System based on energy-saving intelligent water drinking behavior management and excitation device and use method thereof

Abstract

The invention discloses a system based on an energy-saving intelligent water drinking behavior management and excitation device and a use method thereof, and belongs to the technical field of communication, and the system comprises an intelligent behavior recognition module which is used for accurately recognizing the water drinking behavior of a user; the dynamic energy-saving control module realizes fine power consumption control; the behavior prediction and intelligent reminding module is used for predicting a possible time window of a next water drinking behavior; the personalized incentive recommendation module is used for analyzing and layering user states according to the user behavior scoring model, and dynamically recommending incentive statement contents matched with the user states; the point gold coin and business conversion module improves the user participation degree and the platform business conversion efficiency by constructing an incentive distribution model based on a rule engine; the client and network configuration module is used for improving system intelligence, maintainability and user interaction experience. The water drinking recognition accuracy is improved, the false alarm rate is remarkably reduced, the system integration degree and the user experience are improved, and the system is suitable for various scenes such as offices, families and gifts.

Description

Technical Field

[0001] The present invention relates to a system based on an energy-saving intelligent drinking water behavior management and incentive device and a method for using the system, belonging to the field of communication technology, and specifically to the field of intelligent health hardware and Internet of Things communication technology. Background Art

[0002] As people's health awareness increases, scientific drinking water, as an important part of daily behavior management, has received more and more attention. Currently, there are some drinking water reminder devices on the market, but they mainly have the following problems: 1. Rough identification mechanism: Many devices rely on fixed time reminders or simple weight change judgments, which can easily lead to misjudgments or missed judgments, making it difficult to accurately identify actual drinking behavior.

[0003] 2. Unreasonable power consumption control: Most existing products use Bluetooth continuous connection or high-frequency detection strategies, which have high power consumption and poor battery life. Users need to charge frequently, affecting the user experience.

[0004] 3. Lack of intelligence in reminder methods: Reminder mechanisms are mostly timed and cannot be personalized based on user drinking habits, resulting in low reminder efficiency and acceptance.

[0005] 4. Lack of behavioral incentive closed loop: Most devices only provide recording functions, which cannot continuously motivate users to form good drinking habits, and it is difficult to achieve user stickiness and commercial conversion.

[0006] 5. Fragmentation of system functions: There are problems such as insufficient coordination between software and hardware, data transmission delays or imperfect synchronization mechanisms. Some devices lack independent clients or have low functional integration, resulting in a fragmented overall experience and difficulty in forming a unified behavior management closed loop. Summary of the Invention

[0007] The present invention proposes a system based on an energy-saving intelligent drinking water behavior management and incentive device and its use method. The system is specifically applied to an energy-saving intelligent drinking water behavior management and incentive device that integrates weight sensing recognition, low-power control, WiFi communication, user behavior modeling, a dynamic reminder mechanism, and personalized incentive feedback. The system collects user drinking behavior through sensors, realizes automatic recognition, recording, and data upload, and provides predictive reminders, incentive feedback, and gold coin rewards based on the user's historical behavior rhythm, forming a complete health behavior intervention closed loop. The system is suitable for various scenarios such as smart office, family health management, behavior habit cultivation, and health gifts. It belongs to the category of health Internet of Things devices and has good system coordination, energy saving, and user guidance effects. The system has intelligent recognition, energy-saving control, behavior prediction, personalized incentive, and system closed loop capabilities. It can help users improve their drinking habits in a more scientific, intelligent, and energy-saving manner. Through the deep integration of weight sensing recognition, behavior modeling algorithm, dynamic reminder mechanism, low-power scheduling strategy, and personalized incentive system, a complete closed loop from "behavior recognition" to "habit guidance" is constructed.

[0008] The technical solution adopted by the present invention to solve the technical problem is: a system for energy-saving intelligent drinking water behavior management and incentive device, the system comprising: an intelligent behavior recognition module, a dynamic energy-saving control module, a behavior prediction and intelligent reminder module, a personalized incentive recommendation module, a gold coin and business conversion module, and a client and network configuration module; The intelligent behavior recognition module is used to accurately identify the user's drinking behavior; The dynamic energy-saving control module implements refined power consumption control based on the task priority scheduling mechanism and the predictive wake-up strategy; The behavior prediction and intelligent reminder module is used to predict the possible time window for the next drinking behavior; The personalized incentive recommendation module analyzes and stratifies user status according to the user behavior scoring model and dynamically recommends incentive statement content that matches the user; The gold coin and business conversion module collects user drinking behaviors and scores tasks in real time, building an incentive distribution model based on a rule engine to improve user engagement and platform business conversion efficiency. The client and network configuration module is used to improve the system intelligence, maintainability and user interaction experience, and realizes the data linkage, function configuration and user behavior feedback closed loop between the device end and the client.

[0009] Furthermore, the intelligent behavior recognition module includes: real-time data collection based on high-precision weight sensors and analog-to-digital converters, combined with a multi-dimensional behavior judgment algorithm to accurately identify the user's drinking behavior, support self-learning strategies, and adapt and optimize the user's different drinking habits. Among them, the multi-dimensional behavior judgment algorithm includes: using a sliding time window mechanism to achieve real-time judgment. Within each window period, the system continuously collects weight sensor data and constructs a behavior sequence.

[0010] Furthermore, the dynamic energy-saving control module includes: setting different levels of priority according to current and estimated tasks, the device is in deep sleep state by default, and is only awakened once in ultra-low power consumption mode within the set detection cycle. After waking up, the system determines whether there is a high-priority task. If there is no task, it quickly returns to sleep. If the task judgment is established, it enters light sleep / running state and loads the required modules to prepare for execution.

