Method and device for detecting opening and closing states of closed space and refrigerator

Abstract

The invention relates to a method and device for detecting the opening and closing state of a closed space and a refrigerator. The method comprises the following steps: acquiring gas concentration information sensed by a gas sensor in a closed space, wherein the gas concentration information comprises an electrical parameter value representing the concentration of various gases in the closed space; the method further includes detecting, by a machine learning model, an on-off state of the confined space based on the gas concentration information. In this way, the machine learning model can respond to the data sensed by the gas sensor in time, and the on-off state can be accurately determined according to the gas concentration change in the closed space. The mode can be suitable for the on-off state detection of the closed space in various scenes.

Description

Technical Field

[0001] This disclosure relates to the field of intelligent detection technology, and more specifically, to a method, apparatus, and refrigerator for detecting the on / off state of a confined space. Background Technology

[0002] Refrigerators are a common food storage appliance. In our daily use, we take food out and put it in. However, when closing the door, we usually just close it casually or forget to close it after taking things out, which damages the refrigerator's seal. This causes air to circulate between the refrigerator and the outside, resulting in the loss of cold air, which damages the food preservation environment and causes a great waste of electricity.

[0003] In addition, other enclosed spaces, such as chemical reagent storage equipment, wardrobes, warehouses, storage rooms, and chemical workshops, where gas is generated and the internal gas concentration changes due to the switch status, also require switch status detection technology. Summary of the Invention

[0004] Embodiments of this disclosure provide a method, apparatus, and refrigerator for detecting the open / closed state of a confined space.

[0005] In a first aspect of this disclosure, a method for detecting the on / off state of a confined space is provided. The method includes acquiring gas concentration information sensed by a gas sensor in the confined space, wherein the gas concentration information includes electrical parameter values ​​characterizing the concentrations of various gases within the confined space. The method further includes detecting the on / off state of the confined space based on the gas concentration information using a machine learning model.

[0006] In a second aspect of this disclosure, an apparatus for detecting the on / off state of a confined space is provided. The apparatus includes a gas concentration information acquisition module configured to acquire gas concentration information sensed by a gas sensor in the confined space, wherein the gas concentration information includes electrical parameter values ​​characterizing the concentrations of various gases within the confined space. The apparatus also includes an on / off state detection module configured to detect the on / off state of the confined space based on the gas concentration information using a machine learning model.

[0007] In a third aspect of this disclosure, an apparatus comprising a sealed space is provided. The apparatus includes a gas sensor located within the sealed space; a processor; and a memory coupled to the processor and having instructions stored thereon, which, when executed by the processor, cause the method provided according to a first aspect of this disclosure to be performed.

[0008] In a fourth aspect of this disclosure, a computer program product is provided, comprising a computer program that is executed by a processor to implement the method according to the first aspect.

[0009] In a fifth aspect of this disclosure, a machine-readable storage medium is provided. The machine-readable storage medium stores machine-executable instructions, which are executed by a processor to implement the method provided according to a first aspect of this disclosure.

[0010] It should be understood that the description in the Summary of the Invention section is not intended to limit the key or essential features of the embodiments of this disclosure, nor is it intended to restrict the scope of this disclosure. Other features of this disclosure will become readily apparent from the following description. Attached Figure Description

[0011] The above and other features, advantages, and aspects of the embodiments of this disclosure will become more apparent from the accompanying drawings and the following detailed description. In the drawings, the same or similar reference numerals denote the same or similar elements, wherein:

[0012] Figure 1 A schematic diagram of an example environment in which some embodiments of this disclosure may be implemented is shown;

[0013] Figure 2 A flowchart of a method for detecting the opening and closing state of a confined space according to some embodiments of the present disclosure is shown;

[0014] Figure 3A A schematic diagram showing the relationship between the concentration of various gases and time in a closed space under sealed conditions is presented.

[0015] Figure 3B A schematic diagram showing the relationship between the concentration of various gases and time in a closed space under non-closed conditions is presented.

[0016] Figure 4 A schematic diagram illustrates an example process of generating switch states using a machine learning model according to some embodiments of the present disclosure;

[0017] Figure 5 A schematic diagram illustrates an example process of generating switch states using a machine learning model during the training phase, according to some embodiments of the present disclosure.

[0018] Figure 6A The diagram illustrates some example processes for processing gas sensor sensing data to obtain gas concentration information according to some embodiments of the present disclosure;

[0019] Figure 6B Schematic diagrams are shown of some embodiments of processing gas sensor sensing data to obtain gas concentration information according to some embodiments of the present disclosure;

[0020] Figure 6CSchematic diagrams are shown of other embodiments of processing gas sensor sensing data to obtain gas concentration information according to some embodiments of the present disclosure;

[0021] Figure 6D Schematic diagrams are shown of other embodiments of processing gas sensor sensing data to obtain gas concentration information according to some embodiments of the present disclosure;

[0022] Figure 7A A schematic diagram is shown of another example process for processing gas sensor sensing data to obtain gas concentration information according to some embodiments of the present disclosure;

[0023] Figure 7B Schematic diagrams are shown of some embodiments of processing gas sensor sensing data to obtain gas concentration information according to some embodiments of the present disclosure;

[0024] Figure 8A Schematic diagrams are shown of other example processes for processing gas sensor sensing data to obtain gas concentration information according to some embodiments of the present disclosure;

[0025] Figure 8B Schematic diagrams are shown of some embodiments of processing gas sensor sensing data to obtain gas concentration information according to some embodiments of the present disclosure;

[0026] Figure 9 A schematic diagram illustrates an example process for generating three switching states using a machine learning model according to some embodiments of the present disclosure;

[0027] Figure 10 A block diagram of an apparatus for detecting the opening and closing state of a confined space according to some embodiments of the present disclosure is shown; and

[0028] Figure 11 A block diagram of an electronic device that can implement several embodiments of the present disclosure is shown. Detailed Implementation

[0029] Embodiments of this disclosure will now be described in more detail with reference to the accompanying drawings. While some embodiments of this disclosure are shown in the drawings, it should be understood that this disclosure can be implemented in various forms and should not be construed as limited to the embodiments set forth herein. Rather, these embodiments are provided to provide a more thorough and complete understanding of this disclosure. It should be understood that the accompanying drawings and embodiments of this disclosure are for illustrative purposes only and are not intended to limit the scope of protection of this disclosure.

[0030] In the description of embodiments of this disclosure, the term "comprising" and similar terms should be understood as open-ended inclusion, i.e., "including but not limited to". The term "based on" should be understood as "at least partially based on". The term "one embodiment" or "the embodiment" should be understood as "at least one embodiment". The terms "first", "second", etc., may refer to different or the same objects. Other explicit and implicit definitions may also be included below.

[0031] As mentioned above, detecting the open / closed status of enclosed spaces has a significant impact on the storage of food or goods within them. Related technologies utilize temperature or air pressure changes to detect the open / closed status of enclosed spaces. However, on the one hand, temperature and air pressure are easily affected by external conditions; for example, weather changes cause atmospheric pressure variations, and movement of people can lead to changes in air pressure or temperature, resulting in inaccurate readings from temperature or air pressure sensors. On the other hand, temperature and air pressure detection methods require a certain amount of time to accumulate changes before accurately detecting the open / closed status of the enclosed space. This can lead to delays in detection and response.

[0032] In other related technologies, gas concentration values ​​can be used to detect the open / closed state of a confined space. Generally, a closed state is determined by the gas concentration being within a threshold range, and an open state is determined by the gas concentration exceeding a threshold range. However, the change in gas concentration within a confined space is a complex process, making it impossible to use threshold values ​​to determine the concentrations of multiple gases, or even to determine the concentration at a specific state during the change process.

