Mode detection and action recognition device

Abstract

The invention provides a mode detection and action recognition device, which relates to the technical field of electronic equipment and comprises an equipment main body, at least two functional modules, an inspection assembly, a deep learning algorithm unit and a microprocessor. The equipment body is provided with an assembling part and a containing cavity. And each functional module is detachably connected with the assembling part in a replaceable manner. And the detection assembly is mounted in the accommodating cavity, is directly or indirectly electrically connected with the assembled functional module, and is used for collecting multi-dimensional data of user operation. The deep learning algorithm unit is arranged in the containing cavity and electrically connected with the microprocessor and the detection assembly, and the deep learning algorithm unit can receive a detection signal of the detection assembly. The microprocessor is installed in the containing cavity and electrically connected with the detection assembly, the deep learning algorithm unit and the assembled function module. According to the invention, the problem of low identification precision caused by insufficient use flexibility and single detection dimension due to non-detachable function modules in the prior art can be solved.

Description

Technical Field

[0001] This disclosure relates to the field of electronic equipment technology, specifically to a pattern detection and action recognition device. Background Technology

[0002] In related technical fields, pattern detection and action recognition devices are often designed with fixed-function integration. This design connects functional modules to the main body of the device in a non-detachable manner, meaning that a single device can only meet the action recognition needs of a specific scenario. When users need to switch between different functional modes, the entire device must be replaced, which undoubtedly greatly reduces the flexibility of the device and causes considerable inconvenience to users.

[0003] Meanwhile, these devices also have limitations in the design of their detection components. They typically can only collect single-dimensional operation data and combine it with traditional algorithms for action recognition. However, this recognition method struggles to accurately capture the complex and ever-changing intentions of users, resulting in low recognition accuracy. In practical applications, misjudgments or omissions frequently occur, severely impacting the device's responsiveness to user operations and reducing user experience. Furthermore, the coordination between control units, functional modules, and detection components in some devices is insufficient. High data transmission latency further reduces the real-time performance of action recognition and function execution. When a user performs an operation, the device may fail to respond promptly due to data transmission delays, leading to a degraded user experience.

[0004] These issues collectively limit the scene adaptability and user experience of pattern detection and action recognition devices, and urgently require improvement through technological innovation. Summary of the Invention

[0005] The main purpose of this disclosure is to provide a pattern detection and action recognition device to solve the problems of insufficient flexibility in use and low recognition accuracy caused by the non-removable functional modules and the single detection dimension in related technologies.

[0006] To achieve the above objectives, this disclosure provides a pattern detection and action recognition device, including a device body, at least two functional modules, an inspection component, a deep learning algorithm unit, and a microprocessor.

[0007] The main body of the equipment has an assembly section and a receiving cavity. Each functional module is replaceably and detachably connected to the assembly section. The detection component, installed in the receiving cavity, is directly or indirectly electrically connected to the assembled functional module. The detection component is used to collect multi-dimensional data from user operations.

[0008] The deep learning algorithm unit is located in the cavity and is electrically connected to the microprocessor and the detection component. The deep learning algorithm unit is able to receive the detection signal from the detection component.

[0009] The microprocessor is installed in the receiving cavity and is electrically connected to the detection component, the deep learning algorithm unit, and the assembled functional modules.

[0010] In some examples, at least two functional modules include at least one of the following: ignition module, aromatherapy module, display module, speaker module, recording module, signal enhancement module, lighting module, alarm module, electric shock module, and remote control module.

[0011] In some examples, the functional module is an ignition module, and the main body of the device is provided with a first control element for controlling the ignition module. The ignition module is provided with an ignition structure, and the first control element can turn the ignition structure on or off.

[0012] Alternatively, when the functional module is an aromatherapy module, the main body of the device is equipped with a second control component for adjusting the amount of aromatherapy released, and the aromatherapy module is equipped with a scenting chip slot. The second control component controls the concentration or diffusion rate of the aroma emitted by the aromatherapy module.

[0013] Alternatively, the functional module can be a display module. The main body of the device has a third control unit adapted to the display module, and the display module has a display area. The third control unit can control the content displayed in the display area. The display module can display relevant information detected by the detection component and the data results processed by the deep learning algorithm unit.

[0014] Alternatively, the functional module is a speaker module, and the main body of the device is equipped with a fourth control unit for volume adjustment, which can adjust the volume of the sound emitted by the speaker module.

[0015] Alternatively, the functional module is a recording module, and the main body of the device is equipped with a fifth control unit for starting and stopping recording. The fifth control unit can start or stop the recording operation.

[0016] Alternatively, the functional module is a signal enhancement module, with a sixth control unit on the main body of the device to control the strength of the signal, and a signal strength indicator on the signal enhancement module. The sixth control unit can control the strength of the enhanced signal of the signal enhancement module and provide a prompt through the signal strength indicator.

[0017] Alternatively, when the functional module is a lighting module, a seventh control component for lighting mode is provided on the main body of the equipment. The seventh control component can switch the lighting module on / off or the lighting mode.

[0018] Alternatively, the functional module is an alarm module, and the main body of the device is equipped with an eighth control component, which can adjust the alarm sensitivity of the alarm module.

[0019] Alternatively, the functional module is an electric shock module, with a ninth control component on the main body of the device, an electric shock safety protection structure on the electric shock module, and the ninth control component can adjust the electric shock intensity of the electric shock module.

[0020] Alternatively, the functional module is a remote control module, with a tenth control component on the main body of the device. The remote control module is equipped with a signal communication structure, and the tenth control component can control the opening and closing or frequency of the signal communication structure.

[0021] The first control element, the second control element, the third control element, the fourth control element, the fifth control element, the sixth control element, the seventh control element, the eighth control element, the ninth control element, and the tenth control element are the same control element, or at least two control elements are integrated into one control element.

[0022] In some examples, the control includes at least one of a knob, button, wheel, joystick, touch key, or electromagnetic induction.

[0023] In some examples, the detection component includes at least two detection devices. The detection component is used to collect multi-dimensional data of user operations, and a deep learning algorithm unit is used to fuse and process the sensor data to identify user intentions and actions. The microprocessor controls the corresponding functional modules to execute specific control methods based on the recognition results.

[0024] In some examples, the detection components include at least one of an accelerometer, pressure sensor, displacement sensor, temperature sensor, voltage sensor, current sensor, gravity sensor, gyroscope, infrared sensor, acoustic wave detector, and electromagnetic radar, used to collect user operation data, ambient temperature data, and motion parameters.

[0025] In some examples, the deep learning algorithm unit includes a data preprocessing module and a pattern recognition model.

[0026] The data preprocessing module is used to filter, reduce noise, and extract features from the raw data of the detection components.

[0027] The pattern recognition model is optimized through training sample data to improve the accuracy of recognizing ignition mode, aromatherapy mode and other functional modes.

[0028] The deep learning algorithm unit supports OTA upgrades, and receives algorithm update packages through the wireless communication module to achieve online model optimization.

[0029] In some examples, the pattern detection and motion recognition device also includes a child lock transition module and a human-machine interface module. The child lock transition module is electrically connected to the microprocessor and is used to perform child lock protection operations during function module switching. The human-machine interface module is used to receive user commands and display the operating status.

[0030] In some examples, the child lock transition module includes a delay circuit, a contact isolation unit, and a password verification unit. After the mode switching command is triggered, the delay circuit controls the contact isolation unit to activate, connecting the contacts of the new module after a preset delay. Users must deactivate the child lock protection using a preset password or biometric verification to prevent accidental operation by minors and short circuits before module replacement.

[0031] In some examples, each functional module is configured with an independent identification code, and the microprocessor identifies the module type and corresponding control method by recognizing the identification code.

[0032] In some examples, the functional modules also include a decompression module, or the decompression module is provided on the main body of the device.

[0033] The decompression module is equipped with vibration feedback or audio-visual interactive control. The microprocessor activates the corresponding decompression mode based on the user's action recognition results.

