Method and device for generating object model in Internet of Things in smart park, medium and equipment

Abstract

The invention belongs to the technical field of computer vision application, and particularly discloses an object model generation method and device in the smart park Internet of Things, a medium and equipment, and the method comprises the steps: obtaining text information describing to-be-generated equipment; extracting name information and feature information of the to-be-generated equipment according to the text information; retrieving a corresponding object model template in a preset standard object model library according to the name information; determining an object model dimension corresponding to each feature according to the feature information and a preset classification neural network model; and adding each feature in the feature information to the corresponding object model dimension of the corresponding object model template, and generating an object model of the to-be-generated device. The method can greatly improve the efficiency and reduce the cost, reduces the manual coding and debugging, and shortens the development period.

Description

Technical Field

[0001] The present invention belongs to the field of computer vision application technology, and specifically relates to a method, device, medium and equipment for generating object models in a smart park Internet of Things. Background Art

[0002] In today's digital age, the construction of smart parks has become a key trend in urban modernization. With the help of various advanced information technology means, smart parks are committed to building an efficient, intelligent, green and sustainable park environment.

[0003] As the core supporting technology, the Internet of Things interconnects various types of equipment, facilities and sensors widely distributed in the park to realize data collection, transmission and sharing, thus providing a basic basis for the intelligent management and operation decision-making of the park. From intelligent building systems to energy management facilities, from security monitoring networks to environmental monitoring equipment, many physical objects with different functions and types are included in the scope of the Internet of Things. The large amount of real-time data generated by these physical objects contains rich information value.

[0004] As the scale of smart parks continues to expand, the number of IoT devices and the amount of data generated are growing explosively. In a large-scale device and data environment, how to efficiently collect, store, process and analyze data to achieve automatic generation of object models is a huge challenge.

[0005] Traditional manual modeling methods are obviously unable to meet the needs of such massive data and equipment, and existing automated generation technologies often face problems such as performance bottlenecks and difficulty in maintaining data consistency when processing large-scale data, which affects the efficiency and accuracy of object model generation.

[0006] Based on the above problems, the present invention provides a method, device, medium and equipment for generating a physical model in a smart park Internet of Things. Summary of the invention

[0007] Purpose of the invention: The purpose of the present invention is to provide a method, device, medium and equipment for generating a physical model in a smart park Internet of Things.

[0008] Technical solution: To achieve the above purpose, the present invention is implemented by adopting the following technical solution:

[0009] In a first aspect, the present invention provides a method for generating a physical model in a smart park Internet of Things, comprising:

[0010] Obtain text information describing the device to be generated;

[0011] Extracting the name information and feature information of the device to be generated according to the text information;

[0012] According to the name information, a corresponding object model template is retrieved from a preset standard object model library;

[0013] Determine the object model dimension corresponding to each feature according to the feature information and a preset classification neural network model;

[0014] Each feature in the feature information is added to the corresponding physical model dimension of the corresponding physical model template to generate the physical model of the device to be generated.

[0015] In combination with the first aspect, optionally, extracting the name information and feature information of the device to be generated according to the text information includes:

[0016] Perform word segmentation analysis on the text information to determine the part of speech of each word, which includes but is not limited to nouns, verbs and quantifiers;

[0017] According to the parts of speech, all noun word combinations are extracted as a noun set;

[0018] All words in the noun set are respectively input into a preset first classification neural network to determine the words belonging to the name information and the words belonging to the feature information. The input of the first classification neural network is a single word, and the output is a word category, and the word category includes a name category and a feature category.

[0019] In combination with the first aspect, optionally, searching for a corresponding object model template in a preset standard object model library according to the name information includes:

[0020] Calculating similarity values ​​between the words in the name information and the names of all object model templates in the standard model library;

[0021] The physical model template with the largest similarity value is determined as the corresponding physical model template.

[0022] In combination with the first aspect, optionally, the similarity value includes but is not limited to a Euclidean distance value, a Manhattan distance value, and a Hamming distance value.

