Intelligent lock energy consumption optimization control method and intelligent lock

Abstract

The application is suitable for the technical field of intelligent locks, and particularly relates to an intelligent lock energy consumption optimization control method and an intelligent lock. The method comprises the following steps: obtaining historical behavior information of a user using the intelligent lock; obtaining first behavior information by predicting the historical behavior information; in a first time period, when it is determined that the user approaches the intelligent lock, performing a first operation according to a first unlocking mode; in the first time period, when it is determined that the user approaches the intelligent lock, performing a second operation and receiving a third operation on the intelligent lock after the second operation; and starting a corresponding identification module based on the third operation. Therefore, the intelligent lock energy consumption optimization control method can solve the problem of high energy consumption caused by the lag of the control strategy of the identification module of the intelligent lock.

Description

TECHNICAL FIELD

[0001] The application belongs to the technical field of intelligent locks, and particularly relates to an intelligent lock energy consumption optimization control method and an intelligent lock. BACKGROUND

[0002] An intelligent lock is an intelligent device that replaces a traditional mechanical lock to realize door lock switch control based on electronic technology, the Internet of Things and biometric identification technology.

[0003] The core function of the intelligent lock depends on an identification module for identifying fingerprints, passwords, NFC and the like of a user, and in the prior art, the identification module is always on or is polled at a high frequency, which causes standby power consumption to surge. In addition, the identification modules work in parallel, and when a certain identification module is triggered, other identification modules remain in a standby state, which causes invalid power consumption. Therefore, the existing intelligent lock has the problem of high energy consumption caused by a lag in the control strategy for the identification module. SUMMARY

[0004] The embodiments of the application provide an intelligent lock energy consumption optimization control method and an intelligent lock, which can solve the problem of high energy consumption caused by a lag in the control strategy for the identification module of the intelligent lock.

[0005] In a first aspect, the embodiments of the application provide an intelligent lock energy consumption optimization control method, which comprises the following steps.

[0006] Real-time acquisition of historical behavior information of a user using the intelligent lock;

[0007] Prediction of first behavior information according to the historical behavior information; wherein the first behavior information comprises a first time period of next unlocking of the user and a corresponding first unlocking mode;

[0008] In the first time period, when it is determined that the user approaches the intelligent lock, a first operation is performed according to the first unlocking mode; wherein the first operation refers to starting a sensor module and a corresponding identification module of the intelligent lock;

[0009] When it is determined that the user approaches the intelligent lock according to real-time position information, a second operation is performed and a third operation of the intelligent lock after the second operation is received; wherein the second operation refers to starting only the sensor module, and the third operation refers to an input signal of the user.

[0010] The corresponding identification module is started based on the third operation.

[0011] The technical solution described above in the embodiments of the application has at least the following technical effects.

[0012] The intelligent lock energy consumption optimization control method provided in the embodiments of the present application can obtain historical behavior information of a user using the intelligent lock in real time, predict first behavior information according to the historical behavior information, perform a first operation according to a first unlocking mode when it is determined that the user is close to the intelligent lock in a first time period, perform a second operation and receive a third operation on the intelligent lock after the second operation when it is determined that the user is close to the intelligent lock in a time period other than the first time period, and start a corresponding identification module based on the third operation. Therefore, the intelligent lock energy consumption optimization control method provided in the embodiments of the present application combines historical behavior information and real-time location information to form a dynamic strategy of high-precision response in a prediction period and low-power consumption defense in a non-prediction period, which is beneficial to reducing standby energy consumption of the intelligent lock.

[0013] In a possible implementation manner of the first aspect, the method further includes:

[0014] adjusting a working parameter of the intelligent lock according to the first behavior information, wherein the working parameter includes a working cycle of an identification module of the intelligent lock and a sampling frequency of a sensor module.