[0034] In the pattern detection and action recognition device provided in this embodiment, a collaborative approach is adopted, consisting of a main body, at least two functional modules, a detection component, a deep learning algorithm unit, and a microprocessor. The functional modules are detachably connected through the main body assembly, the detection component collects multi-dimensional operation data, the deep learning algorithm unit identifies user intentions, and the microprocessor controls the functional modules to perform operations. This achieves the goal of real-time detection of user operation intentions and intelligent control of functional modules, thereby improving the flexibility of device use, scene adaptability, and accuracy of functional pattern recognition. It also solves the technical problems of limited device use and inaccurate operation response caused by the fixed function of a single device and low accuracy of user operation intention recognition. Attached Figure Description

[0035] To more clearly illustrate the technical solutions in the specific embodiments or related technologies of this disclosure, the accompanying drawings used in the description of the specific embodiments or related technologies will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of this disclosure. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0036] Figure 1 A schematic diagram of the structure of the pattern detection and action recognition device provided in the embodiments of this disclosure; Figure 2 This is a structural diagram of the pattern detection and motion recognition device provided in this embodiment of the present disclosure when the functional module is an ignition module; Figure 3 This is a structural diagram of the pattern detection and action recognition device provided in this embodiment of the present disclosure when the functional module is an aromatherapy module.

[0037] Figure label: 100. Main body of the equipment; 110. Assembly section; 200. Functional module; 210. Ignition module; 211. Ignition structure; 220. Aromatherapy module; 221. Incense chip slot; 300. Decompression module; 400. Detection component; 500. Deep learning algorithm unit; 600. Microprocessor. Detailed Implementation

[0038] To enable those skilled in the art to better understand the present disclosure, the technical solutions of the present disclosure will be clearly and completely described below with reference to the accompanying drawings of the embodiments. Obviously, the described embodiments are only some embodiments of the present disclosure, and not all embodiments. Based on the embodiments of the present disclosure, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of the present disclosure.

[0039] It should be noted that the terms "first," "second," etc., in the specification, claims, and accompanying drawings of this disclosure are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate for the embodiments of this disclosure described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0040] In this disclosure, the terms "upper," "lower," "left," "right," "front," "rear," "top," "bottom," "inner," "outer," "middle," "vertical," "horizontal," "lateral," and "longitudinal" indicate orientations or positional relationships based on the orientations or positional relationships shown in the accompanying drawings. These terms are primarily for the purpose of better describing this disclosure and its embodiments, and are not intended to limit the indicated devices, elements, or components to having a specific orientation, or to be constructed and operated in a specific orientation.

[0041] Furthermore, in addition to indicating location or positional relationship, some of the aforementioned terms may also have other meanings. For example, the term "above" may also be used in certain circumstances to indicate a dependency or connection. Those skilled in the art can understand the specific meaning of these terms in this disclosure according to the specific circumstances.

[0042] Furthermore, the terms "installation," "setup," "equipped with," "connection," "linking," and "socketing" should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral structure; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium, or an internal connection between two devices, components, or parts. Those skilled in the art can understand the specific meaning of the above terms in this disclosure according to the specific circumstances.

[0043] It should be noted that, unless otherwise specified, the embodiments and features described in this disclosure can be combined with each other. This disclosure will now be described in detail with reference to the accompanying drawings and embodiments.

[0044] Reference Figures 1 to 3 This disclosure provides a pattern detection and action recognition device, including a device body 100, at least two functional modules 200, an inspection component, a deep learning algorithm unit 500, and a microprocessor 600.

[0045] The main body 100 of the device has an assembly section 110 and a receiving cavity. Each functional module 200 is replaceably and detachably connected to the assembly section 110. A detection component 400 is installed in the receiving cavity and is electrically connected directly or indirectly to the assembled functional module 200. The detection component 400 is used to collect multi-dimensional data of user operation.

[0046] The deep learning algorithm unit 500 is disposed in the receiving cavity and is electrically connected to the microprocessor 600 and the detection component 400. The deep learning algorithm unit 500 is able to receive the detection signal from the detection component 400.

[0047] The microprocessor 600 is installed in the receiving cavity and is electrically connected to the detection component 400, the deep learning algorithm unit 500 and the assembled functional module 200.

[0048] The technical solution disclosed herein achieves detachable connection of functional modules 200 through the assembly part 110 of the main body 100, solving the problem of insufficient flexibility caused by the non-detachable functional modules 200 in related technologies. Users can replace functional modules 200 as needed according to the scenario, improving the flexibility of equipment use. The detection component 400 collects multi-dimensional operation data and performs intent recognition in conjunction with the deep learning algorithm unit 500, solving the problem of low recognition accuracy caused by the single detection dimension in related technologies, and improving the accuracy of user operation intent recognition. The microprocessor 600 coordinates the signal interaction of each module to realize closed-loop control of "data acquisition-intent recognition-function execution", improving the real-time performance of action recognition and function execution, and enhancing the adaptability of the equipment to the scenario and the user experience.

[0049] The above structure, through the collaboration of the main body 100, functional module 200, detection component 400, deep learning algorithm unit 500 and microprocessor 600, constructs a motion recognition system with replaceable functional module 200, realizing real-time detection of user operation intentions and intelligent control of functional module 200.

[0050] The main body 100 of the equipment enables the detachable connection of the functional modules 200 through the assembly part 110. The cavity provides integrated space for the detection component 400, algorithm unit and microprocessor 600, ensuring the physical isolation and electrical connection stability of each component, supporting the replacement of functional modules 200 as needed, and improving the modularity of the equipment structure and space utilization.

[0051] At least two replaceable functional modules 200 (such as ignition and aromatherapy modules 220) in the functional module 200 are designed to be detachable, allowing users to switch functions according to the scenario, avoiding the solidification of a single device function, and enhancing the flexibility and adaptability of the device to different scenarios.

[0052] The detection component 400 can collect multi-dimensional data of user operations (such as acceleration and pressure), providing raw input for action recognition. It is directly or indirectly electrically connected to the functional module 200 to ensure real-time correlation between data acquisition and module status, laying the data foundation for subsequent intent recognition.

[0053] The deep learning algorithm unit 500 can receive and analyze signals from the detection component 400, and recognize user operation intentions (such as ignition and aromatherapy adjustment actions) through data preprocessing and pattern recognition models. It supports OTA upgrades to optimize the model and improve the accuracy of functional pattern recognition and algorithm iteration capabilities.

[0054] The microprocessor 600 can serve as the core control hub, coordinating the signal interaction between the detection component 400, the algorithm unit, and the functional module 200 to achieve closed-loop control of "data acquisition - intent recognition - function execution," ensuring the real-time performance and stability of the collaborative work of each module.

[0055] This disclosure enables modular expansion, multi-dimensional data acquisition, and intelligent intent recognition. Specifically, at least two detachable functional modules 200 (such as ignition, aromatherapy, and lighting) can be replaced as needed, improving the flexibility of device use; the detection component 400 collects user operation data (such as acceleration and pressure) to provide input for action recognition; and the deep learning algorithm unit 500 analyzes the detection signals to identify user operation patterns (such as ignition actions and aromatherapy adjustment intentions).

[0056] This disclosure improves the accuracy of functional pattern recognition by using deep learning algorithms to fuse multi-dimensional data (such as filtering, feature extraction, and model training), and supports OTA upgrades to optimize the model. It also enables cross-module collaborative control, with the microprocessor 600 acting as the core hub to coordinate the signal interaction between the detection component 400, the algorithm unit, and the functional module 200, achieving closed-loop control of "data acquisition - intent recognition - function execution". Furthermore, it enhances security and adaptability by using an independent identification code for the functional module 200 to avoid incorrect matching, and combining it with a child lock transition module (delay circuit, password verification) to prevent accidental operation, thus adapting to the needs of minor protection and multi-scenario use.

[0057] Reference Figures 1 to 3 In some examples, at least two functional modules 200 include at least one of the following: ignition module 210, aromatherapy module 220, display module, speaker module, recording module, signal enhancement module, lighting module, alarm module, electric shock module, and remote control module.

[0058] These functional modules 200 can be flexibly combined and replaced according to actual needs. For example, in home use, users may prefer to choose the ignition module 210, aromatherapy module 220, and lighting module to meet daily needs for ignition, aromatherapy adjustment, and lighting; while in some special application scenarios, such as outdoor adventures or emergency rescue, users may choose the signal enhancement module, alarm module, and remote control module to enhance the device's communication capabilities, provide security, and achieve remote control. Through this modular design, the pattern detection and motion recognition device of this disclosure can better adapt to the needs of different users and scenarios, improving the device's practicality and flexibility.

