding on the capacity semantic coding feature vectors of the plurality of data acquisition devices to obtain capacity semantic coding feature vectors of the plurality of context data acquisition devices;
Calculating a global unidirectional semantic attention weight value of the feature distribution of each data acquisition task semantic coding feature vector in the plurality of data acquisition task semantic coding feature vectors relative to the plurality of context data acquisition equipment capacity semantic coding feature vectors to obtain a plurality of unidirectional semantic attention weight values;
determining priorities of the plurality of data acquisition tasks based on the plurality of unidirectional semantic attention weight values, and displaying text descriptions of the plurality of data acquisition tasks, capability descriptions of the plurality of data acquisition devices and priorities of the plurality of data acquisition tasks;
The calculating a unidirectional semantic attention weight value of a feature distribution global of each data acquisition task semantic coding feature vector in the plurality of data acquisition task semantic coding feature vectors relative to the plurality of context data acquisition equipment capability semantic coding feature vectors to obtain a plurality of unidirectional semantic attention weight values includes:
Calculating a global unidirectional semantic attention weight value of the feature distribution of each data acquisition task semantic coding feature vector in the plurality of data acquisition task semantic coding feature vectors relative to the plurality of context data acquisition device capability semantic coding feature vectors according to the following attention weight formula to obtain a plurality of unidirectional semantic attention weight values;
wherein, the attention weight formula is:
wherein, Representing the semantically encoded feature vectors of the respective data acquisition tasks,Representation ofIs a matrix of the (c) in the matrix,Equal to the dimension of the semantically encoded feature vector of each of the data acquisition tasks,Is thatIs a matrix of the (c) in the matrix,Equal to the number of feature vectors in the plurality of context data acquisition device capability semantic coding feature vectors,The representation selu activates the function,The variables that represent the input are represented by,Which represents a predetermined parameter that is to be used,Meaning that if it is,Representing each of the plurality of context data acquisition device capability semantic coding feature vectors,Representing a scale of each of the plurality of context data acquisition device capability semantic coding feature vectors,Represents the one-way semantic attention weight value, andRepresenting a predetermined weight parameter.
2. The intelligent acquisition task scheduling method based on digital twinning according to claim 1, wherein performing semantic coding on text descriptions of the plurality of data acquisition tasks to obtain a plurality of data acquisition task semantic coding feature vectors comprises:
Word segmentation processing is carried out on the text descriptions of the plurality of data acquisition tasks so as to convert the text descriptions of the plurality of data acquisition tasks into word sequences composed of a plurality of words;
Using a word embedding layer of a data acquisition task semantic understanding device comprising the word embedding layer to map each word in the word sequence into a word embedding vector so as to obtain a sequence of the word embedding vector;
Using a converter of a data acquisition task semantic comprehener comprising a word embedding layer to carry out global context semantic coding on a sequence of the word embedding vectors based on a converter thought so as to obtain a plurality of global context semantic feature vectors; and
And cascading the plurality of global context semantic feature vectors to obtain the plurality of data acquisition task semantic coding feature vectors.
3. The intelligent acquisition task scheduling method based on digital twinning according to claim 2, wherein performing global semantic association encoding on the plurality of data acquisition device capability semantic encoding feature vectors to obtain a plurality of context data acquisition device capability semantic encoding feature vectors comprises: and passing the plurality of data acquisition device capability semantic coding feature vectors through a context encoder based on a converter module to obtain the plurality of context data acquisition device capability semantic coding feature vectors.
4. The digital twinning-based intelligent acquisition task scheduling method of claim 3, wherein determining priorities of the plurality of data acquisition tasks and displaying text descriptions of the plurality of data acquisition tasks, capability descriptions of the plurality of data acquisition devices, and priorities of the plurality of data acquisition tasks based on the plurality of unidirectional semantic attention weight values comprises:
Determining priorities of the plurality of data acquisition tasks based on the ordering of the plurality of unidirectional semantic attention weight values;
Displaying text descriptions of the plurality of data acquisition tasks, capability descriptions of the plurality of data acquisition devices, and priorities of the plurality of data acquisition tasks.
5. The intelligent acquisition task scheduling method based on digital twinning according to claim 4, further comprising the training step of: for training the data acquisition task semantic comprehension and the context encoder based on the converter module.