[0015] In a possible implementation manner of the first aspect, the historical behavior information includes a plurality of pieces of unlocking information, and the unlocking information includes a time and an unlocking mode of each unlocking of the user; and the predicting the first behavior information according to the historical behavior information includes:

[0016] determining, according to the historical behavior information, a correlation feature and a user feature of each piece of unlocking information, wherein the correlation feature includes a time feature and a corresponding behavior feature of the user using the intelligent lock, the behavior feature is used to reflect the unlocking mode of the user, and the user feature is used to reflect a behavior mode of the user;

[0017] predicting the first behavior information based on the correlation feature and the user feature.

[0018] In a possible implementation manner of the first aspect, the determining, according to the historical behavior information, the correlation feature and the user feature of each piece of unlocking information includes:

[0019] determining, based on each piece of unlocking information in the historical behavior information, a short-term feature of each piece of unlocking information, wherein the short-term feature is used to reflect a local correlation of each piece of unlocking information;

[0020] determining a long-term feature based on the short-term feature of each piece of unlocking information, wherein the long-term feature is used to reflect a periodicity of each piece of unlocking information;

[0021] gate-fusing each short-term feature and each long-term feature to obtain the correlation feature of the intelligent lock.

[0022] In a possible implementation manner of the first aspect, the determining the short-term feature of each piece of unlocking information based on the historical behavior information comprises:

[0023] The first correlation between the pieces of unlocking information is calculated to obtain a feature matrix corresponding to each piece of unlocking information, wherein the first correlation is used to reflect the correlation between the time interval of the pieces of unlocking information and the corresponding unlocking manner in the second time period;

[0024] The short-term feature of each piece of unlocking information is obtained according to the feature matrix.

[0025] In a possible implementation manner of the first aspect, the determining the long-term feature based on the short-term feature of each piece of unlocking information comprises:

[0026] The mutual information value between the pieces of unlocking information is calculated based on the short-term features to obtain a feature vector corresponding to each piece of unlocking information;

[0027] The feature vectors are enhanced according to the similarity between the feature vectors to obtain the long-term feature of each piece of unlocking information.

[0028] In a possible implementation manner of the first aspect, the determining the association feature and the user feature of each piece of unlocking information according to the historical behavior information further comprises:

[0029] A directed graph is constructed based on the time sequence of the pieces of unlocking information in the historical behavior information;

[0030] The attention weight between the nodes with a directed edge in the directed graph is calculated, and the attention feature of each piece of unlocking information is determined according to the attention weight;

[0031] The attention features are aggregated to obtain the user feature.

[0032] In a possible implementation manner of the first aspect, the calculating the attention weight between the nodes with a directed edge in the directed graph and determining the attention feature of each piece of unlocking information according to the attention weight comprises:

[0033] Each node in the directed graph node is initialized based on the pieces of unlocking information to obtain an initial feature of each node;

[0034] The attention weight between the nodes with a directed edge in the directed graph is calculated, and the initial feature of the node is updated according to the attention weight to obtain the attention feature of each piece of unlocking information.

[0035] In a possible implementation manner of the first aspect, the predicting the first behavior information based on the association feature and the user feature comprises:

[0036] According to the association feature, the user feature, and the latest unlocking information, probability distribution information is predicted, wherein the probability distribution information comprises a probability distribution of a time period and an unlocking manner in which the user unlocks next time;

[0037] The first behavior information is obtained according to the probability distribution information.

[0038] In a second aspect, an energy consumption optimization control device of an intelligent lock is provided, comprising:

[0039] An acquisition module is configured to acquire historical behavior information of the user using the intelligent lock in real time;

[0040] A prediction module is configured to predict first behavior information according to the historical behavior information, wherein the first behavior information comprises a first time period and a corresponding first unlocking manner in which the user unlocks next time;

[0041] A first operation module is configured to, in the first time period, perform a first operation according to the first unlocking manner when it is determined that the user approaches the intelligent lock, wherein the first operation refers to starting a sensor module and a corresponding identification module of the intelligent lock.

[0042] A receiving module is configured to, in a time period other than the first time period, perform a second operation when it is determined that the user approaches the intelligent lock, and receive a third operation of the intelligent lock performed by the user after the second operation, wherein the second operation refers to starting only the sensor module, and the third operation refers to an input signal of the user.