[0011] Furthermore, the behavior prediction and intelligent reminder module includes: building a drinking rhythm model based on the user's historical drinking behavior data, using a sliding time window to extract high-frequency drinking time periods in the past N days to form an individualized drinking rhythm vector, and predicting the possible time window for the next drinking behavior by calculating the time deviation between the current time period and the rhythm vector. When the deviation reaches the set threshold, the system triggers the reminder logic. To avoid interference from unexpected time periods, the system supports users to set the daily reminder start and end time periods. All reminder logic is only effective within this time window. Beyond this time period, even if the deviation logic meets the conditions, the system will not trigger any reminders.

[0012] Furthermore, the client and network configuration module include: Multiple WiFi configurations and automatic switching mechanism: Supports multiple WiFi access points preset and save. The system automatically reconnects to the optimal network when the signal is interrupted or the environment changes. The first connection supports turning on AP mode through the device, and the client can scan the code or bind the network with one click. Network interruption data caching and breakpoint resumption: When the device is offline, user behavior data will be cached in the local storage module and compressed and uploaded when the network is restored, reducing communication energy consumption and ensuring data integrity. Client core functional modules: Real-time viewing: Users can view drinking behavior statistics, consecutive days of meeting the standard, and gold coin balance in the app; Custom settings: support setting daily reminder time, whether to enable holiday reminders, and motivational statement style preferences; Incentivized interaction: Users view personalized incentives, like / ignore statements, and influence subsequent recommendation models; Remote OTA upgrade: The system supports issuing firmware update instructions through the client, and the device automatically connects to the server to complete the differential upgrade; User feedback mechanism: built-in problem reporting / abnormal feedback channels for continuous optimization of cloud strategies and client behaviors; Policy issuance and remote linkage mechanism: The system supports sending relevant content from the cloud based on user behavior tags and behavior frequency to achieve system-level adaptive adjustment.

[0013] The present invention also proposes a method for realizing an energy-saving intelligent drinking water behavior management and incentive device system, the method comprising the following steps: Step 1: After powering on, read the local WiFi configuration. If the device is connected to the network successfully, it will enter the deep-sleep state by default. If it fails, it will automatically enter the AP mode and wait for the user to scan the code to connect to the network. Step 2: Wake up the device, collect data every second through the weight sensor, and use the sliding time window algorithm to determine whether it is a valid drinking behavior; Step 3: If it is identified as drinking water behavior, the system generates a record and executes the upload logic; Step 4: Construct an individual drinking rhythm vector based on the user's 7-day drinking record and identify high-frequency periods; Step 5: Build a user behavior scoring model, generate a behavior vector based on the scoring results, and assign it to a strategy pool. Then, based on factors such as time period, holidays, and task status, push corresponding incentive statements. Step 6: If the customer completes the specified behavior, he or she will also receive gold coins and badges as rewards.

[0014] Furthermore, in step 2, using a sliding time window algorithm to determine whether the drinking behavior is effective includes: 1. Within each 60-second sliding time window, the system continuously collects water cup weight data and compares the current weight with the weight of the last drinking record; 2. Detect whether there are n consecutive times (n ≥ 3) of weight decrease or increase, each change exceeds 10 grams (ΔW ≥ 10g); 3. Check whether the deviation of these weight changes is within 10 grams (ΔW ≤ 10g); 4. If the above conditions are met, the system will mark the drinking behavior as valid and generate a drinking record; 5. If the weight variation deviation of three consecutive samples does not exceed 10 grams after the record is generated, the system stops further testing; 6. If the device does not detect any effective drinking behavior or operation within 10 seconds, the device will enter sleep mode to save power.

[0015] Furthermore, step 4 includes: the cloud service generates an individualized drinking rhythm model and reminder time list based on the user's historical drinking behavior, and sends the reminder strategy to the device side when the user uploads the data for the first time. The device stores the reminder schedule locally and carries the local reminder version number for comparison by the cloud side each time it uploads. If the version is inconsistent, the system will automatically update the reminder schedule to ensure that the reminder logic is synchronized in real time. Both the device side and the App client have synchronization capabilities: after the user synchronizes the drinking record through the App, if it is detected that the interval between the next reminder point is greater than the threshold, it will automatically update to the next reminder time point. In the local reminder triggering process, the device wakes up the main control chip by the RTC timer, executes the LED prompt task, and then returns to the sleep state; the App side reminder process is triggered by the cloud task timer scheduling, and reminds the user to drink water through push.

[0016] Furthermore, the dimensions of the user behavior scoring model constructed in step 5 include: reminder response time, consecutive check-in days, target-reaching frequency, active cycle and completion rate. The motivational language styles are divided into: morning greetings, emotional motivation, achievement feedback, flexible encouragement, and holiday greetings. Users set display preferences and frequency. The client supports liking / ignoring motivational language. The system records and provides feedback to optimize subsequent recommendation strategies, forming an intelligent optimization closed loop of behavior → motivation → feedback → re-recommendation.

[0017] The beneficial effects of the present invention are: 1. The present invention can improve the accuracy of drinking water recognition and significantly reduce the false alarm rate.

[0018] 2. The present invention can extend the life cycle of the equipment and reduce the frequency of user maintenance.

[0019] 3. The present invention can provide personalized reminders and incentive mechanisms to promote habit formation.

[0020] 4. The present invention can establish a closed loop of "identification → reminder → incentive → monetization", which is convenient for business expansion.

[0021] 5. The present invention can improve system integration and user experience, and is suitable for various scenarios such as office, home, and gifts. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] Figure 1 This is a schematic diagram of the MCU control circuit of the device of the present invention.

[0023] Figure 2 This is a schematic diagram of the charging management circuit of the device of the present invention.

[0024] Figure 3 This is a schematic diagram of the weighing detection circuit of the device of the present invention.

[0025] Figure 4This is a flowchart of the initialization and binding of the device of the present invention.