[0033] To address this, embodiments of this disclosure propose a scheme for detecting the on / off state of a confined space. In embodiments of this disclosure, a machine learning model can be trained using electrical parameter values ​​representing gas pressure concentration and corresponding on / off states from a dataset of confined space usage. Then, based on the electrical parameter values ​​representing gas concentration sensed by a gas sensor within the confined space, the trained machine learning model generates the on / off state of the confined space.

[0034] In this way, the machine learning model can learn the on / off states corresponding to complex gas concentration changes under various usage scenarios from the usage dataset. This allows the machine learning model to adapt to changes in refrigerator usage patterns and environment over time. It can also respond promptly to data sensed by the gas sensor and extract features from the sensed data to distinguish between normal fluctuations in gas concentration and fluctuations specifically caused by door opening, thus enabling rapid and accurate detection of the on / off state.

[0035] Figure 1 A schematic diagram of an example environment 100 in which some embodiments of this disclosure may be implemented is shown. For example... Figure 1As shown, environment 100 includes a closed space 102. Here, "closed space 102" refers to a space that can be kept closed, and its closed state can be adjusted or changed, for example, by adjusting the on / off state of a device providing the closed space 102. The closed space 102 can be the internal space of a device where gas is generated and its internal gas concentration changes due to the on / off state. For example, the closed space 102 can be provided by a refrigerator, medicine storage cabinet, wardrobe, chemical reagent storage equipment, warehouse, storage room, and chemical workshop, etc. Figure 1 As shown, the enclosed space 102 of environment 100 includes food or articles 104, which may generate volatile gases. The food or articles can be one or more. For example, a refrigerator typically stores various foods, including vegetables, fruits, meats, and seafood. For example, a medicine storage cabinet stores one or more medicines (such as probiotics). For example, a wardrobe stores clothing of various materials. For example, a chemical reagent storage device stores one or more chemical reagents that may be stored together. For example, a warehouse or storage room stores items such as oil tanks or gas cylinders. For example, a chemical workshop stores one or more chemical substances.

[0036] like Figure 1 As shown, the environment 100 also includes a gas sensor 106 disposed within the enclosed space 102, the gas sensor 106 being used to sense changes in gas concentration within the enclosed space 102. Figure 1 As shown, the environment 100 also includes a controller 108, which is communicatively connected to the gas sensor 106. The controller 108 receives data sensed by the gas sensor 106, processes it to obtain gas concentration information, and detects the on / off state of the enclosed space 102 based on the gas concentration information. The controller 108 can be located on the enclosed space side, such as being installed within the enclosed space or integrated into a device within the enclosed space. The controller 108 can also be located remotely, such as at a remote management terminal. Accordingly, the communication between the gas sensor and the controller can be configured with different communication methods based on the location of the devices, such as wired or wireless communication.

[0037] like Figure 1As shown, environment 100 also includes terminal devices 110 located outside the enclosed space 102, such as mobile phones, tablets, and desktop computers. When controller 108 detects that the enclosed space 102 is closed, it waits for the next detection data to perform on / off status detection. When controller 108 detects that the enclosed space 102 is open, it needs to send an alarm message or notification to terminal device 110 to alert the user or administrator. For example, when the controller detects that a refrigerator door is open for a period of time, it sends a reminder to the user's terminal device via SMS or email. The user closes the refrigerator door promptly after receiving the message. This can avoid energy waste and ensure the freshness of food. For example, when the controller detects that a warehouse is open for a period of time, it sends a reminder to the administrator via SMS or email. The administrator closes the warehouse promptly after receiving the message and, if necessary, takes relevant emergency measures to prevent accidents such as explosions. This can avoid potential dangers in a timely manner. Figure 1 As shown, in environment 100, an alarm 112 can also be installed in the enclosed space to issue an alarm on the enclosed space side when the controller detects that the enclosed space is in an open state. The alarm 112 can be an audible alarm, an audible and visual alarm, or a light-emitting alarm, etc.

[0038] like Figure 1 As shown, a ventilation device 114 is also provided within the enclosed space 102 in environment 100. In some scenarios, the ventilation device 114 is used to periodically regulate the gas flow within the enclosed space. The ventilation device 114 can be a ventilation system including a fan. For example, a fan is needed in a refrigerator to regulate the air in the refrigerator's storage compartment to ensure the freshness of food. For example, a ventilation system is needed in a warehouse to regulate the air within the warehouse to prevent excessive accumulation of gases emitted by items, which could lead to poisoning or explosions.

[0039] The gas concentration in a confined space is affected by the ventilation system. Therefore, when applying the embodiments of this disclosure to scenarios with ventilation systems, the factors causing changes in gas concentration in the confined space due to the ventilation system must also be considered. These factors will also need to be considered when using machine learning models for subsequent learning.

[0040] In this way, gas concentration data can be collected to cover as many usage scenarios as possible. The machine learning model, based on gas concentration data that realistically reflects changes in gas concentration within a confined space and the on / off states during use, can learn the on / off states corresponding to complex gas concentration changes in various usage scenarios. By utilizing machine learning, the on / off state detection technology for confined spaces becomes more robust and reliable. During the inference phase, the trained machine learning model can accurately determine the on / off states corresponding to different gas concentration changes.

[0041] It should be understood that the architecture and functionality in example environment 100 are described for illustrative purposes only and do not imply any limitation on the scope of this disclosure. Embodiments of this disclosure can also be applied to other environments with different architectures and / or functionalities.

[0042] Figure 2 A flowchart of a method 200 for detecting the opening / closing state of a confined space according to some embodiments of the present disclosure is shown. Method 200 may, for example, be derived from... Figure 1 The controller 108 in the environment 100 shown executes. For example... Figure 2 As shown in block 202, method 200 can acquire gas concentration information sensed by a gas sensor in a confined space, wherein the gas concentration information includes electrical parameter values ​​characterizing the concentrations of various gases in the confined space. In some embodiments, the gas sensor can be a volatile organic compound (VOC) sensor used to detect volatile organic compounds. When VOC molecules enter the sensor, they undergo a chemical reaction or adsorption with the adsorbent material at a certain reaction temperature, causing changes in the sensor's electrical parameters (such as resistance, current, impedance, and voltage). These changes are proportional to the VOC concentration, and the VOC concentration in the air can be determined by measuring the changes in electrical parameters. This reaction temperature is a built-in temperature of the gas sensor, which is set according to the temperature sensitivity of different reacting gases. For example, hydrogen requires a higher temperature to react. Ammonia, for example, requires a lower temperature to react. The gas sensor reacts according to the set temperature when sensing gas concentration information. When the gas sensor has multiple reaction temperatures, the reactions are performed sequentially at different reaction temperatures. Gas sensors can have a built-in temperature control module to control the reaction temperature of the gas sensor to change according to a preset pattern in order to sense different gases.

[0043] Alternatively, some gas sensors can output electrical signals and, using analog-to-digital converters, process the signals to obtain electrical parameters characterizing the concentrations of various gases in a confined space, such as resistance, current, impedance, and voltage. Some gas sensors integrate analog-to-digital converters to output electrical parameters characterizing the concentrations of various gases in a confined space.

[0044] In such Figure 1 In the environment 100 shown, the controller 108 acquires data sensed by the gas sensor 106, and obtains gas concentration information through data processing. The gas concentration information includes electrical parameter values ​​that characterize the concentration of various gases in the enclosed space 102.