[0023] In combination with the first aspect, optionally, determining the object model dimension corresponding to each feature according to the feature information and a preset classification neural network model includes:

[0024] Each feature in the feature information is input into the preset second classification neural network to determine the object model dimension corresponding to each feature. The input of the second classification neural network is a single word, and the output is a dimension category. The dimension category includes attributes, services and events.

[0025] In combination with the first aspect, optionally, adding each feature in the feature information to the corresponding physical model dimension of the corresponding physical model template to generate the physical model of the device to be generated includes:

[0026] Extracting, according to each feature, text corresponding to the feature in the text describing the device to be generated;

[0027] Perform word segmentation analysis on the corresponding text to determine a verb set or a quantifier set corresponding to each feature;

[0028] Generate a physical model description of each feature according to the corresponding physical model dimension and the verb set or quantifier set corresponding to each feature;

[0029] The physical model description of each feature is added to the corresponding physical model dimension of the corresponding physical model template to generate the physical model of the device to be generated.

[0030] In combination with the first aspect, optionally, extracting, according to each feature, text corresponding to the feature from the text describing the device to be generated, includes:

[0031] Get the current feature and the next adjacent feature;

[0032] All texts between the current feature and the adjacent next feature in the text describing the device to be generated are determined as the corresponding texts, wherein all the texts include the current feature but do not include the adjacent next feature.

[0033] In a second aspect, the present invention provides a device for generating a physical model in a smart park Internet of Things, comprising:

[0034] Text acquisition module: acquires text information describing the device to be generated;

[0035] Name and feature classification module: extract the name information and feature information of the device to be generated according to the text information;

[0036] Object model template retrieval module: according to the name information, retrieve the corresponding object model template in a preset standard object model library;

[0037] Physical model dimension determination module: determines the physical model dimension corresponding to each feature according to the feature information and a preset classification neural network model;

[0038] Physical model generation module: adds each feature in the feature information to the corresponding physical model dimension of the corresponding physical model template to generate the physical model of the device to be generated.

[0039] In a third aspect, the present invention provides a computer-readable storage medium having a computer program stored thereon, characterized in that when the computer program is executed by a processor, it implements a method for generating a physical model in a smart park Internet of Things as described in any one of the first aspects.

[0040] In a fourth aspect, the present invention provides a device, characterized in that it includes:

[0041] A memory for storing instructions;

[0042] A processor is used to execute the instructions so that the device performs operations to implement the method for generating a physical model in the smart park Internet of Things as described in any one of the first aspects.

[0043] Compared with the prior art, the beneficial effect of the method for generating physical models in the smart park Internet of Things of the present invention is that: the present invention obtains text information describing the device to be generated; extracts the name information and feature information of the device to be generated according to the text information; retrieves the corresponding physical model template in a preset standard physical model library according to the name information; determines the physical model dimension corresponding to each feature according to the feature information and a preset classification neural network model; adds each feature in the feature information to the corresponding physical model dimension of the corresponding physical model template to generate the physical model of the device to be generated. The present invention can greatly improve efficiency and reduce costs, reduce manual coding and debugging, and shorten the development cycle.

[0044] The physical model generation device in the smart park Internet of Things in the embodiment of the present application includes the physical model generation method described above, and therefore can have all the technical features and technical effects of the physical model generation method described above, which will not be repeated here. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0046] Figure 1 It is a flow chart of the method for generating object models in the smart park Internet of Things of the present invention;

[0047] Figure 2 This is a structural block diagram of the device for generating a physical model in the smart park Internet of Things of the present invention. DETAILED DESCRIPTION

[0048] The technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present invention.

[0049] In the description of the present invention, it should be noted that the orientations or positional relationships indicated by the terms "top", "bottom", "one side", "the other side", "front", "back", "middle part", "inside", "top end", "bottom end", etc. are based on the orientations or positional relationships shown in the drawings, and are only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore cannot be understood as limiting the present invention; the terms "first", "second", and "third" are only used for descriptive purposes, and cannot be understood as indicating or implying relative importance; in addition, unless otherwise clearly specified and limited, the terms "installed", "connected", and "connected" should be understood in a broad sense, for example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be a direct connection, or it can be an indirect connection through an intermediate medium, or it can be a connection between the two elements. For those of ordinary skill in the art, the specific meanings of the above terms in the present invention can be understood according to specific circumstances.