[0026] Figure 5 This is a flow chart of data collection and uploading of the device of the present invention.

[0027] Figure 6 This is a flow chart of low-power consumption and long-endurance control of the device of the present invention.

[0028] Figure 7 This is a multi-channel drinking water reminder control flow chart of the device of the present invention.

[0029] Figure 8 This is a flow chart of personalized motivational language strategy matching and feedback optimization for the device of the present invention.

[0030] Figure 9 This is a flow chart of the behavior-driven incentive growth and gold coin linkage mechanism of the device of the present invention.

[0031] Figure 10 This is a flow chart of the advertising behavior linkage gold coin incentive and control of the device of the present invention. DETAILED DESCRIPTION

[0032] The present invention will be described in further detail below with reference to the accompanying drawings.

[0033] like Figures 1 to 3 As shown in the figure, the energy-saving intelligent drinking behavior management and incentive device of the present invention comprises: a main control chip module, a weight sensing module, a wireless communication module, a power management module, a reminder module and a supporting client. The structure and function of each part are as follows: 1. Main control chip Use an MCU (such as ESP8266) that supports low-power control and WiFi communication, with deep-sleep mode and interrupt wake-up capabilities to achieve scheduling control of behavior recognition and communication tasks.

[0034] 2. Weight sensing system It consists of strain gauges and high-precision ADC (such as CS8M320), supports real-time acquisition of weight changes from 0g to 5kg, and supports behavioral trend judgment algorithms.

[0035] 3. Wireless communication Supports automatic switching of multiple WiFi groups, and enters AP mode when powered on for the first time (see attached Figure 3 ), users scan the QR code through the client to bind the device MAC address and configure the WiFi information to complete the network connection and identity registration.

[0036] 4. Reminder system It includes local LED light reminders, combined with client APP push and voice prompts to form a three-dimensional reminder system.

[0037] 5. Power Management Using a 750mAh lithium battery and combining it with a module-level low-power strategy, it can support continuous operation for more than 25 days in typical scenarios.

[0038] 6. Client part It provides functions such as water drinking data visualization, reminder time setting, motivational statement preference management, WiFi configuration and OTA remote upgrade.

[0039] in, Figure 1 The diagram shows the circuit structure of the main control chip, power supply, communication, antenna connection, etc., which is used to realize the control and communication functions of the device operation logic; Figure 2 The input and output relationship, indication function and charging status control logic of the lithium battery charging module are illustrated; Figure 3 The diagram illustrates the acquisition and analog-to-digital conversion of weight sensor signals, the communication interface, and the connection method with the main control chip.

[0040] The present invention proposes a system based on an energy-saving intelligent drinking water behavior management and incentive device, which includes an intelligent behavior recognition module, a dynamic energy-saving control module, a behavior prediction and intelligent reminder module, a personalized incentive recommendation module, a gold coin and business conversion module, and a client and network configuration module; wherein: The intelligent behavior recognition module is based on real-time data collection from a high-precision weight sensor and an analog-to-digital converter (ADC), combined with a multi-dimensional behavior judgment algorithm (such as time window, weight change trend, repetition frequency, and sequence pattern) to accurately identify the user's drinking behavior. The module supports self-learning strategies and adapts and optimizes users' different drinking habits.

[0041] The multi-dimensional behavior determination algorithm uses a sliding time window mechanism to achieve real-time judgment. Within each window period (for example, 60 seconds), the system continuously collects weight sensor data and constructs a behavior sequence. The algorithm combines the following multiple indicators for judgment: 1. Within each 60-second sliding time window, the system continuously collects water cup weight data and compares the current weight with the weight of the last drinking record; 2. Detect whether there are n consecutive times (n ≥ 3) of weight decrease or increase, each change exceeds 10 grams (ΔW ≥ 10g); 3. Check whether the deviation of these weight changes is within 10 grams (ΔW ≤ 10g); 4. If the above conditions are met, the system will mark the drinking behavior as valid and generate a drinking record; 5. If the weight variation deviation of three consecutive samples does not exceed 10 grams after the record is generated, the system stops further testing; 6. If the device does not detect any effective drinking behavior or operation within 10 seconds, the device will enter sleep mode to save power.

[0042] When the above-mentioned characteristic judgment criteria are met, the system automatically marks it as "drinking behavior" and records the timestamp. This window algorithm has the ability to filter out behavioral noise, effectively eliminating minor weight fluctuations not caused by drinking, such as desktop vibrations and accidental touches, thereby improving recognition accuracy and reducing system power consumption and false alarm rates. This module also has adaptive learning capabilities, dynamically adjusting the window length and judgment parameters based on the user's historical behavior to adapt to different drinking frequencies and habits.

[0043] The dynamic energy-saving control module implements refined power consumption control based on the task priority scheduling mechanism and predictive wake-up strategy, specifically including the following: Different priority levels are set based on current and predicted tasks (such as water intake identification, data upload, incentive feedback, and client interaction). The device defaults to a deep sleep state and wakes up only once in ultra-low power mode within a set detection period. After waking up, the system determines whether there are high-priority tasks (for example, if the predicted time of water intake is approaching). If not, it quickly returns to sleep. If the task is determined to be successful, it enters a light sleep / running state, loading the required modules in preparation for execution. The WiFi communication module is disabled by default during non-essential tasks and only activates briefly when a data upload task is scheduled. It then shuts down immediately after the upload is complete, significantly reducing wireless communication power consumption. This mechanism, through a "behavior prediction-triggered wakeup + on-demand communication module activation" strategy, achieves event-driven energy-saving scheduling. In typical usage scenarios (e.g., drinking water 10 times per day), the device can achieve stable battery life of over 25 days of continuous operation.