[0045] In box 204, method 200 can use a machine learning model to detect the open / closed state of a confined space based on gas concentration information. For example, the machine learning model can be a trained deep learning model, a neural network, or a support vector machine. The machine learning model can learn the open / closed state of the confined space under different gas concentration changes, including a closed or open state. After training, the machine learning model can be used to detect the open / closed state of the confined space in a timely and accurate manner based on the gas concentration information sensed by the gas sensor. Furthermore, it can issue an alarm for an open / closed state and promptly adjust the confined space from an open / closed state to a closed state.

[0046] In this way, the controller 108 uses a machine learning model for switch status detection, making the detection results robust and reliable. The controller 108 can use the machine learning model to determine the switch status based on the gas concentration information sensed by the gas sensor. In this way, the switch status of a confined space can be accurately identified from complex gas concentration changes, avoiding misidentification or even false alarms.

[0047] Figure 3A This diagram illustrates the relationship between the concentrations of various gases in a sealed space and time. The horizontal axis represents time, and the vertical axis represents any one of the electrical parameters characterizing gas concentration, such as resistance, reactance, voltage, and current. Curve 302 represents the concentration change curve in the sealed space. Curve 302 exhibits a stable trend over continuous time.

[0048] Figure 3B This diagram illustrates the relationship between the concentrations of various gases in a sealed space and time when the sealed space is in an open state. The horizontal axis represents time, and the vertical axis represents electrical parameters characterizing gas concentration, such as resistance, reactance, voltage, and current. Curve 304 represents the concentration change curve in the open state of the sealed space. When a sealed space is switched from a sealed to an open state due to a switching operation and remains open for a certain period, the entry of external air into the sealed space causes a change in gas concentration, disrupting the stable gas concentration change observed in the original sealed state, as shown by curve 304. The open state described herein can include either an unclosed state or a partially closed state. An unclosed state refers to the sealed space being open during an opening operation. A partially closed state refers to the state where the sealed space fails to close effectively during a closing operation.

[0049] Figure 4A schematic diagram of an example process 400 for generating a switch state using a machine learning model according to some embodiments of the present disclosure is shown. In process 400, gas concentration information 402 is input into a trained machine learning model 404, which generates a switch state 406 for a confined space.

[0050] Machine learning model 404 deployed at Figure 1 On the controller 108 in the environment 100 shown. When the controller 108 receives data sensed by the gas sensor, processes the data to obtain gas concentration information 402, and then generates a relevant switch state 406 based on the gas concentration information 402. For example, when the gas concentration information 402 is in a certain state... Figure 3A The machine learning model, based on the gas concentration information 402 and the changing trend of curve 302, generates a switch state 406 that is off. For example, when the gas concentration information 402 shows... Figure 3B Based on the gas concentration information 402, the machine learning model generates a switch state 406 that is either a switch operation state or an on state, according to the changing trend of curve 304.

[0051] During the training phase, the machine learning model 404 needs to be trained based on a training set containing gas concentration information and corresponding switching states. The machine learning model 404 generates predicted switching states based on the gas concentration information in the training set. The model is then trained using a loss function by comparing the predicted switching states with the switching states in the training set.

[0052] Figure 5 A schematic diagram of an example process 500 for generating switch states using a machine learning model during the training phase, according to some embodiments of the present disclosure, is shown. Figure 5 As shown, the confined space usage dataset 502 may include data sensed by gas sensors in similar confined spaces under different usage conditions.

[0053] In some examples, gas sensors detect gas concentration data inside the refrigerator, and the enclosed space usage dataset 502 may include data sensed by gas sensors under different usage conditions. This data is collected using different types of gas sensors (i.e., sensors that detect different types of gases) based on the usage of different types of refrigerators (e.g., single-door and double-door refrigerators). During the data collection process, different foods are placed in each refrigerator, and the food storage conditions vary at different times due to user habits. Furthermore, the ambient temperature and humidity inside the refrigerator will differ depending on the refrigerator's environmental settings, food types, and user habits, and the data sensed by the gas sensors is collected under various ambient temperatures and humidity levels. During the data collection process, different users open and close the refrigerator according to different usage frequencies and speeds. These opening and closing operations include opening (when the refrigerator is not sealed) and closing. When performing a closing operation, the refrigerator may be effectively closed (when the refrigerator is sealed) or not effectively closed (when the refrigerator is not sealed).

[0054] In other examples, gas sensors detect gas concentration data within the warehouse, and the confined space usage dataset 502 may include data from gas sensors detecting gas concentrations in the warehouse under different usage conditions. The collection of this data aims to cover as many usage scenarios as possible within the warehouse environment.

[0055] Depending on the usage scenario, for enclosed spaces within the same scenario, data from gas sensors should be collected to cover various usage conditions under that scenario. In this way, the machine learning model can learn, for different usage scenarios, the changes in gas concentration within the enclosed space under the influence of multiple variables and their corresponding on / off states.

[0056] In process 500, machine learning model 510 determines training dataset 512 from the usage dataset 502 of the confined space. Training dataset 512 includes sample gas concentration information 512-1 and corresponding sample switch states 512-2. Optionally, data for a certain detection period can be obtained from the usage dataset 502 and used as training dataset 512. The detection period can be determined based on what accurately reflects changes in gas concentration within the confined space; it can be one sampling period of the gas sensor or multiple sampling periods of the gas sensor. The sampling period of the gas sensor is the period during which the gas sensor detects the chemical reaction that occurs when gas enters the confined space. At different times within this period, the gas sensor can detect changes in electrical parameters characterizing gas concentration within the confined space.

[0057] In some implementations, during use in a confined space, sample gas concentration information 512-1 and corresponding sample on / off states 512-2 are acquired by the gas sensor under different ambient temperatures and humidity levels within the confined space. The sample gas concentration information 512-1 and corresponding sample on / off states 512-2 during use in the confined space are defined as the confined space usage dataset 502. A training dataset 512 is determined from the usage dataset 502 based on the detection cycle.

[0058] In other implementations, when a ventilation device is used in the application scenario, such as a fan installed inside a refrigerator, the data in dataset 502 includes sample gas concentration information 512-1 sensed by the gas sensor under different ambient temperatures and humidity levels in a confined space when the fan is not running, and the corresponding sample on / off states 512-2; and sample gas concentration information 512-1 sensed by the gas sensor under different ambient temperatures and humidity levels in a confined space when the fan is running, and the corresponding sample on / off states 512-2. By collecting data on gas concentration changes in a confined space when the fan is operating and not operating, the influence of the fan on gas concentration can be taken into account, enabling the training of a machine learning model that is more consistent with the actual application scenario.

[0059] like Figure 5 As shown, the generator 514 of the machine learning model 510 is trained based on sample gas concentration information 512-1 and sample switch states 512-2. The generator 514 can generate a predicted switch state 516 based on the sample gas concentration information 512-1. A generator loss 518 is used to determine the loss between the predicted switch state 516 and the sample switch state 512 corresponding to the sample gas concentration information 512-1. The loss determination can utilize a distance loss function. The generator loss 518 can be a large value, and then it can be backpropagated to the generator 514 to guide the optimization of the generator 514's parameters. This training process can be performed iteratively until the generator 514 can generate a more accurate predicted switch state 516. After the training process is complete, the machine learning model 510 can output the switch state 520 of the enclosed space.

[0060] Through training, machine learning models can learn the on / off states corresponding to different concentration changes under near-real-world usage conditions. Machine learning models tailored to different types of enclosed spaces can be trained to suit their specific usage scenarios. This allows for accurate detection of the on / off states of enclosed spaces. Furthermore, based on the characteristics of the scenario, alarms can be triggered when an enclosed space is not fully enclosed, enabling timely handling of such situations.