[0043] A starting module is configured to start the corresponding identification module based on the third operation.

[0044] In a third aspect, an intelligent lock is provided, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, and the processor implements the method of any one of the first aspect when executing the computer program.

[0045] In a fourth aspect, a computer readable storage medium is provided, and the computer readable storage medium stores a computer program, and the computer program is executable on a processor to implement the method of any one of the first aspect.

[0046] In a fifth aspect, a computer program product is provided, and when the computer program product is executed on an intelligent lock, the intelligent lock performs the method of any one of the first aspect.

[0047] It can be understood that the beneficial effects of the above-mentioned second aspect to the fifth aspect can be referred to the related description in the above-mentioned first aspect, which will not be repeated here. BRIEF DESCRIPTION OF DRAWINGS

[0048] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings needed to be used in the embodiments or prior art description will be briefly introduced as follows. Obviously, the drawings in the following description only some embodiments of the present application, and for those skilled in the art, other drawings can also be obtained without creative labor on the basis of these drawings.

[0049] Figure 1 is a flowchart of the intelligent lock energy consumption optimization control method provided by an embodiment of the present application;

[0050] Figure 2 is a flowchart of the implementation of step S200 in the intelligent lock energy consumption optimization control method provided by an embodiment of the present application;

[0051] Figure 3 is a flowchart of the implementation of steps S210, S211, S212, S215 and S220 in the intelligent lock energy consumption optimization control method provided by an embodiment of the present application;

[0052] Figure 4 is a structural schematic diagram of the intelligent lock energy consumption optimization control device provided by an embodiment of the present application;

[0053] Figure 5 is a structural schematic diagram of the intelligent lock provided by an embodiment of the present application. DETAILED DESCRIPTION

[0054] In the following description, specific details such as specific system structures, techniques, etc. are presented in order to thoroughly understand the embodiments of the present application. However, it should be clear to those skilled in the art that the present application can also be implemented in other embodiments without these specific details. In other cases, detailed descriptions of well-known systems, devices, circuits and methods are omitted to avoid unnecessary details that hinder the description of the present application.

[0055] It should be understood that when used in the specification and the appended claims of the present application, the term "comprising" indicates the presence of the described features, integers, steps, operations, elements and / or components, but does not exclude the presence or addition of one or more other features, integers, steps, operations, elements, components and / or sets thereof.

[0056] It should also be understood that the term "and / or" used in the specification and the appended claims of the present application means any combination of one or more of the associated listed items and all possible combinations, and includes these combinations.

[0057] As used in the description and the appended claims of the application, the term "if' can be interpreted as meaning "when" or "upon" or "in response to a determination" or "in response to a detection" depending on the context. Similarly, the phrase "if determined" or "if detected [the described condition or event]" can be interpreted to mean "upon a determination" or "in response to a determination" or "upon a detection [of the described condition or event]" or "in response to a detection [of the described condition or event]" depending on the context.

[0058] In addition, in the description of the application and the appended claims, the terms "first", "second", "third", etc. are only used to distinguish the description and cannot be understood as indicating or implying relative importance.

[0059] In the description of the application, the reference "one embodiment" or "some embodiments" and the like means that the specific features, structures or characteristics described in connection with the embodiment are included in one or more embodiments of the application. Therefore, the statements "in one embodiment", "in some embodiments", "in other some embodiments", "in further some embodiments" and the like appearing in different places in the specification are not necessarily all referring to the same embodiment, but mean "one or more but not all embodiments", unless otherwise specifically emphasized. The terms "include", "contain", "have" and their variants mean "include but not limited to", unless otherwise specifically emphasized.

[0060] In the related art, the existing technology, the identification module is always on or high frequency polling, which causes the standby power consumption to surge. And the parallel work between each identification module, when a certain identification module is triggered, other identification modules still remain in standby state, which causes invalid energy consumption. Therefore, the existing smart lock has the problem of high energy consumption caused by the lag of the control strategy of the identification module.