[0044] The behavior prediction and intelligent reminder module builds a "drinking rhythm model" based on the user's historical drinking behavior data. It uses a sliding time window to extract high-frequency drinking time periods within the past N days (such as 7 days) to form an individualized drinking rhythm vector. By calculating the time deviation between the current period and the rhythm vector, it predicts the possible time window for the next drinking behavior.

[0045] When the deviation reaches a set threshold (e.g., if you haven't drunk water for more than 20 minutes), the system triggers a reminder. Reminders include: ① local LED lighting; ② client app push notifications. The system also supports adaptive optimization of the rhythm model based on user feedback (e.g., whether or not you drank water after the reminder).

[0046] To avoid unexpected interruptions during unexpected periods, the system allows users to set a daily reminder start and end time window (e.g., 9:00 AM - 8:00 PM). All reminder logic only takes effect within this time window. Outside of this time window, even if the deviation logic conditions are met, the system will not trigger any reminders, thus ensuring a controllable user experience.

[0047] In addition, the system suspends water drinking reminders by default during holidays to prevent unnecessary interruptions to users on non-working days. Users can manually enable "holiday reminder mode" through the client. Once enabled, the system will continue to perform prediction and reminder tasks according to the established rhythm model during holidays.

[0048] The personalized incentive recommendation module analyzes and stratifies user status based on the user behavior scoring model, and dynamically recommends matching incentive statements. The scoring model includes: 1. Response time (e.g., average delay from reminder to drinking water); 2. Frequency of meeting targets (e.g., number of days in the past seven days on which drinking water targets were met); 3. Continuity (e.g., the longest consecutive number of days of clocking in); 4. Active cycle; 5. Completion rate of tasks for the day.

[0049] The system generates user behavior vectors based on the above factors and, combined with user tags (such as "self-disciplined," "volatile," and "low-frequency"), assigns them to corresponding incentive strategy pools. It also incorporates external contextual information into incentive strategy scheduling, including: 1. Time period: such as morning, afternoon, and night → Suitable for "Good Morning Sayings" and "Good Night Sayings"; 2. Date type: such as holidays and weekdays → corresponding to "soft encouragement" or "emotional companionship" statements; 3. Seasonality: such as differences in drinking water reminder styles in winter and summer; 4. Completion dimension: whether the task of the day has been completed or not → use "achievement feedback" or "reminder guidance"; Motivational phrases are categorized into multiple subcategories, such as morning messages, night messages, health reminders, emotional motivation, achievement encouragement, positive feedback, and soft encouragement. The system extracts and pushes these phrases from the corresponding corpus based on strategic matching rules. The client allows users to set display frequency, display time, and phrase style preferences.

[0050] The system also supports user feedback mechanisms (such as likes and ignores), and uses user preferences for subsequent sentence selection strategies, forming an intelligent optimization closed loop of "behavior → incentive → feedback → re-recommendation".

[0051] The gold coin and business conversion module collects user drinking behaviors in real time and scores tasks, building an incentive distribution model based on a rule engine to improve user engagement and platform business conversion efficiency. Specifically, it includes the following: ① Gold coin trigger mechanism The system triggers virtual coins and badge rewards based on the following behaviors: 1. Daily drinking water meets the standard; 2. Punch in the card for a specified number of consecutive days; 3. Respond to reminders on time; 4. The first time a specific function is used or a certain type of task is completed; 5. Each type of behavior is scored according to the weight defined by the system (for example: drinking water on time +1 gold coin, three consecutive days +2 gold coins). The daily upper limit of gold coins can be configured and managed by the cloud.

[0052] ② User account and rating system After the user's gold coins and badges are accumulated, they will be recorded in the client account system, which supports: 1. Check the acquisition history; 2. Current level and progress; 3. Compare with the ranking data of the same city; ③ Gold Coin Usage and Exchange Rules Once the user's gold coins reach the set threshold, the following operations can be performed in the client: 1. Redeem virtual items (such as theme skins, motivational voice packs, leaderboard badges, etc.); 2. Initiate a withdrawal operation. The withdrawal rules are set by the platform (e.g. if the gold coins are ≥ 500, you can apply for a withdrawal of 2 yuan); 3. Triggering a check-in / check-in re-sign-in: If a user fails to complete the water check-in on a certain day, they can use gold coins to re-sign in to restore the streak; after re-signing in, they can continue to participate in the consecutive days reward or ranking statistics.

[0053] 4. The re-signing mechanism supports daily / periodic limit and the threshold can be flexibly adjusted according to the user level.

[0054] ④ Advertising monetization linkage mechanism Advertising is the key entry point for platform business conversion. The system supports the following linkage logic: 1. When users are completing a task, the system may recommend the option of "watching ads to receive coins"; 2. Supports the inclusion of events such as ad exposure, clicks, and stays into the gold coin distribution logic; 3. Dynamically recommend advertising content based on user drinking behavior tags, compliance rates, and preferred time periods (e.g., "users who don't meet standards in summer" → recommended tea brands); 4. Advertising campaign strategies (such as limited-time double gold coins) are dynamically released by the platform.

[0055] ⑤ Platform-adjustable strategy engine The gold coin distribution mechanism, advertising placement, sign-in conditions, daily limits, etc. are all controlled by the cloud-based rule engine and can be dynamically adjusted based on user stratification, operational goals, and holiday strategies to ensure that the incentive logic is consistent with business goals.

[0056] The client and network configuration module serve as the core channels for device-user interaction, cloud synchronization, and remote management. They have the following functions and technical mechanisms: ① Multiple WiFi configurations and automatic switching mechanism Supports multiple WiFi access points (SSID) presets and saves, allowing the system to automatically reconnect to the optimal network when the signal is interrupted or the environment changes. For the first connection, you can enable AP mode on the device, and scan the QR code on the client or connect to the network with one click.