6. The intelligent acquisition task scheduling method based on digital twinning according to claim 5, wherein the training step comprises:
Acquiring training data, wherein the training data comprises training text descriptions of the plurality of data acquisition tasks; and, training capability descriptions of the plurality of data acquisition devices;
carrying out semantic coding on training text descriptions of the plurality of data acquisition tasks to obtain semantic coding feature vectors of the plurality of training data acquisition tasks;
Optimizing the semantic coding feature vectors of the training data acquisition tasks to obtain semantic coding feature vectors of the training data acquisition tasks;
Carrying out semantic coding on the training capacity descriptions of the plurality of data acquisition devices to obtain a plurality of training data acquisition device capacity semantic coding feature vectors;
Passing the plurality of training data acquisition device capability semantic coding feature vectors through a context encoder based on a converter module to obtain the plurality of training context data acquisition device capability semantic coding feature vectors;
Calculating a global unidirectional semantic attention weight value of the feature distribution of each optimization training data acquisition task semantic coding feature vector in the plurality of optimization training data acquisition task semantic coding feature vectors relative to the plurality of training context data acquisition equipment capacity semantic coding feature vectors to obtain a plurality of training unidirectional semantic attention weight values;
calculating a difference loss function value between the training unidirectional semantic attention weight values and the real training unidirectional semantic attention weight values;
training the data acquisition task semantic comprehener and the context encoder based on the converter module based on the difference loss function value.
7. An intelligent acquisition task scheduling device based on digital twinning is characterized by comprising:
the task text description acquisition module is used for acquiring text descriptions of a plurality of data acquisition tasks;
the device capability description module is used for acquiring capability descriptions of a plurality of data acquisition devices;
the task semantic coding module is used for carrying out semantic coding on the text descriptions of the plurality of data acquisition tasks to obtain semantic coding feature vectors of the plurality of data acquisition tasks;
the device capability semantic coding module is used for carrying out semantic coding on the capability descriptions of the plurality of data acquisition devices to obtain capability semantic coding feature vectors of the plurality of data acquisition devices;
The global semantic association coding module is used for carrying out global semantic association coding on the capacity semantic coding feature vectors of the plurality of data acquisition equipment to obtain capacity semantic coding feature vectors of the plurality of context data acquisition equipment;
The attention weight measurement module is used for calculating unidirectional semantic attention weight values of the feature distribution global of each data acquisition task semantic coding feature vector in the plurality of data acquisition task semantic coding feature vectors relative to the plurality of context data acquisition equipment capacity semantic coding feature vectors so as to obtain a plurality of unidirectional semantic attention weight values;
The result generation module is used for determining the priorities of the data acquisition tasks based on the unidirectional semantic attention weight values and displaying text descriptions of the data acquisition tasks, capability descriptions of the data acquisition devices and the priorities of the data acquisition tasks;
The calculating a unidirectional semantic attention weight value of a feature distribution global of each data acquisition task semantic coding feature vector in the plurality of data acquisition task semantic coding feature vectors relative to the plurality of context data acquisition equipment capability semantic coding feature vectors to obtain a plurality of unidirectional semantic attention weight values includes:
Calculating a global unidirectional semantic attention weight value of the feature distribution of each data acquisition task semantic coding feature vector in the plurality of data acquisition task semantic coding feature vectors relative to the plurality of context data acquisition device capability semantic coding feature vectors according to the following attention weight formula to obtain a plurality of unidirectional semantic attention weight values;
wherein, the attention weight formula is:
wherein, Representing the semantically encoded feature vectors of the respective data acquisition tasks,Representation ofIs a matrix of the (c) in the matrix,Equal to the dimension of the semantically encoded feature vector of each of the data acquisition tasks,Is thatIs a matrix of the (c) in the matrix,Equal to the number of feature vectors in the plurality of context data acquisition device capability semantic coding feature vectors,The representation selu activates the function,The variables that represent the input are represented by,Which represents a predetermined parameter that is to be used,Meaning that if it is,Representing each of the plurality of context data acquisition device capability semantic coding feature vectors,Representing a scale of each of the plurality of context data acquisition device capability semantic coding feature vectors,Represents the one-way semantic attention weight value, andRepresenting a predetermined weight parameter.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202410494247.7A CN118069331B (en) | 2024-04-24 | 2024-04-24 | Intelligent acquisition task scheduling method and device based on digital twinning |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202410494247.7A CN118069331B (en) | 2024-04-24 | 2024-04-24 | Intelligent acquisition task scheduling method and device based on digital twinning |
Publications (2)
Publication Number | Publication Date |
---|---|
CN118069331A CN118069331A (en) | 2024-05-24 |
CN118069331B true CN118069331B (en) | 2024-07-09 |
Family
ID=91107815
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202410494247.7A Active CN118069331B (en) | 2024-04-24 | 2024-04-24 | Intelligent acquisition task scheduling method and device based on digital twinning |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN118069331B (en) |
Citations (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN117495421A (en) * | 2023-11-13 | 2024-02-02 | 国家电网有限公司信息通信分公司 | Power grid communication engineering cost prediction method based on power communication network construction |
CN117540935A (en) * | 2024-01-09 | 2024-02-09 | 上海银行股份有限公司 | DAO operation management method based on block chain technology |
Family Cites Families (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN115878330B (en) * | 2023-02-08 | 2023-05-30 | 腾云创威信息科技(威海)有限公司 | Thread operation control method and system |
CN117744127B (en) * | 2024-02-20 | 2024-05-07 | 北京佳芯信息科技有限公司 | Data encryption authentication method and system based on data information protection |
-
2024
- 2024-04-24 CN CN202410494247.7A patent/CN118069331B/en active Active
Patent Citations (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN117495421A (en) * | 2023-11-13 | 2024-02-02 | 国家电网有限公司信息通信分公司 | Power grid communication engineering cost prediction method based on power communication network construction |
CN117540935A (en) * | 2024-01-09 | 2024-02-09 | 上海银行股份有限公司 | DAO operation management method based on block chain technology |
Also Published As
Publication number | Publication date |
---|---|
CN118069331A (en) | 2024-05-24 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN111435463B (en) | Data processing method, related equipment and system | |
CN108629358B (en) | Object class prediction method and device | |
Yu et al. | Pricing information in smart grids: A quality-based data valuation paradigm | |
CN117132135A (en) | Urban informatization management system and method based on digital twinning | |
CN115562940A (en) | Load energy consumption monitoring method and device, medium and electronic equipment | |
CN116663568A (en) | Critical task identification system and method based on priority | |
CN117198421A (en) | Intelligent detection system and method for marine environment | |
US11651271B1 (en) | Artificial intelligence system incorporating automatic model updates based on change point detection using likelihood ratios | |
CN117495421A (en) | Power grid communication engineering cost prediction method based on power communication network construction | |
CN117744855A (en) | Load prediction system and method based on machine learning | |
CN117875638A (en) | Big data-based power supply demand distribution method and system | |
CN118069331B (en) | Intelligent acquisition task scheduling method and device based on digital twinning | |
CN111488517A (en) | Method and device for training click rate estimation model | |
CN110210677B (en) | Bus short-term daily load prediction method and device combining clustering and deep learning algorithm | |
US11636377B1 (en) | Artificial intelligence system incorporating automatic model updates based on change point detection using time series decomposing and clustering | |
US20230419195A1 (en) | System and Method for Hierarchical Factor-based Forecasting | |
CN115809818A (en) | Multidimensional diagnosis and evaluation method and device for auxiliary equipment of pumped storage power station | |
CN114372835B (en) | Comprehensive energy service potential customer identification method, system and computer equipment | |
CN113139129B (en) | Virtual reading trajectory graph generation method and device, electronic equipment and storage medium | |
CN113935407A (en) | Abnormal behavior recognition model determining method and device | |
KR20240011823A (en) | Methods for quantifying product condition and predicting remaining life, and devices and systems | |
CN114004408A (en) | User power load prediction method based on data analysis | |
CN112734205A (en) | Model confidence degree analysis method and device, electronic equipment and computer storage medium | |
Meng et al. | Classification of customer service tickets in power system based on character and word level semantic understanding | |
CN117724858B (en) | Geospatial analysis model scheduling method, device, equipment and medium |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
GR01 | Patent grant | ||
GR01 | Patent grant |