A control method and system for electronic cigarette

Abstract

The present invention provides an electronic cigarette control method and system, the method comprising: using a generative adversarial network based on a camera monitoring video of the current environment to generate a current environment volume size and a current environment electronic cigarette smoke concentration distribution map; using a camera processing model based on the camera monitoring video of the current environment to determine the personal information of each non-smoker, the smoke concentration tolerance threshold of each non-smoker, the position of each electronic cigarette, and the direction and distance between each electronic cigarette and each non-smoker; constructing a knowledge graph; processing the knowledge graph based on a graph neural network model to determine the power of each electronic cigarette; sending the power of each electronic cigarette to a corresponding electronic cigarette, and adjusting the power of each electronic cigarette based on the power of each electronic cigarette. The method can reduce the impact of electronic cigarette smoke on non-smokers in public places.

Description

Technical Field

[0001] The present invention relates to the technical field of electronic cigarettes, and in particular to a control method and system for electronic cigarettes. Background Art

[0002] With the continuous development of modern technology, e-cigarettes, as a new tobacco substitute, have gradually entered people's lives. E-cigarettes are used in various environments, including homes, offices, and public places. However, in public places, when there are too many e-cigarettes, the smoke concentration may exceed the tolerance of the environment, causing non-smokers to feel uncomfortable and even affecting their health.

[0003] Therefore, how to reduce the impact of e-cigarette smoke on non-smokers in public places is a problem that needs to be solved at present. Summary of the invention

[0004] The main technical problem solved by the present invention is how to reduce the impact of electronic cigarette smoke on non-smokers in public places.

[0005] According to a first aspect, the present invention provides an electronic cigarette control method, comprising: determining whether the number of electronic cigarettes connected to the current environment exceeds a set threshold; if the number of electronic cigarettes connected to the current environment exceeds the set threshold, obtaining a camera surveillance video of the current environment; using a generative adversarial network based on the camera surveillance video of the current environment to generate a current environment volume size and a current environment electronic cigarette smoke concentration distribution map; based on the camera surveillance video of the current environment, using a camera processing model to determine the personal information of each non-smoker, the smoke concentration tolerance threshold of each non-smoker, the position of each electronic cigarette, and the direction and distance between each electronic cigarette and each non-smoker; constructing a knowledge graph, the knowledge graph comprising multiple nodes and multiple edges between the multiple nodes, the multiple nodes comprising multiple electronic cigarettes. A sub-cigarette node and multiple non-smoking nodes, each electronic cigarette node establishes an edge with each non-smoking node, the node characteristics of each electronic cigarette node include the location of each electronic cigarette, the node characteristics of each non-smoking node include the current environment volume, the current environment electronic cigarette smoke concentration distribution map, the personal information of each non-smoker, and the smoke concentration tolerance threshold of each non-smoker, and the characteristics of each electronic cigarette node establishing an edge with each non-smoking node include the direction and distance of each electronic cigarette from each non-smoker; the knowledge graph is processed based on the graph neural network model to determine the power of each electronic cigarette; the power of each electronic cigarette is sent to the corresponding electronic cigarette, and the power of each electronic cigarette is adjusted based on the power of each electronic cigarette.

[0006] Furthermore, the camera processing model is a recurrent neural network model.

[0007] Furthermore, the method further includes: if the number of electronic cigarettes connected to the current environment does not exceed a set threshold, determining that the environment is normal, and sending information that the environment is normal to the control terminal.

[0008] Furthermore, the generative adversarial network is trained by gradient descent method.

[0009] Furthermore, the input of the generative adversarial network is the camera monitoring video of the current environment, and the output of the generative adversarial network is the volume size of the current environment and the electronic cigarette smoke concentration distribution map of the current environment.