[0057] ② Network interruption data caching and breakpoint resumption When the device is offline, user behavior data (such as drinking events and reminder responses) will be cached in the local storage module and uniformly compressed and uploaded when the network is restored, reducing communication energy consumption and ensuring data integrity.

[0058] ③Client core function module Real-time viewing: Users can view drinking behavior statistics, consecutive days of meeting the standard, gold coin balance, etc. in the App; Custom settings: support setting daily reminder time, whether to enable holiday reminders, motivational statement style preferences, etc. Incentivized interaction: Users can view personalized incentives, like / ignore statements, and influence subsequent recommendation models; Remote OTA upgrade: The system supports issuing firmware update instructions through the client, and the device can automatically connect to the server to complete differential upgrades; User feedback mechanism: Built-in problem reporting / abnormal feedback channel for continuous optimization of cloud strategies and client behavior.

[0059] ④ Policy issuance and remote linkage mechanism The system supports issuing "reminder frequency optimization strategy", "motivational language display rhythm", "gold coin exchange task recommendation" and other content from the cloud based on user behavior tags and behavior frequency to achieve system-level adaptive adjustment.

[0060] This module realizes the data linkage, function configuration and user behavior feedback closed loop between the device and the client, improving the system intelligence, maintainability and user interaction experience.

[0061] The present invention also proposes a typical embodiment to further illustrate the working principle of the present invention and its specific implementation in terms of drinking behavior detection, data uploading, power consumption optimization, reminder mechanism and behavior incentive. The following describes a typical embodiment in combination with key technical modules: like Figures 4 to 6 As shown in the figure, a behavior recognition and power consumption control process is proposed, which specifically includes: reading the local WiFi configuration (supporting up to 5 groups) after the device is powered on, and automatically entering AP mode if it fails and waiting for the user to scan the code to configure the network.

[0062] After successfully connecting to the Internet, the device enters the deep-sleep state by default and wakes up only in the following situations: 1. Detect weight change; 2. Reaching the reminder time point or the behavior is predicted to be at a high incidence time; 3. Execute scheduled tasks (such as power detection and OTA upgrade); 4. Manual operation by the user.

[0063] After waking up, the device collects data every second through the weight sensor and uses a sliding time window mechanism to determine in real time whether it is a valid "drinking behavior": Within each 60-second sliding time window, the system continuously collects water cup weight data and compares the current weight with the weight of the last drinking record.

[0064] Detect whether there are n consecutive times (n ≥ 3) of weight loss or gain, each change exceeds 10 grams (ΔW ≥ 10g).

[0065] Check that these weight changes are within 10 grams (ΔW ≤ 10g).

[0066] If the above conditions are met, the system will mark the drinking behavior as valid and generate a drinking record.

[0067] If the weight variation deviation of three consecutive samples does not exceed 10 grams after the record is generated, the system stops further testing.

[0068] If the device does not detect any effective drinking behavior or operation within 10 seconds, it will enter sleep mode to save power.

[0069] The system has the ability to filter out noise and adaptively adjust parameters, eliminating interference such as desktop vibration and accidental touch, and improving recognition accuracy.

[0070] If drinking is detected, the system generates a record and executes the upload logic: if the network is available, WiFi is temporarily activated for upload and then shut down; if the network is disconnected, the data is cached locally and subsequently uploaded. All communication modules are disabled by default and only temporarily activated when a task is triggered, effectively extending the device's battery life.

[0071] in, Figure 4This diagram illustrates the network configuration and binding process after the device is powered on for the first time. When the device is powered on, it automatically enters AP (hotspot) mode and turns on the WiFi function, waiting to connect to a mobile device (S1). The user manually connects to the hotspot created by the device using their phone (S2). After the client program detects the connection status, it automatically obtains the device's unique MAC address through a preset interface and uploads it to the server, completing the binding operation between the user identity and the device address (S3). The user then enters and sets the target WiFi network information through the client (S4). The system synchronizes this configuration information to the device through the configuration interface and stores it locally (S5). The device automatically restarts and attempts to connect to the network. Upon successful connection, it transmits the device status and hardware information back to the server, completing the initialization process (S6). This process enables rapid deployment and unique identity binding for the device during first use, providing basic connectivity support for subsequent functions such as drinking behavior recognition, reminder push, and policy synchronization. Figure 5 This demonstrates how the device wakes up via a weight change interrupt, combines a behavior recognition algorithm to determine the user behavior type, and controls data recording and upload based on the recognition results. Specifically, when the device is in deep sleep, the weight sensor module detects a weight change, triggering the system to wake up. The system continuously collects multiple sets of weight data to construct a change sequence. Using set thresholds and pattern recognition logic, it determines whether the action represents "drinking water," "adding water," or "no operation." If the action is deemed valid, the system records the event data and, based on the current network status, selects a strategy for real-time upload or local caching. If the network is available, the system activates the WiFi module for data upload and, upon successful upload, deactivates the communication module to conserve energy. If the upload fails or the network is disconnected, the data is cached in a local queue and uploaded in batches via subsequent scheduled tasks. Furthermore, the system supports detecting any unuploaded data when triggered by periodic tasks such as battery level checks, and initiates a catch-up upload. After the upload is complete, the server synchronizes the uploaded data with the user behavior model, reminder schedule, and incentive strategy, achieving closed-loop data management. Figure 6 It demonstrates how the device can achieve an efficient operation process of "deep sleep → event wake-up → multi-tasking → on-demand networking → back to sleep" based on a multi-source event triggering mechanism and high-priority task queue scheduling. The device is in a deep low-power sleep state by default. The main control chip responds to events such as weight changes, scheduled tasks or user operations through an interrupt mechanism to wake up the system and enter the running state. After waking up, the system executes high-priority tasks including drinking behavior recognition, power detection, local data processing, OTA upgrade detection, etc. If the task requires synchronization of data with the server, the WiFi module is temporarily started for network upload operation. After completion, the WiFi is immediately turned off and the device enters the sleep state again. If there is no current need to connect to the Internet, it will directly return to low-power mode after completing the task locally. This process significantly optimizes the overall energy consumption of the system through multi-task integration and communication on-demand strategy, achieving a battery life of more than 25 days in typical usage scenarios.