[0061] Figure 6AA schematic diagram of an example process 600A for processing gas sensor sensing data to obtain gas concentration information according to some embodiments of the present disclosure is shown. In process 600A, gas sensor 602 sends sensing data 604 characterizing gas concentration to data processing module 606 of controller. Data processing module 606 of controller processes the data characterizing gas concentration and outputs gas concentration information 608 to be fed into a machine learning model.

[0062] In some examples, the data collected by the gas sensor is input to the controller in the form of an electrical signal. Before entering the controller for processing, an existing analog-to-digital converter circuit can be configured to convert the electrical signal into an electrical parameter value. In other examples, the data collected by the gas sensor is input to the controller in the form of an electrical signal. After entering the controller, the analog-to-digital converter circuit within the controller's data processing module 606 converts the electrical signal into an electrical parameter value before proceeding to subsequent processing. In still other examples, the data collected by the gas sensor is input to the controller as an electrical parameter value.

[0063] Figure 6BSchematic diagrams are shown illustrating some embodiments of processing gas sensor sensing data to obtain gas concentration information according to some embodiments of the present disclosure. In process 600B, an example is given where the gas sensor outputs electrical parameter values ​​to the controller. When a gas sensor is installed in a sealed space, when the gas sensor collects an electrical parameter value 620 at a certain moment, it sends that electrical parameter value 620 to the data processing module 606 of the controller. In the diagram, ti represents the acquisition time of the electrical parameter value ci, and i is a natural number. The receiving unit 622 in the data processing module 606 receives the electrical parameter values ​​620 at different times sent by the gas sensor. The receiving unit 622 can receive a set of electrical parameter values ​​624 sensed by the gas sensor at different times. Then, the receiving unit 622 sends the received set of electrical parameter values ​​624 to the first processing unit 626. The first processing unit 626 cleans the data and obtains a set of electrical parameter values ​​628 within the detection period. Data cleaning includes denoising and normalization of the data. In some examples, when the detection period is 11 t, electrical parameter values ​​for 11 consecutive t are obtained from n sets of electrical parameter values ​​624 (i.e., a set of electrical parameter values ​​628 within the detection period in the figure). Then, the first processing unit 626 sends the set of electrical parameter values ​​628 within the detection period to the second processing unit 630. The second processing unit, in chronological order, converts the set of electrical parameter values ​​sensed at different times (i.e., the set of electrical parameter values ​​628 within the detection period in the figure) into a matrix 632 representing gas concentration information. This detection period includes a sampling period, during which the gas sensor reacts at a preset temperature. The temperature control module built into the gas sensor can control it to maintain the same reaction temperature at the multiple times t.

[0064] In some implementations, the detection cycle may include one incomplete sampling cycle, multiple incomplete sampling cycles, or multiple complete sampling cycles. The detection cycle can be set according to the detection requirements.

[0065] Optionally, each row in matrix 632 represents an electrical parameter value at a given time. Multiple electrical parameter values ​​at various times are arranged in chronological order as multiple rows in matrix 632, forming... Figure 6B The matrix 632 shown represents gas concentration information. This embodiment is illustrated only as an example.

[0066] As another example, as described above in conjunction with 628, a detection cycle has multiple data acquisition moments. The electrical parameter values ​​at multiple moments can be arranged into a one-dimensional vector in chronological order.

[0067] The gas concentration information obtained from process 600B can be entered. Figure 5 The pre-trained machine learning model 510 generates the opening and closing states of the sealed space. Correspondingly, during training based on the training set, the sample gas concentration information 512-1 input to the machine learning model can also be used... Figure 6B The process shown yields gas concentration information.

[0068] Using the above method, the machine learning model can learn the on / off state of a confined space corresponding to the changing gas concentration over time at a given reaction temperature. During the inference phase, the gas concentration information in matrix form is input into the machine learning model to generate the on / off states. These on / off states are determined based on the impact of time on gas concentration changes. The machine learning model can detect the on / off state of the confined space from the perspective of reaction time, and the determined on / off states more closely match the on / off states corresponding to different reaction effects (i.e., gas concentrations produced at different reaction degrees) in actual chemical reactions over time.

[0069] Figure 6CSchematic diagrams are shown illustrating other embodiments of processing gas sensor sensing data to obtain gas concentration information according to some embodiments of the present disclosure. In process 600C, an example is given where the gas sensor outputs electrical parameter values ​​to the controller. When a gas sensor is installed in a sealed space, when the gas sensor collects an electrical parameter value 640 at a certain moment, it sends that electrical parameter value 640 to the data processing module 606 of the controller. In the diagram, ti represents the collection time of the electrical parameter value ci at the reaction temperature Ti, where i is a natural number. The receiving unit 642 in the data processing module 606 receives the electrical parameter values ​​640 at different times and different reaction temperatures sent by the gas sensor. The receiving unit 642 can receive a set of electrical parameter values ​​644 sensed by the gas sensor at different times and different reaction temperatures. Then, the receiving unit 642 sends the received set of electrical parameter values ​​644 to the first processing unit 646. The first processing unit 646 cleans the data and obtains a set of electrical parameter values ​​648 within the detection period. Data cleaning includes denoising and normalization of the data. In some examples, when the detection period is 11 t, the electrical parameter values ​​for 11 consecutive t are obtained from n sets of electrical parameter values ​​644 (i.e., a set of electrical parameter values ​​648 within the detection period in the figure). Then, the first processing unit 646 sends the set of electrical parameter values ​​648 within the detection period to the second processing unit 650. The second processing unit 650, in chronological order, converts the set of electrical parameter values ​​sensed at different times and temperatures (i.e., a set of electrical parameter values ​​648 within the detection period in the figure) into a matrix 652 representing gas concentration information. This detection period includes a sampling period, during which the reaction temperature of the gas sensor changes according to a preset pattern, for example, rising from a first reaction temperature to a second reaction temperature at preset temperature intervals, and then falling back to the first reaction temperature at preset temperature intervals, forming a cycle of reaction temperature change. Considering that different times have different effects on the chemical reaction at the same reaction temperature, data from the gas sensor can be collected sequentially at multiple times t for the same reaction temperature. The temperature control module built into the gas sensor can control it to maintain the same reaction temperature at these multiple times t.

[0070] In some implementations, the detection cycle may include one incomplete sampling cycle, multiple incomplete sampling cycles, or multiple complete sampling cycles. The detection cycle can be set according to the detection requirements.

[0071] Optionally, each row in matrix 652 represents an electrical parameter value at different reaction temperatures at a given time. Multiple electrical parameter values ​​at different reaction temperatures at multiple times are arranged in chronological order as multiple rows in matrix 652, forming... Figure 6CThe matrix 652 shown represents gas concentration information. This embodiment is illustrated only as an example. Furthermore, considering that different times have different effects on the chemical reaction at the same reaction temperature, the matrix 652 representing gas concentration information may also include electrical parameter values ​​at the same reaction temperature at different times. That is, different rows in matrix 652 may correspond to the same reaction temperature. For example, multiple electrical parameter values ​​c8, c9, c10 may be collected at multiple times t8, t9, t10, with the gas sensor maintained at the same reaction temperature (T8 = T9 = T10).

[0072] As another example, as described above in conjunction with 648, a detection cycle has multiple data collection times, and the electrical parameter values ​​at multiple times can be arranged into a one-dimensional vector in chronological order.

[0073] The gas concentration information obtained from process 600C can be entered. Figure 5 The pre-trained machine learning model 510 generates the opening and closing states of the sealed space. Correspondingly, during training based on the training set, the sample gas concentration information 512-1 input to the machine learning model can also be used... Figure 6C The process shown yields gas concentration information.