[0010] According to a second aspect, the present invention provides a control system for an electronic cigarette, comprising:

[0011] A judgment module is used to judge whether the number of electronic cigarettes connected to the current environment exceeds a set threshold;

[0012] An acquisition module is used to obtain the camera surveillance video of the current environment if the number of electronic cigarettes connected to the current environment exceeds a set threshold;

[0013] A generation module, used to generate a current environment volume size and a current environment electronic cigarette smoke concentration distribution map using a generative adversarial network based on the camera monitoring video of the current environment;

[0014] A camera processing module, for determining the personal information of each non-smoker, the smoke concentration tolerance threshold of each non-smoker, the position of each electronic cigarette, and the direction and distance between each electronic cigarette and each non-smoker using a camera processing model based on the camera monitoring video of the current environment;

[0015] A construction module is used to construct a knowledge graph, wherein the knowledge graph includes a plurality of nodes and a plurality of edges between the plurality of nodes, wherein the plurality of nodes include a plurality of electronic cigarette nodes and a plurality of non-smoking person nodes, wherein each electronic cigarette node establishes an edge with each non-smoking person node, wherein the node feature of each electronic cigarette node includes the location of each electronic cigarette, and the node feature of each non-smoking person node includes the current environment volume, the current environment electronic cigarette smoke concentration distribution map, the personal information of each non-smoking person, and the smoke concentration tolerance threshold of each non-smoking person, and the feature of each electronic cigarette node establishing an edge with each non-smoking person node includes the direction and distance of each electronic cigarette from each non-smoking person;

[0016] A graph neural network module, used to process the knowledge graph based on a graph neural network model to determine the power of each electronic cigarette;

[0017] The power adjustment module is used to send the power of each electronic cigarette to the corresponding electronic cigarette, and adjust the power of each electronic cigarette based on the power of each electronic cigarette.

[0018] Furthermore, the camera processing model is a recurrent neural network model.

[0019] Furthermore, the system is also used to: if the number of electronic cigarettes connected to the current environment does not exceed a set threshold, determine that the environment is normal, and send information that the environment is normal to the control terminal.

[0020] Furthermore, the generative adversarial network is trained by gradient descent method.

[0021] Furthermore, the input of the generative adversarial network is the camera monitoring video of the current environment, and the output of the generative adversarial network is the volume size of the current environment and the electronic cigarette smoke concentration distribution map of the current environment.

[0022] The present invention provides an electronic cigarette control method and system, the method comprising determining whether the number of electronic cigarettes connected to the current environment exceeds a set threshold; if the number of electronic cigarettes connected to the current environment exceeds the set threshold, obtaining a camera surveillance video of the current environment; using a generative adversarial network based on the camera surveillance video of the current environment to generate a current environment volume size and a current environment electronic cigarette smoke concentration distribution map; using a camera processing model based on the camera surveillance video of the current environment to determine the personal information of each non-smoker, the smoke concentration tolerance threshold of each non-smoker, the position of each electronic cigarette, and the direction and distance between each electronic cigarette and each non-smoker; constructing a knowledge graph, the knowledge graph comprising multiple nodes and multiple edges between the multiple nodes, the multiple nodes comprising multiple electronic cigarette nodes and multiple non-smoker nodes , each electronic cigarette node establishes an edge with each non-smoking node respectively, the node characteristics of each electronic cigarette node include the location of each electronic cigarette, the node characteristics of each non-smoking node include the current environment volume, the current environment electronic cigarette smoke concentration distribution map, each non-smoking person's personal information, and each non-smoking person's smoke concentration tolerance threshold, and the characteristics of each electronic cigarette node establishing an edge with each non-smoking node include the direction and distance of each electronic cigarette from each non-smoking person; the knowledge graph is processed based on the graph neural network model to determine the power of each electronic cigarette; the power of each electronic cigarette is sent to the corresponding electronic cigarette, and the power of each electronic cigarette is adjusted based on the power of each electronic cigarette. This method can reduce the impact of electronic cigarette smoke on non-smokers in public places. BRIEF DESCRIPTION OF THE DRAWINGS

[0023] Figure 1A schematic flow chart of a method for controlling an electronic cigarette provided by an embodiment of the present invention;

[0024] Figure 2 A schematic diagram of a control system of an electronic cigarette provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0025] In an embodiment of the present invention, there is provided Figure 1 A control method of an electronic cigarette is shown, and the control method of the electronic cigarette includes steps S1 to S7:

[0026] Step S1, determining whether the number of electronic cigarettes connected to the current environment exceeds a set threshold.