[0072] like Figure 7 The figure shows a complete reminder mechanism, integrating cloud-based rhythm model construction, reminder policy issuance, local execution, and app push notifications. Based on the user's historical drinking behavior, the cloud service generates a personalized drinking rhythm model and reminder schedule. The reminder strategy (including time points, expiration dates, and version numbers) is also sent to the device when the user uploads their first data. The device stores the reminder schedule locally and, with each subsequent upload, includes the local reminder version number for cloud-based comparison. If a version inconsistency is detected, the system automatically updates the reminder schedule to ensure real-time synchronization of the reminder logic. Furthermore, both the device and the app client have synchronization capabilities: after a user synchronizes their drinking history through the app, if the next reminder interval exceeds a threshold (e.g., 30 minutes), the next reminder time is automatically updated. In the local reminder triggering process, the device's RTC wakes up the main control chip, executes the LED reminder task, and then returns to sleep. The app reminder process is triggered by a cloud-based task scheduler, sending a push notification to remind the user to drink water. This mechanism implements dual closed-loop control of local reminders and cloud strategies. While ensuring the timeliness of reminders, it also takes into account energy-saving operation and personalized guidance to improve user behavior compliance.

[0073] The system constructs an "individual drinking rhythm vector" based on the user's historical 7-day drinking records and identifies high-frequency time periods.

[0074] If the current period deviates from the rhythm by more than 20 minutes, a reminder will be triggered: Local reminder: through LED lighting effects; Client reminder: App push voice / pop-up prompt.

[0075] All reminder logic is only effective within the time window set by the user (such as 09:00~20:00). Reminders are suspended by default on holidays, and manual activation of "holiday reminder mode" is supported.

[0076] The cloud sends reminder policies (including time point, version number, validity period, etc.), and the device performs version comparison every time it uploads to ensure consistent reminder rhythm.

[0077] like Figure 8The figure illustrates how the system constructs individual labels based on user behavior scores, then matches adaptive phrases within a multidimensional strategy pool. Content and weighting are optimized based on user interaction feedback, forming an intelligent closed-loop mechanism of "behavior-recommendation-feedback-adjustment." Specifically, the system first quantifies and scores user behavior data (such as reminder response timeliness, target-compliance frequency, and consistency) to generate multidimensional behavior vectors. Based on these vector features, users are categorized into pre-set behavioral label categories (such as "self-disciplined" and "volatile"). The system then matches adaptive phrases within a multidimensional incentive strategy pool based on the label, current time period, holiday status, water drinking completion status, and season, and displays incentive content through the client interface. Users can interact with incentive phrases (such as "like" or "ignore"). Feedback is recorded and influences subsequent content recommendation strategies. Users can adjust the style, frequency, and preferences of displayed content to enhance their interactive experience. The system also supports periodic cloud-based delivery of new incentive strategy pool content and adjustment of strategy weights to ensure dynamic adaptability and contextual relevance. This process realizes the dynamic linkage of user behavior identification, personalized content matching and strategy feedback optimization, effectively improving the motivation and sustainability of users to develop drinking habits.

[0078] The system builds a user behavior scoring model based on the following dimensions: 1. Remind response time; 2. Number of consecutive days of clocking in; 3. Frequency of reaching the target; 4. Active cycle and completion rate.

[0079] Based on the scoring results, a behavior vector is generated and assigned to a strategy pool (such as "self-disciplined" and "low-frequency"). Corresponding motivational messages are then delivered based on factors such as time period, holidays, and task status. Motivational messages are categorized into morning messages, emotional motivation, achievement feedback, soft encouragement, and holiday greetings. Users can set their display preferences and frequency. The client supports liking or ignoring motivational messages. The system records and provides feedback to optimize subsequent recommendation strategies, forming an intelligent optimization loop of "behavior → motivation → feedback → re-recommendation."

[0080] like Figures 9 and 10 As shown, users will be rewarded with coins and badges by completing the following actions: Drinking water that meets standards; Continuous clocking in; Holiday tasks; Ad viewing behavior.

[0081] Rewards are distributed based on the weight of the behavior (e.g., +1 for drinking water on time, +2 for three consecutive days). Rules such as the gold coin limit and the conditions for re-signing are dynamically controlled by the cloud-based strategy engine.

[0082] Gold coins can be used to: Cash withdrawal (if 100 gold coins are reached, they can be exchanged for 2 yuan); Redeem virtual items (such as skins, voice packs); Make up the sign-in operation (restore the punch-in continuity).

[0083] Advertising incentive strategy: Advertising tasks are presented on the task page in the form of "watch videos to get gold coins", etc., and participation is not mandatory; The system pushes advertising tasks based on user behavior profiles and records behavioral data such as clicks, stays, and skips; Advertising behavior affects the weight of gold coin rewards and the subsequent distribution of advertising types; Support activities such as "Limited-time Double Gold Coins" to achieve coordinated linkage between commercialization and incentive systems.

[0084] The client of the present invention supports multiple user interaction and system configuration functions: Check drinking trends, clock-in records, and gold coin balances; Set reminder time period, sentence style and gold coin usage preferences; Submit motivational feedback and report exceptions; Receive and execute OTA firmware upgrades (differential updates); Synchronize behavioral data and policy content with the cloud in real time.