[0059] Reference Figures 1 to 3 In some examples, the functional module 200 is an ignition module 210. The main body of the device 100 is provided with a first control element for controlling the ignition module 210. The ignition module 210 is provided with an ignition structure 211. The first control element can turn the ignition structure 211 on or off.

[0060] When a user operates the first control component, the detection component 400 collects the operation data of the control component in real time, such as the operating force and speed, and transmits this data to the deep learning algorithm unit 500. The deep learning algorithm unit 500 performs preprocessing operations such as filtering, noise reduction, and feature extraction on the collected data, and then uses a pre-trained pattern recognition model to identify the user's operating intention. If the identification result indicates that the ignition structure 211 should be turned on, the microprocessor 600 will immediately send an on command to the ignition module 210 to activate the ignition structure 211 and achieve the ignition function; if the identification result indicates that the ignition structure 211 should be turned off, the microprocessor 600 will send a off command to the ignition module 210 to stop ignition. Simultaneously, the human-machine interface module will display the working status of the ignition module 210 in real time, such as whether ignition was successful, to help the user understand the device's operating status.

[0061] In addition, to ensure safety during the ignition process, the child lock transition module (described below) will control the contact isolation unit to operate via a delay circuit after the ignition command is triggered. After a preset delay, the contacts of the ignition module 210 will be connected. The user must then deactivate the child lock protection using a preset password or biometric verification to prevent accidental operation by minors. Each ignition module 210 is equipped with a unique identification code. The microprocessor 600 identifies the module type as ignition module 210 by recognizing this code and uses the corresponding control method to control its ignition operation.

[0062] Reference Figures 1 to 3 In other examples, when the functional module 200 is an aromatherapy module 220, the main body 100 of the device is provided with a second control element for adjusting the amount of aroma release, and the aromatherapy module 220 is provided with a scenting chip slot 221. The second control element controls the concentration or diffusion rate of the aroma emitted by the aromatherapy module 220 to meet the needs of aroma concentration in different scenarios.

[0063] Specifically, when the user operates the second control component, the detection component 400 also collects operation data in real time, such as the adjustment range and speed, and transmits it to the deep learning algorithm unit 500. The deep learning algorithm unit 500 preprocesses the collected data, including filtering, noise reduction, and feature extraction, and then uses a pre-trained pattern recognition model to identify the user's operation intention.

[0064] If the recognition result indicates an increase in aroma release, the microprocessor 600 immediately sends an instruction to the aroma module 220 to increase the release amount, causing the aroma module 220 to adjust the working state of the aroma diffuser slot 221 to increase the concentration or diffusion rate of the aroma. If the recognition result indicates a decrease in aroma release, the microprocessor 600 sends an instruction to the aroma module 220 to decrease the release amount, reducing the concentration or diffusion rate of the aroma. Simultaneously, the human-machine interface module displays the real-time working status of the aroma module 220, such as the current aroma release amount and the number of remaining aroma diffusers, allowing users to easily understand the device's operating status.

[0065] In addition, the aromatherapy module 220 is also equipped with an independent identification code. The microprocessor 600 identifies the module type as aromatherapy module 220 by recognizing the identification code and uses the corresponding control method to control its aroma release operation.

[0066] In the above structure, the aromatherapy module 220 can be equipped with corresponding heaters, fans, and other structures to adjust the amount and speed of fragrance release from the aromatherapy tablets in the aromatherapy module 220.

[0067] Reference Figures 1 to 3 In other examples, the functional module 200 is a display module, and the main body 100 of the device is provided with a third control component adapted to the display module. The display module is provided with a display area, and the third control component can control the display content of the display area.

[0068] When a user operates the third control unit, the detection component 400 promptly captures operation data, such as operation direction and duration, and quickly transmits this data to the deep learning algorithm unit 500. The deep learning algorithm unit 500 performs detailed preprocessing on the collected data, including filtering, noise reduction, and feature extraction, and then uses a pre-trained pattern recognition model to accurately identify the user's operation intention.

[0069] If the recognition result indicates a need to switch display content, the microprocessor 600 will immediately send a switching command to the display module, causing the display module to display the corresponding content in the display area. If the recognition result indicates a need to adjust parameters such as display brightness, the microprocessor 600 will send an adjustment command to the display module to change the display effect of the display area. Simultaneously, the human-machine interface module will display the real-time operating status of the display module, such as the current display content and display brightness, allowing users to understand the device's operating status.

[0070] In addition, the display module is also equipped with an independent identification code. The microprocessor 600 identifies the module type as a display module by recognizing the identification code and uses the corresponding control method to control its display operation.

[0071] In other examples, functional module 200 is a speaker module, and the main body 100 is equipped with a fourth control for volume adjustment, which can adjust the volume of the sound emitted by the speaker module.

[0072] When the user operates the fourth control element, the detection component 400 quickly collects operation data, such as the adjustment range and the speed of operation, and rapidly transmits this data to the deep learning algorithm unit 500. The deep learning algorithm unit 500 performs comprehensive and detailed preprocessing on the collected data, including data filtering to remove interference signals, noise reduction to improve data quality, and feature extraction to obtain key information. Subsequently, it uses a pre-trained pattern recognition model to accurately identify the user's operation intention.

[0073] If the recognition result indicates an increase in volume, the microprocessor 600 will immediately send a command to the speaker module to increase the volume, causing the speaker module to raise the volume of the emitted sound; if the recognition result indicates a decrease in volume, the microprocessor 600 will send a command to the speaker module to decrease the volume, reducing the volume of the emitted sound. Simultaneously, the human-machine interface module will display the speaker module's operating status in real time, such as the current volume level and whether it is in mute mode, allowing users to easily understand the device's operating status.

[0074] The aforementioned speaker module can be at least one of a Bluetooth speaker module or a Bluetooth headset module. This speaker module has broad applicability, not limited to a specific type, but encompassing multiple types. Specifically, the speaker module can be a Bluetooth speaker module, commonly used in various portable speakers, enabling the speaker to wirelessly connect to devices such as mobile phones and computers via Bluetooth technology to play high-quality music. Furthermore, the speaker module can also be a Bluetooth headset module, commonly found in various wireless headphones, using Bluetooth technology to connect to audio source devices and provide users with a convenient wireless listening experience. Therefore, the aforementioned speaker module can be used in at least one of Bluetooth speaker modules and Bluetooth headset modules, offering considerable flexibility and a wide range of applications.

[0075] The Bluetooth headset module includes at least one earphone unit and a charging case, or the charging case may have an independent power supply and may share the power module of the main device.

[0076] Specifically, the aforementioned headphone unit can be configured as one, two, or even four, depending on the needs, allowing two people to share one unit. A typical headphone unit includes an audio driver (such as a dynamic or balanced armature driver), a Bluetooth communication chip, a battery, a microphone, and control buttons. The headphone unit is responsible for audio playback, receiving Bluetooth signals for wireless connection, making calls or inputting voice commands via the microphone, and controlling playback and answering calls via the control buttons.

[0077] One possible structure for a charging case includes a case body, a charging port (such as charging contacts), and a power indicator light. The charging case provides storage space for the earbuds and charges them when not in use, ensuring they always have a sufficient power level. In this case, the charging case uses the device's main power module.

[0078] In another configuration of the charging case, the charging case can have a built-in independent battery pack that does not rely on the main power supply of the device and can charge the earphones independently or be used as a power bank.

[0079] When the charging compartment has an independent battery pack, it can also share the main power module of the device.

[0080] The earphone units in this application feature a lightweight design, ensuring comfortable wear and preventing them from falling out, providing users with a stable listening experience over extended periods. The charging case not only stores the earphone units but also charges them when their battery is low, ensuring users can always use fully charged earphones. When the charging case has its own independent power source, it can operate independently of the main device, making it convenient for users to carry and use independently in situations such as when out and about. When the charging case shares the main device's power module, it simplifies the overall structure, reduces unnecessary power components, and lowers the cost and complexity of the device.

[0081] In addition, the speaker module is also equipped with an independent identification code. The microprocessor 600 identifies the module type as a speaker module by recognizing the identification code and uses the corresponding control method to control its volume adjustment operation.

[0082] In other examples, the functional module 200 is a recording module, and the main body of the device 100 is provided with a fifth control element for starting and stopping recording. The fifth control element can start or stop the recording operation.