[0027] The set threshold refers to a pre-set upper limit or threshold of the number of electronic cigarettes. The system can determine whether the number of electronic cigarettes in the current environment exceeds the set threshold by monitoring the connection status of the electronic cigarette devices. As an example, the system monitors the Bluetooth connection status of the electronic cigarettes in the current environment and identifies the number of connected electronic cigarette devices.

[0028] Step S2: if the number of electronic cigarettes connected to the current environment exceeds a set threshold, a camera monitoring video of the current environment is obtained.

[0029] Camera surveillance video is a video signal captured by a camera device that can be used to monitor and record activities and situations in the current environment.

[0030] In some embodiments, if the number of electronic cigarettes connected to the current environment does not exceed a set threshold, the environment is determined to be normal, and information indicating that the environment is normal is sent to the control terminal.

[0031] Step S3, using a generative adversarial network based on the camera monitoring video of the current environment to generate a current environment volume size and a current environment electronic cigarette smoke concentration distribution map.

[0032] The generative adversarial network is trained by gradient descent method.

[0033] The input of the generative adversarial network is the camera monitoring video of the current environment, and the output of the generative adversarial network is the volume size of the current environment and the distribution map of the electronic cigarette smoke concentration in the current environment.

[0034] Generative Adversarial Network (GAN) consists of a generator and a discriminator to generate realistic data samples. The generator tries to generate samples that look similar to real data, while the discriminator tries to distinguish the generated samples from real samples.

[0035] The generator part of the generative adversarial network can learn the environmental features in the camera surveillance video, including the visual features of smoke. Through the learned environmental features, the generator can try to generate data samples similar to the real environment, including the distribution of smoke. The camera surveillance video can capture the visual information of the current environment, including the color, density, and diffusion of the smoke. The generative adversarial network will learn the visual features of the smoke in the environment and use these features to model the concentration distribution of the smoke. The generator will generate a smoke concentration distribution map that matches the actual smoke distribution in the current environment based on the input camera surveillance video.

[0036] Step S4, based on the camera surveillance video of the current environment, using the camera processing model to determine the personal information of each non-smoker, the smoke concentration tolerance threshold of each non-smoker, the position of each electronic cigarette, and the direction and distance between each electronic cigarette and each non-smoker.

[0037] The camera processing model is a recurrent neural network model. The input of the camera processing model is the camera monitoring video of the current environment, and the output of the camera processing model is the personal information of each non-smoker, the smoke concentration tolerance threshold of each non-smoker, the position of each electronic cigarette, and the direction and distance between each electronic cigarette and each non-smoker.

[0038] The recurrent neural network model includes the recurrent neural network (RNN). The recurrent neural network model can process sequence data, capture sequence information, and output results based on the correlation between the previous and next data in the sequence. By processing the camera surveillance video of the current environment in a continuous time period through the recurrent neural network model, it is possible to output features that comprehensively consider the correlation between the sequences at each time point, making the output features more accurate and comprehensive.

[0039] The camera processing model can extract personal information of people by analyzing the camera surveillance video. For example, facial recognition technology can identify the identity characteristics of each person. At the same time, object detection technology can be used to detect the location of the e-cigarette.