[0074] Using the above method, the machine learning model can learn the on / off state of a confined space corresponding to the changing gas concentration over time at different reaction temperatures. During the inference phase, inputting the matrix-style gas concentration information into the machine learning model quickly generates relatively accurate on / off states. This machine learning model can detect the on / off state of a confined space from two dimensions: reaction time and reaction temperature. The determined on / off states more closely match the on / off states corresponding to different reaction effects (i.e., gas concentrations produced at different reaction degrees) resulting from changes in actual chemical reactions over time and reaction temperature.

[0075] Enclosed spaces often contain various gases, and different gases exhibit varying sensitivities to reactions at different temperatures. Some gases are sensitive at high temperatures, such as hydrogen and methane, while others, such as ammonia, are sensitive at low temperatures. Therefore, collecting data at multiple temperatures, and further collecting data at the same temperature over different time periods, can provide a more comprehensive reflection of gas concentration changes within an enclosed space.

[0076] Figure 6DSchematic diagrams are shown illustrating other embodiments of processing gas sensor sensing data to obtain gas concentration information according to some embodiments of the present disclosure. In process 600D, an example is given where the gas sensor outputs electrical parameter values ​​to the controller. When a gas sensor is installed in a sealed space, when the gas sensor collects an electrical parameter value 640 at a certain moment, it sends that electrical parameter value 640 to the data processing module 606 of the controller. In the diagram, ti represents the collection time of the electrical parameter value ci at the reaction temperature Ti, where i is a natural number. The receiving unit 662 in the data processing module 606 receives the electrical parameter values ​​660 at different times and different reaction temperatures sent by the gas sensor. The receiving unit 662 can receive a set of electrical parameter values ​​664 sensed by the gas sensor at different times and different reaction temperatures. Then, the receiving unit 662 sends the received set of electrical parameter values ​​664 to the first processing unit 666. The first processing unit 666 cleans the data and obtains a set of electrical parameter values ​​668 within the detection period. Data cleaning includes denoising and normalization of the data. In some examples, when the detection period includes 6 sampling periods, and each sampling period includes 10 t, data for 6 consecutive sampling periods is obtained from n sets of electrical parameter values ​​664 using a sliding window method. The sliding window size can be determined based on the time required for the gas sensor to complete the reaction at different reaction temperatures. In the figure, the sliding window size is 10t. Sliding window one is used to obtain 10 consecutive t of electrical parameter values ​​from the n electrical parameter values, i.e., one set of electrical parameter values ​​for the first sampling period. Sliding window two is used to obtain the next 10 consecutive t of electrical parameter values ​​from the n electrical parameter values, i.e., one set of electrical parameter values ​​for the second sampling period. This process continues until 6 sets of electrical parameter values ​​for 6 sampling periods are obtained (e.g., ...). Figure 6D (668 in the figure). Then, the first processing unit 666 sends multiple sets of electrical parameter values ​​668 within the detection period to the second processing unit 670. The second processing unit 670 converts a set of electrical parameter values ​​sensed at different times and temperatures (i.e., multiple sets of electrical parameter values ​​668 within the detection period in the figure) into a matrix 672 representing gas concentration information in chronological order.

[0077] Optionally, each row in matrix 672 represents the electrical parameter value at different reaction temperatures at a given time. The electrical parameter values ​​at multiple times at different reaction temperatures are arranged in chronological order as multiple rows in matrix 672, forming... Figure 6DThe matrix 672 shown represents gas concentration information. This embodiment is illustrated only as an example. Furthermore, it is considered that different times have different effects on the chemical reaction at the same reaction temperature. The matrix 672 representing gas concentration information may also include electrical parameter values ​​at the same reaction temperature at different times; that is, different rows in matrix 672 may correspond to the same reaction temperature. For example, multiple electrical parameter values ​​c9, c10, c11 may be collected at multiple times t9, t10, t11, with the gas sensor maintained at the same reaction temperature (T9 = T10 = T11).

[0078] In other embodiments, each row of the matrix (not shown) represents the electrical parameter values ​​at multiple reaction temperatures during a sampling period. A sampling period has multiple data acquisition moments, and multiple sampling periods constitute a detection period. The electrical parameter values ​​at multiple moments within a sampling period are arranged chronologically within the same row of the matrix. Another implementation involves arranging multiple sampling periods chronologically as different rows of the matrix to form a matrix representing gas concentration information.

[0079] Based on the gas concentration information obtained from process 600D, you can input... Figure 5 The pre-trained machine learning model 510 generates the opening and closing states of the sealed space. Correspondingly, during training based on the training set, the sample gas concentration information 512-1 input to the machine learning model can also be used... Figure 6D The process shown yields gas concentration information.

[0080] Using the above method, the machine learning model can learn the on / off state of a confined space corresponding to the changing gas concentration over time at different reaction temperatures. During the inference phase, inputting the matrix-style gas concentration information into the machine learning model quickly generates relatively accurate on / off states. This machine learning model can detect the on / off state of the confined space from three dimensions: reaction time, reaction temperature, and sampling period. The determined on / off states more closely match the actual on / off states corresponding to different reaction effects (referring to the gas concentrations produced at different reaction degrees) resulting from changes in reaction time and temperature in chemical reactions.

[0081] Enclosed spaces often contain various gases, and different gases exhibit varying sensitivities to reactions at different temperatures. Some gases, such as hydrogen and methane, are sensitive at high temperatures, while others, like ammonia, are sensitive at low temperatures. Therefore, collecting data at multiple temperatures, and at the same temperature for different time periods, including data from multiple sampling cycles, can comprehensively reflect the objective changes in gas concentration within an enclosed space over consecutive sampling periods.

[0082] Apart from Figure 6B , Figure 6C , Figure 6D In addition to the illustrated process, the gas concentration information in this embodiment of the disclosure can also be represented as a matrix including electrical parameter values ​​at a certain reaction temperature under multiple sampling periods. This embodiment takes into account the continuity of data changes of the gas sensor under multiple sampling periods, and the switching state corresponding to the gas concentration changes under multiple sampling periods can be learned through a machine learning model.

[0083] Figure 7A A schematic diagram of another example process 700A for processing gas sensor sensing data to obtain gas concentration information according to some embodiments of the present disclosure is shown. Multiple gas sensors are arranged in a confined space, including, but not limited to, four gas sensors. The four gas sensors can be of the same type, i.e., gas sensors used to detect the same type of gas. Alternatively, the four gas sensors can be partially of the same type and partially of different types; for example, two gas sensors are used to sense the gas concentration information of gas A, and the other two are used to sense the gas concentration information of gas B. The four gas sensors can also be of different types, sensing the gas concentration information of gases A, B, C, and D respectively. In a scenario, multiple gas sensors can be distributed individually within the confined space, or multiple gas sensors can be integrated into one or more sensing modules, with the sensing modules placed within the confined space. Embodiments of the present disclosure can configure a certain number of gas sensors according to the needs of the scenario, and can also configure one or more types of gas sensors according to the types of gases present in the confined space of the scenario. Here, four gas sensors are used as an example for explanation.

[0084] In process 700A, gas sensors 702-1, 702-2, 702-3, and 702-4 send the sensed gas concentration data 704-1, 704-2, 704-3, and 704-4 to the controller's data processing module 706. The controller's data processing module 706 processes the gas concentration data and outputs gas concentration information 708, which is then fed into the machine learning model.

[0085] In some examples, data collected by the gas sensor is input to the controller as an electrical signal. Before entering the controller for processing, an existing analog-to-digital converter (ADC) can be configured to convert the electrical signal into electrical parameter values. In other examples, the data collected by the gas sensor is input to the controller as an electrical signal. After entering the controller, the ADC within the controller converts the electrical signal into electrical parameter values ​​before proceeding with subsequent processing. In still other examples, the data collected by the gas sensor is input to the controller as electrical parameter values.