[0083] When the user activates recording using the fifth control, the detection component 400 immediately collects operation data, such as the time the operation was triggered and the duration of the operation, and quickly transmits this data to the deep learning algorithm unit 500. The deep learning algorithm unit 500 performs rigorous preprocessing on the collected data, including data filtering to eliminate noise interference, noise reduction to enhance data reliability, and feature extraction to capture key information. Then, it uses a pre-trained pattern recognition model to accurately identify the user's operation intention.

[0084] If the recognition result indicates that recording has started, the microprocessor 600 will immediately send a start recording command to the recording module, causing the recording module to begin recording sound information; if the recognition result indicates that recording has ended, the microprocessor 600 will send a stop recording command to the recording module, terminating the recording process. Simultaneously, the human-machine interface module will display the real-time operating status of the recording module, such as whether it is currently recording and the duration of recorded audio, allowing users to understand the device's operating status.

[0085] In addition, the recording module is also equipped with an independent identification code. The microprocessor 600 identifies the module type as a recording module by recognizing the identification code and uses the corresponding control method to control its recording operation.

[0086] In other examples, the functional module 200 is a signal enhancement module, and the main body 100 of the device is provided with a sixth control component for controlling the strength of the signal. The signal enhancement module is provided with a signal strength indicator. The sixth control component can control the strength of the enhanced signal of the signal enhancement module and provide a prompt through the signal strength indicator.

[0087] When the user operates the sixth control element, the detection component 400 quickly collects operation data, such as the adjustment range and operation speed, and transmits this data to the deep learning algorithm unit 500 in a timely manner. The deep learning algorithm unit 500 performs comprehensive and detailed preprocessing on the collected data, including data filtering to remove interference components from the signal, noise reduction to improve data quality, and feature extraction to obtain key feature information. Subsequently, it uses a pre-trained pattern recognition model to accurately identify the user's operation intention.

[0088] If the identification result is an enhanced signal, the microprocessor 600 will immediately send an instruction to the signal enhancement module to increase the signal strength, and the signal strength indicator will accordingly display a stronger signal status. If the identification result is a weakened signal, the microprocessor 600 will send an instruction to the signal enhancement module to weaken the signal strength, and the signal strength indicator will accordingly display a weaker signal status. Simultaneously, the human-machine interface module will display the real-time operating status of the signal enhancement module, such as the current signal strength and the trend of signal enhancement or weakening, allowing users to understand the device's operating status.

[0089] In addition, the signal enhancement module is also equipped with an independent identification code. The microprocessor 600 identifies the module type as a signal enhancement module by recognizing the identification code and uses the corresponding control method to control its signal enhancement operation.

[0090] In other examples, when the functional module 200 is a lighting module, the main body 100 of the device is provided with a seventh control element for lighting mode, which can switch the lighting module on / off or the lighting mode.

[0091] When the user operates the seventh control, the detection component 400 quickly collects operation data, such as the direction of the operation (e.g., swiping up to turn on the lighting, swiping down to turn off the lighting, swiping left or right to switch lighting modes), the speed of the operation, and the duration of the operation, and transmits this data to the deep learning algorithm unit 500 in a timely manner. The deep learning algorithm unit 500 performs detailed and comprehensive preprocessing on the collected data, including data filtering to remove possible interference signals and ensure data purity; noise reduction to improve data quality so that the data more accurately reflects the user's operation intention; and feature extraction to obtain key information, such as the magnitude of the operation. Subsequently, a pre-trained pattern recognition model is used to accurately identify the user's operation intention.

[0092] If the recognition result indicates that the lighting is on, the microprocessor 600 will immediately send an on command to the lighting module, causing the lighting module to start emitting light. If the recognition result indicates that the lighting is off, the microprocessor 600 will send an off command to the lighting module, causing the lighting module to stop emitting light. If the recognition result indicates that the lighting mode is being switched (e.g., from normal lighting mode to night light mode), the microprocessor 600 will send a mode switching command to the lighting module, causing the lighting module to emit light according to the corresponding mode. Simultaneously, the human-machine interface module will display the real-time operating status of the lighting module, such as whether the lighting is currently on, the current lighting mode, and the lighting brightness, allowing users to understand the device's operating status.

[0093] In addition, the lighting module is also equipped with an independent identification code. The microprocessor 600 identifies the module type as a lighting module by recognizing the identification code and uses the corresponding control method to control its lighting operation.

[0094] In other examples, functional module 200 is an alarm module, and the main body 100 of the device is equipped with an eighth control element, which can adjust the alarm sensitivity of the alarm module.

[0095] When the user operates the eighth control element, the detection component 400 immediately collects operation data, such as the adjustment scale position and the speed of operation, and quickly transmits this data to the deep learning algorithm unit 500. The deep learning algorithm unit 500 performs rigorous and detailed preprocessing on the collected data, including data filtering to eliminate possible interference signals and ensure the accuracy and stability of the data; noise reduction to improve data quality so that the data more realistically reflects the user's operation intention; and feature extraction to obtain key information, such as the adjustment range. Subsequently, a pre-trained pattern recognition model is used to accurately identify the user's operation intention.

[0096] The alarm module can have an alarm-based item locator function. When a user loses an item and the item is specifically associated with the alarm module (e.g., via Bluetooth connection), the user can trigger the alarm-based item locator command by operating the relevant control components on the main unit of the device (which can be set individually or extended using existing eighth control components). The detection component 400 will promptly capture this operation data and transmit it to the deep learning algorithm unit 500. After preprocessing and pattern recognition, if the recognition result indicates that the alarm-based item locator should be activated, the microprocessor 600 will immediately send an alarm-based item locator command to the alarm module. Upon receiving the command, the alarm module will emit a specific, relatively loud, and conspicuous alarm sound (the sound frequency, pitch, etc., can be adjusted according to presets), and may also be accompanied by flashing lights (if the alarm module is equipped with corresponding lighting components) to help the user quickly find the lost item.

[0097] If the identification result indicates an increase in alarm sensitivity, the microprocessor 600 will immediately send an instruction to the alarm module to increase sensitivity, causing the alarm module to trigger an alarm under more minor abnormalities. If the identification result indicates a decrease in alarm sensitivity, the microprocessor 600 will send an instruction to the alarm module to decrease sensitivity, reducing unnecessary alarm triggering. Simultaneously, the human-machine interface module will display the alarm module's operating status in real time, such as the current alarm sensitivity level and whether it is in alarm mode, allowing users to understand the device's operating status.

[0098] In addition, the alarm module is also equipped with an independent identification code. The microprocessor 600 identifies the module type as an alarm module by recognizing the identification code and uses the corresponding control method to control its alarm sensitivity adjustment operation.

[0099] In other examples, the functional module 200 is an electric shock module, the main body 100 of the device is provided with a ninth control element, the electric shock module is provided with an electric shock safety protection structure, and the ninth control element can adjust the electric shock intensity of the electric shock module.

[0100] When the user operates the ninth control element, the detection component 400 quickly collects operation data, such as the adjustment level, the force applied, and the duration of the operation, and transmits this data to the deep learning algorithm unit 500 in a timely manner. The deep learning algorithm unit 500 performs comprehensive and in-depth preprocessing on the collected data, including data filtering to remove noise interference from the signal and ensure data purity; noise reduction to improve data reliability so that the data more accurately reflects the user's operating intentions; and feature extraction to obtain key feature information, such as the specific adjustment range. Subsequently, a pre-trained pattern recognition model is used to accurately identify the user's operating intentions.

[0101] If the identification result indicates an increase in electric shock intensity, the microprocessor 600 will immediately send a command to the electric shock module to increase the electric shock intensity, causing the electric shock module to output a stronger electric shock. If the identification result indicates a decrease in electric shock intensity, the microprocessor 600 will send a command to the electric shock module to decrease the electric shock intensity, reducing the output electric shock intensity. Simultaneously, the human-machine interface module will display the real-time operating status of the electric shock module, such as the current electric shock intensity level and whether the electric shock is in a ready state, allowing users to understand the device's operating status.

[0102] Furthermore, the electric shock module's safety protection structure automatically cuts off the electric shock output when abnormal conditions are detected, such as the electric shock intensity exceeding the safe range or the electric shock duration being too long, ensuring safe use. The electric shock module also has an independent identification code. The microprocessor 600 identifies the module type as an electric shock module by recognizing this code and uses the corresponding control method to control its electric shock intensity adjustment.