[0050] Embodiment 1

[0051] like Figure 1 As shown, an embodiment of the present invention provides a method for generating a physical model in a smart park Internet of Things, including:

[0052] S1: Obtain text information describing the device to be generated.

[0053] In a specific embodiment, the text information describing the device to be generated generally starts with the device name, and then describes the device characteristics. For example, for a thermometer, the text information describing it can be: a thermometer that can display temperature and humidity, where the temperature measurement range is minus twenty degrees to one hundred degrees above zero, and the humidity measurement range is zero to eighty degrees.

[0054] S2: Extracting the name information and feature information of the device to be generated according to the text information.

[0055] In a specific embodiment, extracting the name information and feature information of the device to be generated according to the text information includes:

[0056] S2-1: performing word segmentation analysis on the text information to determine the part of speech of each word, wherein the part of speech includes but is not limited to nouns, verbs and quantifiers;

[0057] Specifically, the word segmentation analysis first adopts a rule-based word segmentation method to perform preliminary segmentation of text information based on a pre-set vocabulary library and grammatical rules. For example, for common punctuation marks, such as periods, commas, question marks, etc., they are used as natural separation boundaries to divide the text into several preliminary fragments. At the same time, for some specific vocabulary combination rules, such as fixed phrases, proper nouns, etc., they are accurately identified and divided through rule matching to avoid incorrect segmentation. Then, a statistical word segmentation model is used. This model is trained based on a large-scale corpus and determines the boundaries of word segmentation by calculating the probability of word occurrence and co-occurrence probability. For example, for some vocabulary strings with multiple segmentation possibilities, according to the probability values ​​corresponding to different segmentation methods in the corpus, the segmentation scheme with the highest probability is selected, thereby improving the accuracy and rationality of word segmentation. Then, deep learning word segmentation technology is introduced, such as using a recurrent neural network (RNN) or a convolutional neural network (CNN). These models can automatically learn the semantic and grammatical features of the text and perform in-depth understanding and word segmentation on the text. They can capture long-distance dependencies in the text and complete word segmentation tasks more accurately even when faced with complex text structures, metaphorical expressions or emerging vocabulary.

[0058] S2-2: extracting all noun word combinations into a noun set according to the part of speech;

[0059] S2-3: Input all the words in the noun set into a preset first classification neural network respectively to determine the words belonging to the name information and the words belonging to the feature information. The input of the first classification neural network is a single word, and the output is a word category, which includes a name category and a feature category.

[0060] Specifically, the preset first classification neural network can adopt a random forest model framework, and perform training and learning based on the name words and feature words of the equipment in the existing object model library, and finally generate a neural network model with a single word as input and a word category as output, wherein the word category includes a name category and a feature category.

[0061] S3: According to the name information, a corresponding object model template is retrieved from a preset standard object model library.

[0062] In a specific embodiment, searching for a corresponding object model template in a preset standard object model library according to the name information includes:

[0063] S3-1: Calculate the similarity between the words in the name information and the names of all object model templates in the standard model library;

[0064] Specifically, the similarity value includes, but is not limited to, a Euclidean distance value, a Manhattan distance value, and a Hamming distance value.

[0065] S3-2: Determine the physical model template with the largest similarity value as the corresponding physical model template.

[0066] S4: Determine the object model dimension corresponding to each feature based on the feature information and a preset classification neural network model.

[0067] In a specific embodiment, determining the object model dimension corresponding to each feature according to the feature information and a preset classification neural network model includes:

[0068] S4-1: Input each feature in the feature information into the preset second classification neural network respectively to determine the object model dimension corresponding to each feature, the input of the second classification neural network is a single word, and the output is a dimension category, and the dimension category includes attributes, services and events.