[0086] In some implementations, the gas concentration information obtained by processing the sensing data of each gas sensor can be based on... Figure 6B , Figure 6C ,or Figure 6D The example process involves processing the data from each gas sensor to obtain a matrix representing the gas concentration. This matrix includes electrical parameter values ​​at different times within the detection period, or electrical parameter values ​​at different times and temperatures within the detection period, or electrical parameter values ​​at different times and temperatures across multiple sampling periods. The matrix from each gas sensor is then input into a machine learning model. Each gas sensor can collect data for different types of gases.

[0087] In some implementations, the controller needs to normalize the gas concentration information fed into the machine learning model. This can be done by normalizing data from different gas sensors, including normalizing different types of electrical parameter values ​​(e.g., some gas sensors output resistance values, while others output current values) and normalizing electrical parameter values ​​with different data formats (e.g., some gas sensors output data containing decimal points, while others output integers). In examples using neural network modeling, a batch normalization layer can be added to the network to normalize the data from different gas sensors. In other examples, normalization can be performed before feeding the data into the model using a normalization formula.

[0088] Using the above method, the machine learning model can learn the on / off states of a confined space corresponding to different types of gas sensors and gas concentrations under various conditions. During the inference phase, inputting the matrix-style gas concentration information into the machine learning model allows for the rapid generation of relatively accurate on / off states.

[0089] Figure 7BSchematic diagrams are shown illustrating some embodiments of processing gas sensor sensing data to obtain gas concentration information according to some embodiments of the present disclosure. In process 700B, an example is given where the gas sensor outputs electrical parameter values ​​to the controller. When four gas sensors are installed in a confined space, when the gas sensors collect electrical parameter values ​​740-1, 740-2, 740-3, and 740-4 at a certain moment, they send these values ​​to the data processing module 706 of the controller. In the diagram, ti represents the acquisition time of the electrical parameter value ci at the reaction temperature Ti, where i is a natural number. The receiving unit 742 in the data processing module 706 receives the electrical parameter values ​​740-1, 740-2, 740-3, and 740-4 at different times and different reaction temperatures sent by the gas sensors. The receiving unit 742 can receive a set of electrical parameter values ​​744-1, 744-2, 744-3, and 744-4 sensed by the gas sensor at different times and under different reaction temperatures. Then, the receiving unit 662 sends the received set of electrical parameter values ​​740-1, 740-2, 740-3, and 740-4 from each gas sensor to the first processing unit 746. The first processing unit 746 cleans the data from each gas sensor and extracts multiple sets of electrical parameter values ​​748-1, 748-2, 748-3, and 748-4 within the detection period. Data cleaning includes denoising and normalization. In some examples, when the detection period includes 6 sampling periods, and each sampling period includes 10 t, data from 6 consecutive sampling periods is obtained from n sets of electrical parameter values ​​using a sliding window method. The specific process can be found in [reference needed]. Figure 6D The process is shown in the figure. Afterwards, the first processing unit 746 sends multiple sets of electrical parameter values ​​748-1, 748-2, 748-3, and 748-4 within the detection period to the second processing unit 750. The second processing unit 750, in chronological order, converts the multiple sets of electrical parameter values ​​sensed at different times and reaction temperatures (i.e., the multiple sets of electrical parameter values ​​748-1, 748-2, 748-3, and 748-4 within the detection period in the figure) into matrices 752-1, 752-2, 752-3, and 752-4 representing gas concentration information.

[0090] Optionally, each row in matrices 752-1, 752-2, 752-3, and 752-4 represents the electrical parameter value at different reaction temperatures at a given time. The electrical parameter values ​​at multiple times and different reaction temperatures are arranged in chronological order as multiple rows in matrices 752-1, 752-2, 752-3, and 752-4, forming... Figure 7BThe matrices 752-1, 752-2, 752-3, and 752-4 shown represent gas concentration information. This embodiment is illustrated only as an example. Furthermore, it is considered that different times at the same reaction temperature have different effects on the chemical reaction. The matrices 752-1, 752-2, 752-3, and 752-4 containing gas concentration information include electrical parameter values ​​at the same reaction temperature at different times.

[0091] Optionally, each row in matrices 752-1, 752-2, 752-3, and 752-4 represents the electrical parameter values ​​at different reaction temperatures under one sampling period. One sampling period has multiple data acquisition moments, and multiple sampling periods constitute one detection period. The electrical parameter values ​​at multiple moments within a sampling period are arranged in chronological order within the same row of matrices 752-1, 752-2, 752-3, or 752-4 within another implementation. Multiple sampling periods are arranged in chronological order as different rows of the corresponding matrices to form matrices 752-1, 752-2, 752-3, or 752-4 representing gas concentration information.

[0092] Based on the gas concentration information obtained from process 700B, you can input... Figure 5 The pre-trained machine learning model 510 generates the opening and closing states of the sealed space. Correspondingly, during training based on the training set, the sample gas concentration information 512-1 input to the machine learning model can also be used... Figure 7B The process shown yields gas concentration information.

[0093] Using the above method, the machine learning model can learn the on / off state of a confined space corresponding to the changing gas concentration over time at different reaction temperatures. During the inference phase, inputting the matrix-style gas concentration information into the machine learning model allows for the rapid generation of relatively accurate on / off states. The machine learning model can detect the on / off state of the confined space from four dimensions: time, reaction temperature, sampling period, and channels (referring to gas sensors detecting different types of gases). The determined on / off states better reflect the actual chemical reactions of different gases, resulting in different reaction effects (i.e., gas concentrations produced at different reaction degrees) under varying reaction times and temperatures.

[0094] Enclosed spaces often contain various gases, and different gases exhibit varying sensitivities to reactions at different temperatures. Some gases are sensitive at high temperatures, such as hydrogen and methane, while others, like ammonia, are sensitive at low temperatures. Therefore, by collecting data from multiple gas sensors of different types at various reaction temperatures, and at the same reaction temperature for different time periods, including data from multiple sampling periods, a more comprehensive reflection of the objective changes in gas concentration within the enclosed space over consecutive sampling periods can be achieved.

[0095] Figure 8A This is an embodiment of a gas sensor. In process 800A, the gas sensor 802 sends the sensed data 804, representing the gas concentration, to the data processing module 806 of the controller. The data processing module 806 of the controller processes the data representing the gas concentration and outputs gas concentration information 808, which is then fed into a machine learning model.

[0096] like Figure 8A The gas concentration information 800 is represented as a curve. The receiving unit of the data processing module 808 receives the electrical parameter values ​​acquired by the gas sensor at different times. Then, the first processing unit of the data processing module 808 obtains a set of electrical parameter values ​​within the detection period. Afterwards, the second processing unit of the data processing module converts the set of electrical parameter values ​​within the detection period to generate a curve with the horizontal axis representing time and the vertical axis representing the electrical parameter values. In some examples, such as... Figure 8A In one example, when the gas sensor operates at a certain reaction temperature, it generates a curve showing the change of electrical parameter values ​​over time at that reaction temperature, which is then input into the machine learning model. In other examples, when the gas sensor operates at different reaction temperatures, it generates multiple curves showing the change of electrical parameter values ​​over time, each curve representing a specific reaction temperature, which are then input into the machine learning model.

[0097] Figure 8B This is an embodiment with four gas sensors. In process 800A, four sensors are installed in a confined space. This embodiment can be configured with a certain number of gas sensors according to the needs of the scenario; here, four gas sensors are used as an example for explanation.