[0040] The personal information of each non-smoker includes age, gender, height, weight, activity level, location, etc. As an example, people of different ages may have different tolerances to smoke, and the younger or older they are, the lower their tolerance may be. Physical fitness factors such as weight and height may affect a person's physical function and tolerance to electronic smoke. The more active a person is, the stronger his tolerance to electronic cigarettes is, and the higher the smoke concentration tolerance threshold is.

[0041] The smoke concentration tolerance threshold of each non-smoker refers to the upper limit or maximum allowable value of the smoke concentration in the environment that the non-smoker can tolerate or bear. When the smoke concentration exceeds this threshold, it may have a negative impact on the health of non-smokers.

[0042] In some embodiments, the camera processing model includes an information determination layer, a smoke concentration tolerance threshold layer, and a position determination layer. The information determination layer, the smoke concentration tolerance threshold layer, and the position determination layer all include a recurrent neural network structure. The input of the information determination layer is the camera monitoring video of the current environment, the output of the information determination layer is the personal information of each non-smoker and the position of each electronic cigarette, the input of the smoke concentration tolerance threshold layer is the personal information of each non-smoker, the output of the smoke concentration tolerance threshold layer is the smoke concentration tolerance threshold of each non-smoker, the input of the position determination layer is the position of each electronic cigarette, the personal information of each non-smoker, and the output of the position determination layer is the direction and distance between each electronic cigarette and each non-smoker.

[0043] Different layers are responsible for different levels of information abstraction and processing. For example, the information determination layer is responsible for extracting the location information of people and electronic cigarettes from the surveillance video, the smoke concentration tolerance threshold layer is responsible for determining the smoke concentration tolerance threshold based on personal information, and the location determination layer is responsible for determining the relative position relationship between people and electronic cigarettes. Through such layered processing, complex information can be processed more effectively. Building multiple layers can improve the modularity of the system, process information more effectively, and improve the accuracy and efficiency of the system.

[0044] Step S5, constructing a knowledge graph, wherein the knowledge graph includes multiple nodes and multiple edges between the multiple nodes, the multiple nodes include multiple electronic cigarette nodes and multiple non-smoking nodes, each electronic cigarette node establishes an edge with each non-smoking node respectively, the node characteristics of each electronic cigarette node include the location of each electronic cigarette, the node characteristics of each non-smoking node include the current environment volume size, the current environment electronic cigarette smoke concentration distribution map, the personal information of each non-smoker, and the smoke concentration tolerance threshold of each non-smoker, and the characteristics of each electronic cigarette node establishing an edge with each non-smoking node include the direction and distance of each electronic cigarette from each non-smoking person.

[0045] The electronic cigarette node is a node representing each electronic cigarette, and the node features include the location of each electronic cigarette.

[0046] The non-smoking node is a node representing each non-smoker. The node features include the current environment volume, the current environment electronic cigarette smoke concentration distribution map, the personal information of each non-smoker, and the smoke concentration tolerance threshold of each non-smoker.

[0047] Edges represent the association between nodes. Each e-cigarette node establishes an edge with each non-smoking node. The characteristics of the edge include the direction and distance of each e-cigarette from each non-smoking node.

[0048] Step S6: Process the knowledge graph based on the graph neural network model to determine the power of each electronic cigarette.

[0049] The graph neural network model includes a graph neural network (GNN) and a fully connected layer. The graph neural network is a neural network that directly acts on a knowledge graph. The knowledge graph is a data structure consisting of two parts: nodes and edges. The input of the graph neural network model is the knowledge graph, and the output of the graph neural network model is the power of each electronic cigarette.

[0050] If the power of the e-cigarette is too high, the concentration of e-cigarette smoke in the environment may be too high. If the power of the e-cigarette is too low, smokers may not feel the pleasure of smoking, affecting their experience. Therefore, by building a knowledge graph and using a graph neural network model, the appropriate power of each e-cigarette is output after comprehensive consideration.