[0085] The cloud has a user stratification strategy engine that can dynamically issue policies based on behavior tags and scores: Personalized reminder rhythm; The motivational speech shows rhythm and style; Ad recommendation priority; Gold coin and task distribution logic.

[0086] The device side combines the reminder time version number with the behavioral data for two-way synchronization to ensure the consistency and intelligence of the system behavior management logic, and build a "cloud-end-human" three-in-one health intervention closed loop.

[0087] in, Figure 9This system illustrates how it achieves positive incentives and sustained intervention in users' water drinking behavior through goal guidance, badge level rewards, coin distribution, and user account growth feedback. Users can see in-app notifications about their water drinking goals, such as "Complete one more water drink to earn a blue badge + coins," fostering anticipation and proactive motivation. The system assesses completed tasks and, if incentive conditions are met (such as consecutive check-in days or completion of holiday tasks), issues a badge of the corresponding level, along with semantic feedback (such as "Keep it up, healthy guy!"). Each badge is associated with a fixed coin reward (e.g., +2 for green, +5 for blue, +10 for holiday special badges), which is directly credited to the user's account and updated with badge records and coin balances. The client also displays feedback on users' accumulated achievements, growth levels, and badge progress, reinforcing their sense of accomplishment. Users can use coins to withdraw, redeem virtual items, and re-check in. The system also supports cloud-based delivery of personalized challenges and limited-time incentive events (such as "3-Day Challenge" and "Double Coin Day") to further encourage users to develop water drinking habits and maintain their active behavior. This incentive mechanism improves user behavior compliance and achieves long-term maintenance of health intervention effects through the positive closed-loop logic of "behavior → goal guidance → immediate rewards → account growth → re-guidance". Figure 10 This demonstration demonstrates how the system seamlessly integrates advertising tasks with an incentive system, while ensuring a robust user experience. This system achieves the dual goals of a non-intrusive monetization path and personalized behavior guidance. Advertising incentive tasks are centrally configured by the cloud-based advertising task configuration center. These tasks include information such as ad content type, gold coin reward value, behavioral tags for the task triggering users (e.g., those with poor reminder response or insufficient water intake), and delivery frequency. The system accurately categorizes users based on their historical behavior and pushes appropriate advertising tasks, achieving "tag-driven" differentiated ad distribution.

[0088] Unlike traditional interruptive advertising, the present invention adopts a "task-based incentive advertising" strategy. The advertising content is presented in the form of an optional entrance in the task list or gold coin page. Users can choose whether to participate independently to avoid interrupting the user's main process operation. Before the advertisement is triggered, the system clearly informs the user of the gold coin value, reward props or limited badges that can be obtained after completing this advertisement, thereby improving the user's cognitive transparency and participation initiative.

[0089] While users are watching ads, the system collects real-time ad behavior data (such as clicks, duration of viewing, completion, or skipping) and incorporates this behavioral feedback into user behavior profiling and ad preference analysis modules. If the system detects a low tolerance for ads (e.g., frequent skipping), it automatically reduces the frequency of subsequent ad pushes or blocks corresponding ad tasks, achieving dynamic adaptation of delivery strategies and user experience protection.

[0090] After completing an advertising task, the system will issue corresponding gold coin rewards based on the task settings, mark the source as "advertising incentives," and record it separately in the account. Gold coin rewards can be used in conjunction with other incentive channels such as badges, re-signing, withdrawals, and redemptions to form a complete behavioral monetization closed loop.

[0091] The advertising incentive mechanism of the present invention has the following core features: non-mandatory triggering, transparent and visible rewards, users can close it on their own, advertising behavior affects subsequent strategy push, supports personalized behavior targeted recommendations, and rewards can be traced and managed in a hierarchical manner. It effectively improves the platform's commercialization capabilities while ensuring user experience and behavioral compliance, and is one of the important commercial scalability modules of the present invention.

[0092] The above scheme is only a description of the preferred implementation method of this application, but the scope of protection of this application is not limited to this. Any person familiar with the technology can easily implement it within the scope of the description of this application without changing the changes or replacements of the basic principles involved in the claims, which should be covered by the scope of protection of this application, that is, the scope of protection of this application should be based on the scope of protection of the claims.

Claims

1. A system for energy-saving intelligent drinking water behavior management and incentive device, characterized in that: The system includes: an intelligent behavior recognition module, a dynamic energy-saving control module, a behavior prediction and intelligent reminder module, a personalized incentive recommendation module, a gold coin and business conversion module, and a client and network configuration module; The intelligent behavior recognition module is used to accurately identify the user's drinking behavior; The dynamic energy-saving control module implements refined power consumption control based on the task priority scheduling mechanism and the predictive wake-up strategy; The behavior prediction and intelligent reminder module is used to predict the possible time window for the next drinking behavior; The personalized incentive recommendation module analyzes and stratifies user status according to the user behavior scoring model and dynamically recommends incentive statement content that matches the user; The gold coin and business conversion module collects user drinking behaviors and scores tasks in real time, building an incentive distribution model based on a rule engine to improve user engagement and platform business conversion efficiency. The client and network configuration module is used to improve the system intelligence, maintainability and user interaction experience, and realizes the data linkage, function configuration and user behavior feedback closed loop between the device end and the client.

2. A system for energy-saving intelligent drinking water behavior management and incentive device according to claim 1, characterized in that: The intelligent behavior recognition module includes: real-time data collection based on high-precision weight sensors and analog-to-digital converters, combined with a multi-dimensional behavior judgment algorithm to accurately identify users' drinking behavior, support self-learning strategies, and adapt and optimize users' different drinking habits. Among them, the multi-dimensional behavior judgment algorithm includes: using a sliding time window mechanism to achieve real-time judgment. Within each window period, the system continuously collects weight sensor data and constructs a behavior sequence.