[0098] In some implementations, when gas sensors 802-1, 802-2, 802-3, and 802-4 collect electrical parameter values ​​804-1, 804-2, 804-3, and 804-4 at a certain moment, they send these values ​​to the data processing module 806 of the controller. The receiving unit in the data processing module 806 receives the electrical parameter values ​​804-1, 804-2, 804-3, and 804-4 at different moments sent by the gas sensors. The receiving unit can receive a set of electrical parameter values ​​sensed by the gas sensors at different moments. Then, the receiving unit sends the received set of electrical parameter values ​​from each gas sensor to the first processing unit. The first processing unit cleans the data from each gas sensor and extracts a set of electrical parameter values ​​within the detection period (see reference...). Figure 6B(Example). Then, the first processing unit sends a set of electrical parameter values ​​within the detection period to the second processing unit. The second processing unit 750 converts the set of electrical parameter values ​​sensed at different times into curves 808-1, 808-2, 808-3, and 808-4 representing gas concentration information in chronological order.

[0099] In some implementations, the data processing unit receives electrical parameter values ​​at different times and at different reaction temperatures. For each gas sensor, after processing by the first processing unit and the second processing unit, multiple curves characterizing the data collected at different reaction temperatures during the detection period are obtained.

[0100] In some implementations, the data processing unit receives electrical parameter values ​​at different times and at different reaction temperatures. For each gas sensor, after processing by the first processing unit and the second processing unit, multiple curves characterizing the data collected at different reaction temperatures over multiple sampling periods are obtained.

[0101] Based on the gas concentration information obtained under processes 800A, 800B, or other implementation methods described above, it can be input. Figure 5 The trained machine learning model 510 generates the opening and closing states of the enclosed space. Correspondingly, when training based on the training set, the sample gas concentration information 512-1 input to the machine learning model can also be processed according to processes 800A, 800B or other embodiments described above to obtain gas concentration information.

[0102] Using the above method, the machine learning model can learn the on / off state of a confined space corresponding to the time-varying gas concentration under different conditions (referring to one or more of the following: different reaction times, different reaction temperatures, different sampling periods, and different channels). During the inference phase, inputting the gas concentration information in curve form into the machine learning model can quickly generate relatively accurate on / off states.

[0103] Figure 9 A schematic diagram of an example process 900 for generating three switching states using a machine learning model according to some embodiments of the present disclosure is shown. Figure 9 As shown, the gas concentration information 902 is input into the trained machine learning model 904, and the output is the opening and closing status 906 of the sealed space.

[0104] During the training phase, the training dataset includes sample gas concentration information and sample on / off states. The sample on / off states include the open state (i.e., the non-sealed state mentioned earlier), the closed state (i.e., the sealed state mentioned earlier), or the on / off operation state. The on / off operation state refers to the state of opening and closing operations performed within a certain threshold time. This threshold time is defined based on usage habits and the typical time for opening and closing operations. For example, if food is typically placed in or out of a refrigerator within 2 minutes of use, then 2 minutes can be defined as the threshold time.

[0105] Based on the above training dataset, according to Figure 5 The process shown is used for training, which yields three switch states that can learn different gas concentration changes. In the inference phase, the machine learning model 904 determines the switch state of the enclosed space as open state 906-1, closed state 906-2, or switch operation state 906-3 based on the gas concentration information 902.

[0106] When the machine learning model determines that the closed space is in an "off" or "on / off" state, it waits for subsequent data collection from the gas sensor to perform further on / off status detection. When the machine learning model determines that the closed space is in an "on" state, it needs to send an alarm message or notification to the user or management terminal to promptly address the unsealed state of the closed space.

[0107] Figure 10 A block diagram of an apparatus 1000 for detecting the opening and closing state of a confined space, according to some embodiments of the present disclosure, is shown. (See reference...) Figure 10 The device 1000 includes a gas concentration information acquisition module 1002, configured to acquire gas concentration information sensed by a gas sensor in the enclosed space, wherein the gas concentration information includes electrical parameter values ​​characterizing the concentrations of various gases in the enclosed space. The device 1000 also includes a switch state detection module 1004, configured to detect the switch state of the enclosed space based on the gas concentration information using a machine learning model.

[0108] In some embodiments, the gas concentration information acquisition module 1002 is configured to: acquire a set of electrical parameter values ​​sensed by the gas sensor at different times; and convert the set of electrical parameter values ​​sensed at different times into a matrix representing the gas concentration information in chronological order.

[0109] In some embodiments, the gas concentration information acquisition module 1002 is configured to: acquire different electrical parameter values ​​sensed by the gas sensor at different reaction temperatures for each moment; and convert the set of electrical parameter values ​​sensed at different times and different reaction temperatures into a matrix representing the gas concentration information in chronological order.

[0110] In some embodiments, the gas sensor includes multiple sensors, and the gas concentration information acquisition module 1002 is configured to: acquire different electrical parameter values ​​sensed by each of the multiple sensors at different reaction temperatures; and convert the set of electrical parameter values ​​sensed by each gas sensor at different times and at different reaction temperatures into a matrix representing the gas concentration information.

[0111] In some embodiments, the gas concentration information acquisition module 1002 is configured to: acquire a set of electrical parameter values ​​sensed by the gas sensor at different times and at different reaction temperatures; and generate a curve representing the gas concentration information based on the set of electrical parameter values.

[0112] In some embodiments, the gas sensor includes multiple sensors, and the gas concentration information acquisition module 1002 is configured to: acquire a set of electrical parameter values ​​sensed by each of the multiple sensors at different times and at different reaction temperatures; and convert the set of electrical parameter values ​​sensed by each gas sensor at different times and at different reaction temperatures into a curve representing the gas concentration information.

[0113] In some embodiments, multiple sets of electrical parameter values ​​under multiple consecutive sampling periods can be sensed based on any of the above embodiments, and a matrix or curve representing gas concentration information can be formed.

[0114] Embodiments of this disclosure also provide an apparatus comprising a sealed space. The apparatus includes a gas sensor, a processor, and a memory located within the sealed space, wherein the memory is coupled to the processor and has instructions stored thereon that, when executed by the processor, cause the execution of methods provided according to various aspects of this disclosure. In some embodiments, the apparatus comprising a sealed space may be a refrigerator, a medicine storage cabinet, a wardrobe, a chemical reagent storage device, a warehouse, a storage room, or a chemical workshop, etc.

[0115] Figure 11 A schematic block diagram of an example device 1100 that can be used to implement embodiments of the present disclosure is shown. Figure 11 As shown, device 1100 includes a processor 1101, which can perform various appropriate actions and processes based on computer program instructions loaded into random access memory (RAM) 1103 according to computer program instructions stored in read-only memory (ROM) 1102. The RAM 1103 may also store various programs and data required for the operation of device 1100. The processor 1101, ROM 1102, and RAM 1103 are interconnected via bus 1104. Input / output (I / O) interface 1105 is also connected to bus 1104.

[0116] The various processes and procedures described above, such as method 200, can be executed by processor 1101. For example, in some embodiments, method 200 may be implemented as a software program tangibly contained in a machine-readable medium. In some embodiments, part or all of the software program may be loaded and / or installed on device 1100 via ROM 1102. When the software program is loaded into RAM 1103 and executed by processor 1101, one or more actions of method 200 described above may be performed.

[0117] The functions described above in this document can be performed at least in part by one or more hardware logic components. For example, exemplary types of hardware logic components that can be used, without limitation, include: field programmable gate arrays (FPGAs), application-specific integrated circuits (ASICs), application-specific standard products (ASSPs), systems-on-a-chip (SoCs), payload programmable logic devices (CPLDs), and so on.