[0051] The knowledge graph represents the e-cigarette nodes and non-smoker nodes and the relationship between them as a graph structure. The knowledge graph integrates various information (such as the location of e-cigarettes, the size of the environmental volume, the distribution map of smoke concentration, personal information, etc.) into one structure. The characteristics of the edge include distance and direction, and the spatial position relationship between e-cigarettes and non-smokers can be considered. This information can help the model better understand the propagation path and impact range of e-cigarettes. The knowledge graph can integrate the feature information of multiple nodes, including the location of e-cigarettes, the size of the environmental volume, the distribution map of smoke concentration, personal information, etc. Integrating this multi-source information into the knowledge graph can provide more comprehensive input data, which helps the graph neural network model to learn and predict the power of e-cigarettes more accurately.

[0052] Step S7, sending the power of each electronic cigarette to the corresponding electronic cigarette, and adjusting the power of each electronic cigarette based on the power of each electronic cigarette.

[0053] The control system may refer to sending instructions for controlling the power of the electronic cigarette to the electronic cigarette device to adjust the working intensity of the heater and the amount of smoke generated. As an example, the power adjustment instruction can be sent from the control system to the electronic cigarette device via a Bluetooth or Wi-Fi connection.

[0054] Based on the same inventive concept, Figure 2 A schematic diagram of a control system of an electronic cigarette provided by an embodiment of the present invention, wherein the control system of the electronic cigarette comprises:

[0055] A judgment module 21 is used to judge whether the number of electronic cigarettes connected to the current environment exceeds a set threshold;

[0056] The acquisition module 22 is used to acquire the camera monitoring video of the current environment if the number of electronic cigarettes connected to the current environment exceeds a set threshold;

[0057] A generation module 23 is used to generate a current environment volume size and a current environment electronic cigarette smoke concentration distribution map using a generative adversarial network based on the camera monitoring video of the current environment;

[0058] A camera processing module 24 is used to determine the personal information of each non-smoker, the smoke concentration tolerance threshold of each non-smoker, the position of each electronic cigarette, and the direction and distance between each electronic cigarette and each non-smoker using a camera processing model based on the camera monitoring video of the current environment;

[0059] A construction module 25 is used to construct a knowledge graph, wherein the knowledge graph includes a plurality of nodes and a plurality of edges between the plurality of nodes, wherein the plurality of nodes include a plurality of electronic cigarette nodes and a plurality of non-smoking person nodes, wherein each electronic cigarette node establishes an edge with each non-smoking person node, wherein the node feature of each electronic cigarette node includes the location of each electronic cigarette, and the node feature of each non-smoking person node includes the current environment volume, the current environment electronic cigarette smoke concentration distribution map, the personal information of each non-smoking person, and the smoke concentration tolerance threshold of each non-smoking person, and the feature of each electronic cigarette node establishing an edge with each non-smoking person node includes the direction and distance of each electronic cigarette from each non-smoking person;

[0060] A graph neural network module 26, used to process the knowledge graph based on a graph neural network model to determine the power of each electronic cigarette;

[0061] The power adjustment module 27 is used to send the power of each electronic cigarette to the corresponding electronic cigarette, and adjust the power of each electronic cigarette based on the power of each electronic cigarette.