[0069] Specifically, the second classification neural network model is a multi-category classification model, which can be constructed using a multi-layer perceptron. When preparing data, first collect texts from multiple channels of the Internet of Things, and then manually annotate them according to the classification standards after screening, cleaning, and word segmentation, and then divide the data set in proportion. In terms of model architecture design, the input layer determines the number of neurons to receive word semantics based on the word vector dimension, and sets a hidden layer with an activation function to enhance the expressiveness. The output layer sets neurons based on the number of categories and uses Softmax to obtain the category probability distribution. In the training and optimization stage, first randomly initialize the parameters, select the cross entropy loss function and optimization algorithms such as Adam, perform forward propagation, calculate losses, back propagate and update parameters according to batch data, and iterate for multiple rounds until the performance of the verification set is stable.

[0070] S5: Add each feature in the feature information to the corresponding physical model dimension of the corresponding physical model template to generate the physical model of the device to be generated.

[0071] In a specific embodiment, adding each feature in the feature information to the corresponding physical model dimension of the corresponding physical model template to generate the physical model of the device to be generated includes:

[0072] S5-1: extracting, according to each feature, text corresponding to the feature from the text describing the device to be generated;

[0073] Specifically, extracting the text corresponding to each feature from the text describing the device to be generated according to each feature includes:

[0074] S5-1-1: Get the current feature and the next adjacent feature;

[0075] S5-1-2: Determine all texts between the current feature and the adjacent next feature in the text describing the device to be generated as the corresponding text, wherein all the texts include the current feature but do not include the adjacent next feature.

[0076] S5-2: performing word segmentation analysis on the corresponding text to determine a verb set or a quantifier set corresponding to each feature;

[0077] S5-3: Generate a physical model description for each feature according to the corresponding physical model dimension and the verb set or quantifier set corresponding to each feature;

[0078] S5-4: Add the physical model description of each feature to the corresponding physical model dimension of the corresponding physical model template to generate the physical model of the device to be generated.

[0079] The present invention provides a method for generating a physical model in a smart park Internet of Things, which can greatly improve efficiency and reduce costs, reduce manual coding and debugging, and shorten the development cycle.

[0080] Embodiment 2

[0081] like Figure 2 As shown, an embodiment of the present invention provides a device for generating a physical model in a smart park Internet of Things, comprising:

[0082] Text acquisition module: acquires text information describing the device to be generated;

[0083] Name and feature classification module: extract the name information and feature information of the device to be generated according to the text information;

[0084] Object model template retrieval module: according to the name information, retrieve the corresponding object model template in a preset standard object model library;

[0085] Physical model dimension determination module: determines the physical model dimension corresponding to each feature according to the feature information and a preset classification neural network model;

[0086] Physical model generation module: adds each feature in the feature information to the corresponding physical model dimension of the corresponding physical model template to generate the physical model of the device to be generated.

[0087] The present invention provides a device for generating a physical model in a smart park Internet of Things, which can greatly improve efficiency and reduce costs, reduce manual coding and debugging, and shorten the development cycle.

[0088] Embodiment 3

[0089] An embodiment of the present invention provides a computer-readable storage medium having a computer program stored thereon. When the computer program is executed by a processor, a method for generating a physical model in a smart park Internet of Things as described in Embodiment 1 is implemented.

[0090] Embodiment 4

[0091] An embodiment of the present invention provides a device, including:

[0092] A memory for storing instructions;

[0093] The processor is used to execute the instructions so that the device performs the operations of the method for generating a physical model in the smart park Internet of Things as described in any one of the first embodiments.

[0094] Those skilled in the art will appreciate that the embodiments of the present application may be provided as methods, systems, or computer program products. Therefore, the present application may adopt the form of a complete hardware embodiment, a complete software embodiment, or an embodiment in combination with software and hardware. Moreover, the present application may adopt the form of a computer program product implemented in one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) that include computer-usable program code.