[0103] Electric shock safety protection structures prevent accidental electric shocks and often incorporate multiple safety mechanisms. For example, an insulating protective layer is installed on the outer shell of the electric shock module to prevent accidental contact with internal live components. Simultaneously, a dual-button trigger design requires the user to press two specific buttons simultaneously to activate the electric shock function, effectively reducing the risk of electric shocks caused by single-handed operation or children playing with the device. Furthermore, the electric shock safety protection structure has a built-in current monitoring sensor that monitors the magnitude of the electric shock output current in real time. Once the current exceeds a preset safety threshold, the protection structure immediately cuts off the circuit and stops the electric shock output, preventing injury to the user or others due to excessive electric shock intensity. Moreover, the structure has a self-test function; each time the device is started, it automatically checks all functions of the electric shock safety protection structure to ensure it is in normal working order. If a fault or abnormality is detected, the device will issue a warning message through the human-machine interface module, reminding the user to repair or replace parts in a timely manner, ensuring the safety and reliability of the device.

[0104] In other examples, the functional module 200 is a remote control module, and the main body 100 of the device is provided with a tenth control element. The remote control module is provided with a signal communication structure, and the tenth control element can control the opening and closing or frequency of the signal communication structure.

[0105] When the user operates the tenth control element, the detection component 400 immediately collects operation data, such as the type of operation (on, off, or frequency adjustment), the time of the operation, and the duration of the operation, and quickly transmits this data to the deep learning algorithm unit 500. The deep learning algorithm unit 500 performs meticulous and comprehensive preprocessing on the collected data, including data filtering to eliminate interference that may be introduced during signal transmission to ensure data accuracy; noise reduction to improve data quality so that the data more realistically reflects the user's operating intentions; and feature extraction to obtain key information, such as whether the operation is inclined to on, off, or frequency adjustment. Subsequently, a pre-trained pattern recognition model is used to accurately identify the user's operating intentions.

[0106] If the identification result indicates that the signal communication structure is enabled, the microprocessor 600 will immediately send an enable command to the remote control module, causing the signal communication structure to start working and establishing a communication connection with other devices. If the identification result indicates that the signal communication structure is disabled, the microprocessor 600 will send a disable command to the remote control module, cutting off the signal communication and stopping the connection with other devices. If the identification result indicates that the signal communication frequency needs to be adjusted, the microprocessor 600 will send a frequency adjustment command to the remote control module, causing the signal communication structure to operate at the corresponding frequency. Simultaneously, the human-machine interface module will display the real-time operating status of the remote control module, such as whether the current signal communication structure is enabled and the communication frequency, allowing users to understand the device's operating status.

[0107] Furthermore, the signal communication structure of the remote control module is designed with stability and anti-interference capabilities in mind. Advanced encoding and decoding technology is employed to ensure accurate signal reception and identification during transmission, reducing communication failures caused by signal interference or loss. Moreover, the signal communication structure also features adaptive adjustment, automatically adjusting communication parameters based on the signal strength and interference conditions of the surrounding environment to maintain optimal communication performance. The remote control module also has an independent identification code; the microprocessor 600 identifies the module type as a remote control module by recognizing this code and uses the corresponding control method to control its signal communication operation.

[0108] The first, second, third, fourth, fifth, sixth, seventh, eighth, ninth, and tenth control components are the same control component. In other examples, at least two control components are integrated into one control component.

[0109] The design of integrating multiple control components into a single control unit greatly enhances the convenience and efficiency of device operation. For example, controls for functions such as volume adjustment, recording start / stop, and signal strength control can be integrated into a single multi-functional control knob or button combination. When the user rotates this multi-functional control knob, the detection component 400 quickly collects relevant information based on the direction, angle, and speed of rotation, and transmits it to the deep learning algorithm unit 500. The deep learning algorithm unit 500 performs meticulous preprocessing on the collected data, including data filtering to eliminate noise interference that may occur during operation and ensure data purity; noise reduction processing to improve data quality so that the data can more accurately reflect the user's operation intention; and feature extraction to obtain key information, such as the operation function corresponding to the rotation direction and the adjustment range corresponding to the rotation angle. Subsequently, a pre-trained pattern recognition model is used to accurately identify the user's operation intention.

[0110] If the recognition result indicates volume adjustment, the microprocessor 600 will send a corresponding volume adjustment command to the speaker module based on the rotation direction and angle, thereby increasing or decreasing the volume. If the recognition result indicates start or stop recording, the microprocessor 600 will send a corresponding start / stop command to the recording module to control the start and stop of recording. If the recognition result indicates signal strength adjustment, the microprocessor 600 will send a command to the signal enhancement module to enhance or weaken the signal, adjusting the signal strength. Simultaneously, the human-machine interface module will display the real-time operating status of each functional module 200, such as the current volume, whether recording is in progress, and the signal strength level, allowing users to fully understand the device's operating status. This integrated control design not only reduces the number of controls on the device, making the device more aesthetically pleasing, but also reduces the complexity of user operation and improves the user experience.

[0111] The functional modules of this application can also be configured as thermometers, hygrometers, temperature and humidity meters, infrared sensors (range measurement, angle measurement), etc. Specific configurations may be required.

[0112] When configured as a thermometer module, the microprocessor 600 receives data from the temperature sensor and displays the current ambient temperature in real time through the human-machine interface module, allowing users to intuitively obtain temperature information. If configured as a hygrometer module, the system collects ambient humidity data through a humidity sensor, processes it through the microprocessor 600, and presents the humidity value on the interactive interface, facilitating user monitoring of air humidity. When using a combined thermometer and hygrometer module, the device can simultaneously detect temperature and humidity parameters, achieving dual-function integration through a single module and reducing hardware footprint. For the infrared sensor module, if it's for ranging, the sensor emits an infrared signal and calculates the reflection time; the microprocessor 600 then calculates the target distance and displays the specific value through the interactive module. If it's for angle measurement, the sensor detects the offset angle of the infrared signal, and the processor calculates and outputs the angle data, meeting users' needs for spatial orientation measurement. The operating status of all modules is dynamically updated through the human-machine interface module, ensuring users can always monitor the device's operating parameters.

[0113] Reference Figures 1 to 3 In some examples, the control includes at least one of a knob, button, wheel, joystick, touch key, and electromagnetic induction.

[0114] Knob controls are characterized by their intuitive operation and precise adjustment. Users can continuously adjust parameters by rotating the knob, such as gradually increasing or decreasing volume and smoothly adjusting signal strength. Their surfaces are typically designed with anti-slip textures to ensure users can accurately perceive the rotation range during operation and avoid accidental operation due to slipping.

[0115] Button controls are known for their simplicity and efficiency, making them suitable for scenarios requiring quick triggering or switching of functions. For example, recording can be started or stopped with a single press, while mode switching can be done by long press or combination of buttons. Some high-end buttons also integrate pressure-sensing technology, enabling them to perform different operations based on the pressure applied, further expanding their functional boundaries.

[0116] Wheel controls combine the continuous adjustment of a knob with the precise triggering of a button, making them ideal for scenarios requiring multi-dimensional control. For example, during signal adjustment, users can rotate the wheel to adjust basic parameters while simultaneously pressing the center of the wheel to confirm settings or switch adjustment modes. This design significantly improves operational efficiency.

[0117] Joystick controls excel in scenarios requiring spatial positioning or directional control, such as moving a cursor or adjusting device orientation by tilting the joystick forward, backward, left, or right. Their internal six-axis sensor accurately captures the tilting motion, and combined with force feedback technology, provides users with a realistic tactile experience.

[0118] Touch-sensitive controls, with their advantages of no mechanical wear and being waterproof and dustproof, are widely used in devices that require frequent cleaning or operate in harsh environments. Capacitive touch technology ensures responsive operation, while multi-touch functionality supports the simultaneous execution of multiple commands, such as adjusting the display ratio with two fingers or switching interfaces with three fingers.

[0119] Electromagnetic induction controllers represent the future trend of contactless operation. Users can trigger preset functions simply by bringing their hands or specialized tools close to the sensing area. This design has significant advantages in scenarios such as medical or food processing where maintaining equipment cleanliness is crucial, and different levels of access control can be achieved by adjusting the sensing distance.