3. The system for energy-saving intelligent drinking water behavior management and incentive device according to claim 1 is characterized in that: The dynamic energy-saving control module includes: setting different levels of priority according to current and estimated tasks. The device is in deep sleep state by default and is only awakened once in ultra-low power consumption mode within a set detection cycle. After waking up, the system determines whether there is a high-priority task. If there is no task, it quickly returns to sleep. If the task judgment is successful, it enters light sleep / running state and loads the required modules to prepare for execution.

4. The system for energy-saving intelligent drinking water behavior management and incentive device according to claim 1, characterized in that: The behavior prediction and intelligent reminder module includes: building a drinking rhythm model based on the user's historical drinking behavior data, using a sliding time window to extract high-frequency drinking time periods in the past N days to form an individualized drinking rhythm vector, and predicting the possible time window for the next drinking behavior by calculating the time deviation between the current time period and the rhythm vector. When the deviation reaches a set threshold, the system triggers the reminder logic. To avoid interference from unexpected time periods, the system supports users to set the start and end time periods of daily reminders. All reminder logic is only effective within this time window. Beyond this time period, even if the deviation logic meets the conditions, the system will not trigger any reminders.

5. The system for energy-saving intelligent drinking water behavior management and incentive device according to claim 1 is characterized in that: The client and network configuration module include: Multiple WiFi configurations and automatic switching mechanism: Supports multiple WiFi access points preset and save. The system automatically reconnects to the optimal network when the signal is interrupted or the environment changes. The first connection supports turning on AP mode through the device, and the client can scan the code or bind the network with one click. Network interruption data caching and breakpoint resumption: When the device is offline, user behavior data will be cached in the local storage module and compressed and uploaded when the network is restored, reducing communication energy consumption and ensuring data integrity. Client core functional modules: Real-time viewing: Users can view drinking behavior statistics, consecutive days of meeting the standard, and gold coin balance in the app; Custom settings: support setting daily reminder time, whether to enable holiday reminders, and motivational statement style preferences; Incentivized interaction: Users view personalized incentives, like / ignore statements, and influence subsequent recommendation models; Remote OTA upgrade: The system supports issuing firmware update instructions through the client, and the device automatically connects to the server to complete the differential upgrade; User feedback mechanism: built-in problem reporting / abnormal feedback channels for continuous optimization of cloud strategies and client behaviors; Policy issuance and remote linkage mechanism: The system supports sending relevant content from the cloud based on user behavior tags and behavior frequency to achieve system-level adaptive adjustment.

6. A method for realizing an energy-saving intelligent drinking water behavior management and incentive device system, characterized in that: The method comprises: Step 1: After powering on, read the local WiFi configuration. If the device is connected to the network successfully, it will enter the deep-sleep state by default. If it fails, it will automatically enter the AP mode and wait for the user to scan the code to connect to the network. Step 2: Wake up the device, collect data every second through the weight sensor, and use the sliding time window algorithm to determine whether it is a valid drinking behavior; Step 3: If it is identified as drinking water behavior, the system generates a record and executes the upload logic; Step 4: Construct an individual drinking rhythm vector based on the user's 7-day drinking record and identify high-frequency periods; Step 5: Build a user behavior scoring model, generate a behavior vector based on the scoring results, and assign it to a strategy pool. Then, based on factors such as time period, holidays, and task status, push corresponding incentive statements. Step 6: If the customer completes the specified behavior, he or she will also receive gold coins and badges as rewards.

7. A method for realizing an energy-saving intelligent drinking water behavior management and incentive device system according to claim 6, characterized in that: In step 2, using the sliding time window algorithm to determine whether the drinking behavior is effective includes: ① Within each 60-second sliding time window, the system continuously collects water cup weight data and compares the current weight with the weight of the last drinking record; ② Check whether there are n consecutive times (n ≥ 3) of weight decrease or increase, with each change exceeding 10 grams (ΔW ≥ 10g); ③ Check whether the deviation of these weight changes is within 10 grams (ΔW≤10g); ④ If the above conditions are met, the system will mark the drinking behavior as valid and generate a drinking record; ⑤ If the weight variation deviation of three consecutive samples does not exceed 10 grams after the record is generated, the system stops further testing; ⑥ If the device does not detect any effective drinking behavior or operation within 10 seconds, the device will enter sleep mode to save power.

8. A method for realizing an energy-saving intelligent drinking water behavior management and incentive device system according to claim 6, characterized in that: The step 4 includes: the cloud service generates an individualized drinking rhythm model and reminder time list based on the user's historical drinking behavior, and sends the reminder strategy to the device side when the user uploads the data for the first time. The device stores the reminder schedule locally and carries the local reminder version number for comparison by the cloud side each time it uploads. If the version is inconsistent, the system will automatically update the reminder schedule to ensure that the reminder logic is synchronized in real time. Both the device side and the App client have synchronization capabilities: after the user synchronizes the drinking record through the App, if it is detected that the interval between the next reminder point is greater than the threshold, it will automatically update to the next reminder time point. In the local reminder triggering process, the device wakes up the main control chip by the RTC timer, executes the LED reminder task, and then returns to the sleep state; the App side reminder process is triggered by the cloud task timer scheduling, and reminds the user to drink water through push.

9. A method for realizing an energy-saving intelligent drinking water behavior management and incentive device system according to claim 6, characterized in that: The dimensions of the user behavior scoring model constructed in step 5 include: reminder response time, consecutive check-in days, target-reaching frequency, active cycle and completion rate. The motivational language styles are divided into: morning greetings, emotional motivation, achievement feedback, soft encouragement, and holiday greetings. Users set display preferences and frequency. The client supports liking / ignoring motivational language. The system records and provides feedback to optimize subsequent recommendation strategies, forming an intelligent optimization closed loop of behavior → motivation → feedback → re-recommendation.

Images