[0095] The present application is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of the processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0096] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1A function specified in one or more boxes.

[0097] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0098] The above is only a preferred embodiment of the present invention. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the technical principles of the present invention. These improvements and modifications should also be regarded as the scope of protection of the present invention.

Claims

1. A method for generating object models in the Internet of Things in a smart park, characterized by: include: Obtain text information describing the device to be generated; Extracting the name information and feature information of the device to be generated according to the text information; Retrieving a corresponding object model template from a preset standard object model library according to the name information; Determine the object model dimension corresponding to each feature according to the feature information and a preset classification neural network model; Each feature in the feature information is added to the corresponding physical model dimension of the corresponding physical model template to generate the physical model of the device to be generated.

2. The method for generating a physical model in the smart park Internet of Things according to claim 1 is characterized in that: Extracting the name information and feature information of the device to be generated according to the text information includes: Perform word segmentation analysis on the text information to determine the part of speech of each word, wherein the part of speech includes noun, verb and quantifier; According to the parts of speech, all noun word combinations are extracted as a noun set; All words in the noun set are respectively input into a preset first classification neural network to determine the words belonging to the name information and the words belonging to the feature information. The input of the first classification neural network is a single word, and the output is a word category, and the word category includes a name category and a feature category.

3. The method for generating a physical model in the smart park Internet of Things according to claim 1, characterized in that: According to the name information, a corresponding object model template is retrieved from a preset standard object model library, including: Calculating similarity values ​​between the words in the name information and the names of all object model templates in the standard model library; The physical model template with the largest similarity value is determined as the corresponding physical model template.

4. The method for generating a physical model in the smart park Internet of Things according to claim 3 is characterized in that: The similarity values ​​include Euclidean distance value, Manhattan distance value, and Hamming distance value.

5. The method for generating object models in the smart park Internet of Things according to claim 1 is characterized in that: Determining the object model dimension corresponding to each feature according to the feature information and a preset classification neural network model includes: Each feature in the feature information is input into the preset second classification neural network to determine the object model dimension corresponding to each feature. The input of the second classification neural network is a single word, and the output is a dimension category. The dimension category includes attributes, services and events.

6. The method for generating object models in the smart park Internet of Things according to claim 1 is characterized in that: Adding each feature in the feature information to the corresponding physical model dimension of the corresponding physical model template to generate the physical model of the device to be generated includes: Extracting, according to each feature, text corresponding to the feature in the text describing the device to be generated; Perform word segmentation analysis on the corresponding text to determine a verb set or a quantifier set corresponding to each feature; Generate a physical model description of each feature according to the corresponding physical model dimension and the verb set or quantifier set corresponding to each feature; The physical model description of each feature is added to the corresponding physical model dimension of the corresponding physical model template to generate the physical model of the device to be generated.

7. The method for generating a physical model in the smart park Internet of Things according to claim 6, characterized in that: Extracting the text corresponding to each feature from the text describing the device to be generated according to each feature includes: Get the current feature and the next adjacent feature; All texts between the current feature and the adjacent next feature in the text describing the device to be generated are determined as the corresponding texts, wherein all the texts include the current feature but do not include the adjacent next feature.

8. A device for generating object models in the Internet of Things in a smart park, characterized in that: include: Text acquisition module: acquires text information describing the device to be generated; Name and feature classification module: extract the name information and feature information of the device to be generated according to the text information; Object model template retrieval module: according to the name information, retrieve the corresponding object model template in a preset standard object model library; Physical model dimension determination module: determines the physical model dimension corresponding to each feature according to the feature information and a preset classification neural network model; Physical model generation module: adds each feature in the feature information to the corresponding physical model dimension of the corresponding physical model template to generate the physical model of the device to be generated.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the method for generating a physical model in the smart park Internet of Things as described in any one of claims 1 to 7 is implemented.

10. A device, characterized in that: include: A memory for storing instructions; A processor is used to execute the instructions so that the device performs operations to implement the method for generating a physical model in the smart park Internet of Things as described in any one of claims 1-7.

Images