[0120] In some examples, the detection component 400 includes at least two detection devices. The detection component 400 is used to collect multi-dimensional data of user operations, and the deep learning algorithm unit 500 is used to fuse the sensor data to identify user intentions and actions. The microprocessor 600 controls the corresponding functional module 200 to execute specific control methods based on the recognition results.

[0121] The collaborative operation of various detection devices in the detection component 400 enables comprehensive capture of user operation information from different dimensions. For example, the accelerometer can record dynamic changes in the operation, such as the swing speed of the joystick and the rotation acceleration of the knob; while the position sensor can accurately locate the position of the operation, such as the touch coordinates of the touch key and the offset position of the joystick. This multi-dimensional data complements each other, providing rich and accurate input for the deep learning algorithm unit 500.

[0122] After receiving multi-dimensional data from the detection component 400, the deep learning algorithm unit 500 uses advanced algorithms for fusion processing. By correlating and analyzing data from different sensors, it uncovers hidden user intentions and action patterns. For example, by combining the rotation speed recorded by the accelerometer and the rotation angle determined by the position sensor, it can more accurately determine the user's intention in operating the knob—whether it's a fine adjustment or a significant change in parameters.

[0123] After receiving the user's intent and action results identified by the deep learning algorithm unit 500, the microprocessor 600 responds quickly. Based on different recognition results, it sends precise control commands to the corresponding functional modules 200. For example, if the user intent is to increase the volume, the microprocessor 600 sends a command to the speaker module to increase the volume; if it's to switch recording modes, it sends a corresponding switching command to the recording module. This precise control method ensures that the device can accurately execute various functions according to the user's expectations, improving the device's intelligence level and user experience.

[0124] In some examples, the detection component 400 includes at least one of an accelerometer, a pressure sensor, a displacement sensor, a temperature sensor, a voltage sensor, a current sensor, a gravity sensor, a gyroscope, an infrared sensor, an acoustic detector, and an electromagnetic radar, for collecting user operation data, ambient temperature data, and motion parameters.

[0125] Accelerometers can accurately capture the acceleration of an object, recording changes in acceleration whether it's rotation, linear motion, or complex trajectories. During device operation, they can be used to monitor the acceleration of a knob during rotation, determining whether the user is rotating it rapidly or adjusting it slowly.

[0126] Pressure sensors can detect the amount of pressure applied to a device. When a user presses a button, the pressure sensor outputs a corresponding electrical signal based on the pressure applied, providing crucial data for identifying the user's operational intent. For example, in high-end buttons, pressure-sensing technology enables different pressure levels to correspond to different operational functions.

[0127] Displacement sensors are used to measure changes in the position of objects. In devices, they can accurately determine the touch coordinates of touch keys, the offset position of joysticks, etc., providing the deep learning algorithm unit 500 with important information about the operation position and assisting in judging the user's operation actions.

[0128] Temperature sensors can monitor ambient temperature data in real time. In some temperature-sensitive equipment applications, temperature sensors can feed back ambient temperature information to the system so that the equipment can make corresponding adjustments based on temperature changes, ensuring the normal operation of the equipment.

[0129] Voltage and current sensors are used to measure the magnitude of voltage and current, respectively. In electrical-related functions such as electric shock modules, they can monitor the voltage and current output of the electric shock in real time to ensure that the intensity of the electric shock is within a safe range. They also provide data support for the electric shock safety protection structure, and trigger the protection mechanism immediately upon detecting an anomaly.

[0130] Gravity sensors can detect changes in the orientation and attitude of equipment in a gravitational field. During equipment use, they can help determine the equipment's tilt angle, placement direction, etc., providing a basis for functions that require operational adjustments based on the equipment's attitude.

[0131] Gyroscopes can measure the angular velocity and angular displacement of an object. In scenarios such as joystick control, gyroscopes can accurately capture the angular velocity and angular displacement of the joystick's swing, providing accurate data for recognizing the user's spatial positioning and directional control operations.

[0132] Infrared sensors detect the presence and distance of surrounding objects by emitting and receiving infrared light. In devices, they can be used to achieve contactless operation detection; for example, near electromagnetic induction control components, infrared sensors can help determine if a hand or special tool is near the sensing area.

[0133] Acoustic detectors can emit and receive sound waves, acquiring information by analyzing their reflection and propagation characteristics. In scenarios requiring sound wave detection for operational monitoring or environmental awareness, acoustic detectors can play a crucial role, such as detecting changes in sound intensity and frequency around equipment to help determine the user's operating environment.

[0134] Electromagnetic radar uses electromagnetic waves to detect information such as the position and velocity of targets. In equipment, it can be used to monitor the surrounding environment or assist in electromagnetic induction-based operational control, providing multi-dimensional data support for the intelligent operation of the equipment. These different types of sensors work together to comprehensively collect user operation data, ambient temperature data, and motion parameters, jointly providing strong support for the precise control and intelligent operation of the equipment.

[0135] Each type of detection device in the detection assembly 400 is provided with at least two, and when any detection device fails, it automatically switches to other backup detection devices.

[0136] This dual- or multi-backup detection device design greatly enhances the reliability and stability of the equipment. During actual operation, if a detection device malfunctions or experiences performance degradation, the system can quickly detect this anomaly and automatically and seamlessly switch to other normally functioning backup detection devices. This ensures that the detection component 400 continuously and accurately collects user operation data, ambient temperature data, and motion parameters. For example, if one accelerometer fails due to prolonged use or external interference, the system will immediately activate another backup accelerometer to continue accurately capturing the object's acceleration, ensuring accurate monitoring of operations such as knob rotation and joystick movement. Similarly, if a pressure sensor malfunctions, a backup pressure sensor will promptly take over, accurately sensing the pressure applied to the equipment and providing reliable data for recognizing user intent. This automatic switching mechanism not only avoids equipment malfunctions caused by the failure of a single detection device but also reduces the need for manual user intervention, improving ease of use and maintenance efficiency, and further guaranteeing the equipment's intelligence level and user experience.

[0137] In some examples, the deep learning algorithm unit 500 includes a data preprocessing module and a pattern recognition model.

[0138] The data preprocessing module is used to filter, reduce noise, and extract features from the raw data of the detection component 400.

[0139] The pattern recognition model is optimized through training sample data to improve the accuracy of recognizing ignition mode, aromatherapy mode and other functional modes.

[0140] The Deep Learning Algorithm Unit 500 supports OTA upgrades, enabling online model optimization by receiving algorithm update packages via a wireless communication module.

[0141] The data preprocessing module, as the core of the front-end processing of the deep learning algorithm unit 500, is responsible for the initial optimization of the raw data from the detection component 400. Filtering operations can specifically eliminate high-frequency noise and random interference in the data, ensuring smooth data curves that reflect true operational characteristics. Noise reduction further removes background noise through adaptive algorithms, improving the data signal-to-noise ratio to a usable level. The feature extraction stage uses time-domain analysis and frequency-domain transformation to accurately extract key operational parameters such as rotation angle, pressure value, and displacement trajectory, providing high-quality input for subsequent pattern recognition.

[0142] The pattern recognition model employs a hybrid architecture of convolutional neural networks and recurrent neural networks. Through continuous iterative optimization using millions of training samples, it has achieved accurate classification of over 20 functional modes, including ignition and aromatherapy modes. Its unique multimodal fusion mechanism can simultaneously process multi-dimensional sensor data such as acceleration, pressure, and displacement, maintaining a recognition accuracy of over 98.7% even in complex operating scenarios. A specially designed anti-interference training strategy enables the model to effectively distinguish between intentional operation and environmental disturbances; for example, it can accurately identify knob adjustment intentions even in vibration-affected environments.

[0143] The OTA upgrade function is achieved through a 4G / Wi-Fi dual-mode communication module integrated on the main control board, supporting differential upgrade technology to compress the update package size to 15% of the original model. An A / B partition backup mechanism is used during the upgrade process to ensure automatic rollback to the previous stable version in case of algorithm update failure. The upgraded model not only optimizes the existing pattern recognition logic but also quickly adapts to 200 newly added functional modules through transfer learning technology. For example, the most recent upgrade added support for gesture recognition of electromagnetic induction control components. The entire upgrade process can be completed within 3 minutes without user intervention.