Claims

1. A method for controlling an electronic cigarette, characterized in that: include: Determine whether the number of electronic cigarettes connected to the current environment exceeds the set threshold; If the number of electronic cigarettes connected to the current environment exceeds the set threshold, the camera surveillance video of the current environment is obtained; Based on the camera surveillance video of the current environment, a generative adversarial network is used to generate the current environment volume size and the current environment electronic cigarette smoke concentration distribution map; Based on the camera surveillance video of the current environment, a camera processing model is used to determine the personal information of each non-smoker, the smoke concentration tolerance threshold of each non-smoker, the position of each electronic cigarette, and the direction and distance between each electronic cigarette and each non-smoker, wherein the camera processing model is a recurrent neural network model; Constructing a knowledge graph, wherein the knowledge graph includes a plurality of nodes and a plurality of edges between the plurality of nodes, wherein the plurality of nodes include a plurality of electronic cigarette nodes and a plurality of non-smoking person nodes, wherein each electronic cigarette node establishes an edge with each non-smoking person node, wherein the node feature of each electronic cigarette node includes the location of each electronic cigarette, and the node feature of each non-smoking person node includes the current environment volume size, the current environment electronic cigarette smoke concentration distribution map, the personal information of each non-smoking person, and the smoke concentration tolerance threshold of each non-smoking person, and the feature of each electronic cigarette node establishing an edge with each non-smoking person node includes the direction and distance of each electronic cigarette from each non-smoking person; Processing the knowledge graph based on a graph neural network model to determine the power of each electronic cigarette; The power of each electronic cigarette is sent to the corresponding electronic cigarette, and the power of each electronic cigarette is adjusted based on the power of each electronic cigarette.

2. The electronic cigarette control method according to claim 1, characterized in that: The method further includes: if the number of electronic cigarettes connected to the current environment does not exceed a set threshold, determining that the environment is normal, and sending information indicating that the environment is normal to the control terminal.

3. The electronic cigarette control method according to claim 1, characterized in that: The generative adversarial network is trained by gradient descent method.

4. The electronic cigarette control method according to claim 3, characterized in that: The input of the generative adversarial network is the camera monitoring video of the current environment, and the output of the generative adversarial network is the volume size of the current environment and the distribution map of the electronic cigarette smoke concentration in the current environment.

5. An electronic cigarette control system, characterized in that: include: A judgment module is used to judge whether the number of electronic cigarettes connected to the current environment exceeds a set threshold; An acquisition module is used to obtain the camera surveillance video of the current environment if the number of electronic cigarettes connected to the current environment exceeds a set threshold; A generation module, used to generate a current environment volume size and a current environment electronic cigarette smoke concentration distribution map using a generative adversarial network based on the camera monitoring video of the current environment; A camera processing module, for determining the personal information of each non-smoker, the smoke concentration tolerance threshold of each non-smoker, the position of each electronic cigarette, and the direction and distance between each electronic cigarette and each non-smoker using a camera processing model based on the camera monitoring video of the current environment, wherein the camera processing model is a recurrent neural network model; A construction module is used to construct a knowledge graph, wherein the knowledge graph includes a plurality of nodes and a plurality of edges between the plurality of nodes, wherein the plurality of nodes include a plurality of electronic cigarette nodes and a plurality of non-smoking person nodes, wherein each electronic cigarette node establishes an edge with each non-smoking person node, wherein the node feature of each electronic cigarette node includes the location of each electronic cigarette, and the node feature of each non-smoking person node includes the current environment volume, the current environment electronic cigarette smoke concentration distribution map, the personal information of each non-smoking person, and the smoke concentration tolerance threshold of each non-smoking person, and the feature of each electronic cigarette node establishing an edge with each non-smoking person node includes the direction and distance of each electronic cigarette from each non-smoking person; A graph neural network module, used to process the knowledge graph based on a graph neural network model to determine the power of each electronic cigarette; The power adjustment module is used to send the power of each electronic cigarette to the corresponding electronic cigarette, and adjust the power of each electronic cigarette based on the power of each electronic cigarette.

6. The electronic cigarette control system according to claim 5, characterized in that: The system is also used to: if the number of electronic cigarettes connected to the current environment does not exceed a set threshold, determine that the environment is normal, and send information that the environment is normal to the control terminal.

7. The electronic cigarette control system according to claim 5, characterized in that: The generative adversarial network is trained by gradient descent method.

8. The electronic cigarette control system according to claim 7, characterized in that: The input of the generative adversarial network is the camera monitoring video of the current environment, and the output of the generative adversarial network is the volume size of the current environment and the distribution map of the electronic cigarette smoke concentration in the current environment.

Images