[0118] The program code used to implement the methods of this disclosure may be written in any combination of one or more programming languages. This program code may be provided to a processor or controller of a general-purpose computer, special-purpose computer, or other programmable data processing apparatus, such that when executed by the processor or controller, the program code causes the functions / operations specified in the flowcharts and / or block diagrams to be implemented. The program code may be executed entirely on a machine, partially on a machine, as a standalone software package partially on a machine and partially on a remote machine, or entirely on a remote machine or server.

[0119] This disclosure can be a method, apparatus, system, and / or program product. The program product may include a machine-readable storage medium on which machine-readable program instructions for performing various aspects of this disclosure are loaded. The machine-readable program instructions described herein can be downloaded from the machine-readable storage medium to various computing / processing devices, or downloaded via a network, such as the Internet, a local area network, a wide area network, and / or a wireless network, to an external computer or external storage device. The network may include copper transmission cables, fiber optic transmissions, wireless transmissions, routers, firewalls, switches, gateway computers, and / or edge servers. A network adapter card or network interface in each computing / processing device receives the machine-readable program instructions from the network and forwards them to the machine-readable storage medium in the respective computing / processing device.

[0120] Machine program instructions used to perform the operations of this disclosure may be assembly instructions, instruction set architecture (ISA) instructions, machine instructions, machine-dependent instructions, microcode, firmware instructions, state setting data, or source code or object code written in any combination of one or more programming languages, including object-oriented programming languages ​​such as Smalltalk, C++, etc., and conventional procedural programming languages ​​such as the "C" language or similar programming languages. Machine-readable program instructions may be executed entirely on a user's computer, partially on a user's computer, as a standalone software package, partially on a user's computer and partially on a remote computer, or entirely on a remote computer or server. In cases involving a remote computer, the remote computer may be connected to the user's computer via any type of network—including a local area network (LAN) or a wide area network (WAN)—or may be connected to an external computer (e.g., via the Internet using an Internet service provider). In some embodiments, electronic circuitry, such as programmable logic circuitry, field-programmable gate arrays (FPGAs), or programmable logic arrays (PLAs), is personalized by utilizing state information from the machine-readable program instructions to implement various aspects of this disclosure.

[0121] In the context of this disclosure, a machine-readable medium can be a tangible medium that may contain or store a program for use by or in conjunction with an instruction execution system, apparatus, or device. A machine-readable medium can be a machine-readable signal medium or a machine-readable storage medium. Machine-readable media can be, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, apparatus, or devices, or any suitable combination of the foregoing. More specific examples of machine-readable storage media include electrical connections based on one or more wires, portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fiber, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination of the foregoing. Furthermore, although operations are depicted in a specific order, this should be understood as requiring that such operations be performed in the specific order shown or in sequential order, or requiring that all illustrated operations be performed to achieve the desired result. In certain environments, multitasking and parallel processing may be advantageous. Similarly, while several specific implementation details are included in the foregoing discussion, these should not be construed as limiting the scope of this disclosure. Certain features described in the context of individual embodiments may also be implemented in combination in a single implementation. Conversely, various features described in the context of a single implementation may also be implemented individually or in any suitable sub-combination in multiple implementations.

[0122] Although the subject matter has been described using language specific to structural features and / or methodological logic, it should be understood that the subject matter defined in the appended claims is not necessarily limited to the specific features or actions described above. Rather, the specific features and actions described above are merely illustrative examples of implementing the claims.

Claims

1. A method (200) for detecting an open-close state of an enclosed space, comprising: obtaining (202) gas concentration information sensed by a gas sensor in the enclosed space, wherein the gas concentration information comprises values of electrical parameters representing concentrations of a plurality of gases in the enclosed space; and detecting (204), by a machine learning model, the open-close state of the enclosed space based on the gas concentration information. 2.The method (200) of claim 1, wherein obtaining (202) the gas concentration information sensed by the gas sensor in the enclosed space comprises: obtaining a set of values of electrical parameters sensed by the gas sensor at a plurality of time instants. 3.The method (200) of claim 2, wherein obtaining the set of values of electrical parameters sensed by the gas sensor at a plurality of time instants comprises: obtaining a set of values of electrical parameters sensed by the gas sensor at a plurality of time instants and a plurality of reaction temperatures, wherein each time instant corresponds to one reaction temperature, and different time instants correspond to the same or different reaction temperatures. 4.The method (200) of claim 3, wherein obtaining (202) the gas concentration information sensed by the gas sensor in the enclosed space further comprises: generating a curve representing the gas concentration information based on the set of values of electrical parameters. 5.The method (200) of claim 3, wherein obtaining (202) the gas concentration information sensed by the gas sensor in the enclosed space further comprises: generating a matrix representing the gas concentration information based on the set of values of electrical parameters. 6.The method (200) of any one of claims 2-5, wherein the gas sensor comprises a plurality of sensors, and obtaining the set of values of electrical parameters sensed by the gas sensor at a plurality of time instants and a plurality of reaction temperatures comprises: obtaining a set of values of electrical parameters sensed by each sensor of the plurality of sensors at a plurality of time instants and a plurality of reaction temperatures. 7.The method (200) of claim 4 or 5, wherein obtaining the set of values of electrical parameters sensed by the gas sensor at a plurality of time instants and a plurality of reaction temperatures comprises: obtaining a set of values of electrical parameters sensed by the gas sensor at a plurality of consecutive sampling periods, and in each of the sampling periods, obtaining a set of values of electrical parameters sensed by the gas sensor at a plurality of collection time instants and a plurality of reaction temperatures, wherein each collection time instant corresponds to one reaction temperature, and different collection time instants in the same sampling period correspond to different or the same reaction temperatures. 8.The method (200) of claim 5, wherein the gas sensor senses a set of values of electrical parameters at a plurality of consecutive sampling periods, and generating the matrix representing the gas concentration information based on the set of values of electrical parameters comprises: arranging a set of values of electrical parameters at a plurality of collection time instants in the same sampling period in time order as different columns of the matrix, and arranging a set of values of electrical parameters at different sampling periods in time order as different rows of the matrix.

9. The method (200) according to claim 1, wherein detecting (204) the opening / closing state of the enclosed space by a machine learning model based on the gas concentration information comprises: Based on the gas concentration information, the machine learning model determines the switch state of the enclosed space as open, closed, or in operation.

10. The method (200) according to claim 1, wherein the machine learning model is trained by learning the relationship between the gas concentration information and the opening / closing state of the enclosed space.

11. The method (200) according to claim 10, wherein the gas concentration information required for model training is obtained by the gas sensor sensing the gas in the sealed space under different ambient temperatures and humidity conditions when the sealed space is in different switching states.

12. The method (200) according to any one of claims 1-5, 8-11, wherein the electrical parameter value includes at least one of resistance value, voltage value, and current value.

13. The method (200) according to any one of claims 1-5, 8-11, wherein the opening and closing state of the sealed space indicates the opening and closing state of the refrigerator door, and the gas sensor is a volatile organic compound sensor.

14. A device (1000) for detecting the opening and closing state of a confined space, comprising: A gas concentration information acquisition module (1002) is configured to acquire gas concentration information sensed by a gas sensor in the enclosed space, wherein the gas concentration information includes electrical parameter values ​​characterizing the concentrations of various gases within the enclosed space; and The switch state detection module (1004) is configured to detect the switch state of the enclosed space based on the gas concentration information using a machine learning model.

15. An apparatus comprising a sealed space, comprising: A gas sensor is located in the enclosed space; processor; as well as A memory coupled to the processor and having instructions stored thereon, which, when executed by the processor, cause the method according to any one of claims 1-13 to be performed.

Images