[0144] This disclosure may also include a power supply module, or an external power supply may be used as needed. The power supply module can provide stable power support for various components such as the main body 100, functional module 200, detection component 400, deep learning algorithm unit 500, and microprocessor 600, ensuring normal operation of the device; the external power supply option can be flexibly adapted according to the usage scenario, improving the ease of use and battery life of the device.

[0145] For example, the power module uses a combination of high-density lithium polymer batteries and an intelligent power management chip, which increases energy density by 30% compared to traditional solutions, allowing the device to run continuously for 12 hours under full load. The intelligent power management chip uses dynamic voltage regulation technology to adjust the power supply strategy according to the real-time load of each functional module. For instance, during recording standby, it reduces system power consumption to 18% of the operating state, effectively extending battery life. The module also integrates a triple safety mechanism of overcharge protection, over-discharge protection, and short-circuit protection. When the battery temperature exceeds the safety threshold, it automatically cuts off the power and triggers a buzzer alarm. For scenarios requiring long-term operation, the device has a standard Type-C charging port on the back, supporting 18W fast charging, which can charge the battery from 0% to 80% in 45 minutes.

[0146] In some examples, the pattern detection and motion recognition device also includes a child lock transition module and a human-machine interface module. The child lock transition module is electrically connected to the microprocessor 600 and is used to perform child lock protection operations during the switching of function modules 200. The human-machine interface module is used to receive user commands and display the operating status.

[0147] As a core component of the device's safety mechanism, the child lock transition module employs dual-verification logic to implement child lock protection. When the user manually controls and replaces function module 200, the microprocessor 600 immediately activates the child lock transition module. This module first monitors whether the device is stationary using an accelerometer, and simultaneously confirms whether the operation is applied with normal adult force using pressure sensor data. If abnormal vibration or pressure below a preset threshold is detected, the system automatically locks power supply towards function module 200 and triggers audible and visual warnings to prevent children from accidentally touching electrical contacts at the connection points, which could lead to abnormal device parameters or dangerous operation. After confirming safety, the module sends an unlock signal via an electromagnetic induction controller, allowing normal power supply to the newly switched function module 200.

[0148] The human-computer interaction module integrates a high-sensitivity capacitive touchscreen and multi-color LED status indicators, forming an intuitive operation feedback system. The touchscreen uses 5-point touch technology, supporting complex gesture recognition such as swiping, zooming, and long-pressing. Users can simultaneously monitor 12 key parameters, including acceleration, temperature, and pressure, through a customized interface layout. The LED indicators display the device status in real time using color coding (green / yellow / red). For example, green indicates normal operation, yellow indicates the battery level is below 20%, and flashing red indicates a 400 fault in the detection component. The module's built-in voice interaction function supports bilingual (Chinese and English) command recognition. Users can directly control the device using voice commands such as "increase volume" or "switch mode." The system provides clear voice feedback through a bone conduction speaker, ensuring accurate interaction even in noisy environments. A specially designed vibration feedback mechanism generates tactile cues of varying intensities when operations are confirmed, such as a short vibration when parameters are adjusted correctly and a continuous vibration when errors occur, creating a multimodal interactive experience.

[0149] In some examples, the child lock transition module includes a delay circuit, a contact isolation unit, and a password verification unit. After the mode switching command is triggered, the delay circuit controls the contact isolation unit to activate, connecting the contacts of the new module after a preset delay. Users must deactivate the child lock protection using a preset password or biometric verification to prevent accidental operation by minors and short circuits before module replacement.

[0150] The delay circuit, serving as the core of the child lock transition module's time control, employs a high-precision RC timing element in collaboration with the microcontroller. When the microprocessor 600 receives a mode switching command, it immediately sends a trigger signal to the delay circuit. This circuit precisely calculates the delay time (typically set to 3 seconds) using a charge-discharge curve, ensuring a safe interval between the power outage and power restoration of the functional module 200. During this period, the contact isolation unit remains open via an electromagnetic relay. Its double-break structure design can withstand instantaneous current surges of 220V / 10A, effectively preventing arcing short circuits that may occur during module replacement.

[0151] The password verification unit is integrated into the secure encryption area of ​​the main control chip, supporting dual verification modes of 6-digit numeric password and fingerprint recognition. Upon first use, users must enter a management password via the touchscreen; the system uses the AES-256 encryption algorithm to store the password hash value. After the delay, the module will flash an LED indicator to prompt for verification information. If three consecutive incorrect entries are made, the system will automatically lock and send an alarm signal to the bound mobile phone. For versions supporting biometric recognition, the fingerprint sensor uses capacitive liveness detection technology, capable of penetrating a 0.3mm thick glass cover, maintaining a 99.2% recognition accuracy rate in both dry and wet hand environments. A specially designed emergency unlock mechanism allows administrators to enter recovery mode by pressing and holding the power button + volume down button for 5 seconds, resetting verification information via security questions when the password is forgotten.

[0152] In some examples, each functional module 200 is configured with an independent identification code, and the microprocessor 600 identifies the module type and corresponding control method by recognizing the identification code.

[0153] This identification mechanism uses a non-volatile memory chip (such as EEPROM) to permanently store a unique identifier for the module, and then uses a 128-bit encryption algorithm to generate a dynamic verification key. When a new module is connected, the microprocessor 600 first reads the basic ID from the memory chip via the I2C communication interface, and then sends a random challenge number to trigger the key generation process. The module's built-in encryption coprocessor completes the SHA-256 hash operation within 200ms and returns an encrypted response packet containing a timestamp.

[0154] The main control chip can verify the authenticity of modules by comparing them with a pre-stored module feature library (e.g., a parameter model covering over 200 functional modules). It can also automatically load the corresponding control parameter set. For example, when a new electromagnetic induction module is connected, the system will automatically call its dedicated sensitivity curve (adjustable range of 0.5-5.0N) and accidental touch prevention threshold (≥0.8N trigger), eliminating the need for manual configuration by the user. For third-party compatible modules, the system will activate a safety sandbox mode, limiting its maximum output power to no more than 60% of the rated value. Simultaneously, it will continuously monitor the operating current curve, immediately cutting off power and recording a fault log when abnormal fluctuations (±15% threshold) are detected.

[0155] This mechanism also supports hot-swapping of modules. When replacing modules without power interruption, the identification process is completed within 350ms, ensuring seamless device functionality. A specially designed module aging compensation algorithm continuously tracks the success rate of identification code reading. When the memory chip is detected to have exceeded 100,000 write cycles, it automatically switches to the redundant storage area and prompts the user to replace the module.

[0156] In some examples, functional module 200 also includes decompression module 300, or decompression module 300 is provided on device body 100.

[0157] The decompression module 300 is equipped with vibration feedback or audio-visual interactive control, and the microprocessor 600 activates the corresponding decompression mode based on the user's action recognition results.

[0158] The decompression module 300, a crucial component for enhancing user experience, helps alleviate operational stress through multi-dimensional sensory feedback. When the microprocessor 600 determines, via motion recognition algorithms, that the user is in a high-intensity operational state (such as continuously rotating a knob rapidly more than 5 times or applying pressure exceeding 80% of the rated value), it automatically triggers the decompression mode. The vibration feedback mechanism employs a linear resonant actuator (LRA), capable of generating precise vibration frequencies from 10-200Hz. Through preset "wave-like," "pulse-like," and "gradually increasing" vibration waveforms, it simulates a massage-like tactile experience. For example, when the system detects that the user has been gripping the device for an extended period, it activates a gradual increasing vibration, starting from an initial low frequency of 15Hz and gradually increasing to a high frequency of 120Hz, which automatically decays after 8 seconds, effectively relaxing hand muscles.

[0159] The audio-visual interactive control is achieved through the coordinated operation of a multi-color LED light ring and bone conduction speakers. The light ring uses a circular array of 16 RGB-LEDs, capable of displaying over 16 million color combinations. When entering decompression mode, it changes according to a preset sequence of "breathing light - rainbow gradient - shooting star flash." Simultaneously, the bone conduction speakers play psychoacoustic-optimized natural sound effects, including three scenes: forest birdsong (200-400Hz), babbling brook (500-800Hz), and gentle ocean waves (100-300Hz). The sound pressure level is controlled within the range of 45-55dB, ensuring no auditory interference while creating a sense of spatial immersion through binaural effects. A specially designed intelligent adjustment algorithm dynamically adjusts the brightness of the light ring based on ambient light sensor data, automatically reducing the brightness to 30% in low-light environments to prevent excessive light from affecting the user's relaxation experience.

[0160] The decompression module 300 also supports user-defined decompression schemes, and can store three sets of personalized configurations through the human-computer interaction module. Users can select their preferred vibration intensity (adjustable from 1 to 10 levels), light color (independent adjustment of hue / saturation / brightness), and sound effect type via the touchscreen. The system will encrypt and store the configuration data in non-volatile memory. When a specific operation gesture is detected (such as double-tapping the top of the device repeatedly), the user's preset decompression scheme will be automatically invoked. For enterprise-level application scenarios, the decompression module 300 also integrates a stress monitoring function. By analyzing changes in the user's operation force and frequency, it generates a daily stress index report (0-100 points). When the stress value exceeds 80 points for three consecutive days, relaxation training suggestions will be pushed through the APP, forming a complete stress management closed loop.

[0161] It should be noted that in the pattern detection and motion recognition device, the ignition module 210's ignition start control method covers at least one of pulse ignition, arc ignition, or heating wire ignition, and the microprocessor 600 can automatically select the appropriate ignition method based on the battery power. The aroma diffusion control method of the aroma diffusion module 220 includes ultrasonic atomization, heating evaporation, or compression spray, and the microprocessor 600 will adjust the atomization amount or evaporation rate based on the aroma diffusion liquid level data.

[0162] The human-machine interaction module consists of an LED indicator group and touch buttons. The LED indicator group uses different colors and flashing frequencies to indicate the ignition mode, aromatherapy mode or fault status. This module is used to receive user manual mode switching commands and display the current working status.

[0163] The microprocessor 600 has a built-in fault diagnosis unit for monitoring the operating status of the sensor module, functional module 200, and power supply system. When an anomaly is detected, a protection mechanism is automatically triggered, such as emergency power cut-off, mode lock-up, or fault code reporting. The fault code is fed back to the user in real time through the human-machine interface module.

[0164] The power module is electrically connected to the microprocessor 600 and is used to store recovered braking energy or electrical energy during idle periods. The microprocessor 600 employs a low-power architecture, automatically shutting down unnecessary sensing units in standby mode, retaining only the core wake-up module to reduce static power consumption.

[0165] The microprocessor 600 combines ambient temperature data, aromatherapy liquid level data, and battery power data collected by the detection component 400 to dynamically adjust the output parameters of the functional module 200. Specific control methods for the functional module 200 can be customized via a mobile terminal APP, allowing users to adjust ignition intensity, aromatherapy concentration, or motion recognition sensitivity parameters.

[0166] Although embodiments of the present disclosure have been described in conjunction with the accompanying drawings, those skilled in the art can make various modifications and variations without departing from the spirit and scope of the present disclosure, and such modifications and variations all fall within the scope defined by the appended claims.

Claims

1. A pattern detection and action recognition device, characterized in that, include: The main body of the equipment has an assembly section and a receiving cavity; At least two functional modules, each of which is alternatively and detachably connected to the assembly; A detection component is installed in the receiving cavity and is directly or indirectly electrically connected to the assembled functional module. The detection component is used to collect multi-dimensional data of user operations. A deep learning algorithm unit is disposed in the receiving cavity and electrically connected to the detection component. The deep learning algorithm unit is capable of receiving the detection signal from the detection component. A microprocessor is installed in the receiving cavity and is electrically connected to the detection component, the deep learning algorithm unit, and the assembled functional module, respectively.

2. The pattern detection and action recognition device according to claim 1, characterized in that, At least two of the functional modules include at least one of the following: ignition module, aromatherapy module, display module, speaker module, recording module, signal enhancement module, lighting module, alarm module, electric shock module, and remote control module.

3. The pattern detection and action recognition device according to claim 2, characterized in that, The functional module is an ignition module. The main body of the device is provided with a first control component for controlling the ignition module. The ignition module is provided with an ignition structure. The first control component can turn the ignition structure on or off. Alternatively, when the functional module is an aromatherapy module, the main body of the device is provided with a second control component for adjusting the amount of aromatherapy release, the aromatherapy module is provided with a scenting chip slot, and the second control component controls the concentration or diffusion rate of the aroma emitted by the aromatherapy module. Alternatively, the functional module is a display module, and the main body of the device is provided with a third control component adapted to the display module. The display module is provided with a display area, and the third control component can control the display content of the display area. Alternatively, the functional module is the speaker module, and the main body of the device is equipped with a fourth control component for volume adjustment, which can adjust the volume of the sound emitted by the speaker module; Alternatively, the functional module is a recording module, and the main body of the device is provided with a fifth control element for starting and stopping recording, which can start or stop the recording operation. Alternatively, the functional module is a signal enhancement module, the main body of the device is provided with a sixth control component for controlling the strength of the signal, the signal enhancement module is provided with a signal strength indicator, and the sixth control component can control the strength of the enhanced signal of the signal enhancement module and provide prompts through the signal strength indicator; Alternatively, when the functional module is a lighting module, the main body of the device is provided with a seventh control component for lighting mode, which can switch the on / off state of the lighting module or the lighting mode. Alternatively, the functional module is an alarm module, and the main body of the device is provided with an eighth control component, which can adjust the alarm sensitivity of the alarm module; Alternatively, the functional module is an electric shock module, the main body of the device is provided with a ninth control component, the electric shock module is provided with an electric shock safety protection structure, and the ninth control component can adjust the electric shock intensity of the electric shock module; Alternatively, the functional module is a remote control module, the main body of the device is provided with a tenth control component, the remote control module is provided with a signal communication structure, and the tenth control component can control the opening and closing or frequency of the signal communication structure. The first control element, the second control element, the third control element, the fourth control element, the fifth control element, the sixth control element, the seventh control element, the eighth control element, the ninth control element, and the tenth control element are the same control element, or at least two control elements are integrated into one control element.

4. The pattern detection and action recognition device according to claim 3, characterized in that, The control unit includes at least one of the following: knob, button, wheel, joystick, touch key, and electromagnetic induction.

5. The pattern detection and action recognition device according to claim 1, characterized in that, The detection component includes at least one of an accelerometer, a pressure sensor, a displacement sensor, a temperature sensor, a voltage sensor, a current sensor, a gravity sensor, a gyroscope, an infrared sensor, an acoustic detector, and an electromagnetic radar, and is used to collect user operation data, ambient temperature data, and motion parameters.

6. The pattern detection and action recognition device according to claim 2, characterized in that, The deep learning algorithm unit includes a data preprocessing module and a pattern recognition model; The data preprocessing module is used to filter, reduce noise, and extract features from the raw data of the detection component. The pattern recognition model is optimized through training sample data to improve the accuracy of recognizing ignition mode, aromatherapy mode and other functional modes; The deep learning algorithm unit supports OTA upgrades and receives algorithm update packages through the wireless communication module to achieve online model optimization.

7. The pattern detection and action recognition device according to any one of claims 1 to 6, characterized in that, It also includes a child lock transition module and a human-machine interaction module. The child lock transition module is electrically connected to the microprocessor and is used to perform child lock protection operations during function module switching. The human-machine interaction module is used to receive user commands and display the working status.

8. The pattern detection and action recognition device according to claim 7, characterized in that, The child lock transition module includes a delay circuit, a contact isolation unit, and a password verification unit. After the mode switching command is triggered, the delay circuit controls the contact isolation unit to operate, and connects the contacts of the new module after a preset delay. Users need to deactivate the child lock protection by using a preset password or biometric verification to prevent minors from accidentally operating the lock and to prevent short circuits in the contacts before module replacement.

9. The pattern detection and action recognition device according to any one of claims 1 to 6, characterized in that, Each of the aforementioned functional modules is equipped with an independent identification code. The microprocessor identifies the module type and corresponding control method by recognizing the identification code.

10. The pattern detection and action recognition device according to any one of claims 1 to 6, characterized in that, The functional module further includes a decompression module, or the device body is provided with a decompression module; The decompression module is equipped with vibration feedback or audio-visual interactive control, and the microprocessor activates the corresponding decompression mode based on the user's action recognition result.

Images