[0061] To solve the above problems, the embodiments of the application provide a smart lock energy consumption optimization control method and a smart lock. In the method, historical behavior information of a user using the smart lock is obtained; first behavior information is predicted according to the historical behavior information; in a first time period, when it is determined that the user is close to the smart lock, a first operation is performed according to a first unlocking mode; in a time period other than the first time period, when it is determined that the user is close to the smart lock, a second operation is performed and a third operation on the smart lock after the second operation is received; and a corresponding identification module is started based on the third operation. Therefore, the smart lock energy consumption optimization control method provided by the embodiments of the application combines historical behavior information and real-time location information to form a dynamic strategy of high-precision response in a prediction period and low-power defense in a non-prediction period, which is conducive to reducing the standby energy consumption of the smart lock.

[0062] The intelligent lock energy consumption optimization control method provided by the embodiments of the present application can be applied to the intelligent lock, at this time the intelligent lock is the execution subject of the intelligent lock energy consumption optimization control method provided by the embodiments of the present application, and the embodiments of the present application do not limit the specific type of the intelligent lock.

[0063] For example, the intelligent lock can include a plurality of sensor modules and a plurality of identification modules, the sensor modules are arranged on the outer panel of the intelligent lock, the identification modules are built-in in the inner panel of the intelligent lock, and the sensor modules and the identification modules are in one-to-one correspondence and are communicatively connected. Among them, the sensor module is used to receive the input signal of the user, which can be a semiconductor capacitive sensor, a mechanical key matrix, a capacitive touch screen or a face recognition camera, but is not limited thereto; the identification module is used to analyze the input signal of the user to determine whether to unlock, and the identification module can be a capacitive fingerprint identification module (including a fingerprint algorithm chip (such as ARM Cortex-M0)), a password identification module (including a key scanning chip (such as TM1638)), a touch password / figure identification module (including a touch controller (such as STM32F407)) or a face recognition module (including a 3D structured light chip), but is not limited thereto.

[0064] The intelligent lock can also include a communication module for supporting remote control of a mobile phone APP and the like, temporary password sharing, and unlocking synchronization recording, wherein the communication module can include a Bluetooth 5.0 chip (such as Nordic nRF52840) and a WiFi module and the like; a lock body driving module for receiving the control instruction sent by the main control board to drive the lock tongue to retract and extend to realize unlocking / locking, wherein the lock body driving module can include a DC motor, a speed reduction gear set and a lock tongue control board and the like; a security alarm module for detecting illegal prying or continuous password errors, wherein the security alarm module can include an accelerometer, a trial-and-error locking chip and a buzzer and the like; an environmental sensor, a power supply and a main control board (MCU) responsible for coordinating the work of each module and processing data and the like, but is not limited thereto.

[0065] In order to better understand the intelligent lock energy consumption optimization control method provided by the embodiments of the present application, the specific implementation process of the intelligent lock energy consumption optimization control method provided by the embodiments of the present application will be exemplarily introduced below.

[0066] Figure 1 The schematic flowchart of the intelligent lock energy consumption optimization control method provided by the embodiments of the present application is shown, and the intelligent lock energy consumption optimization control method includes:

[0067] S100, real-time acquisition of historical behavior information of the user using the intelligent lock.

[0068] It can be understood that the real-time acquisition of the historical behavior information of the user using the intelligent lock means that the sensor module of the intelligent lock is used to perceive the user operation in real time, and the time and the unlocking mode of the unlocking are recorded synchronously when the user unlocks successfully.

[0069] Exemplarily, the main control board of the smart lock is in communication connection with the sensor module and the identification module, the main control board can integrate a real-time clock (RTC) chip (such as DS3231) to record the time stamp of each unlocking, the unlocking mode (fingerprint / password / NFC / Bluetooth), upload the data to the cloud database or local storage through the communication module of the smart lock, and format the data into a structured log (such as JSON) to obtain historical behavior information. And the data can be aggregated according to the user ID in the historical behavior information.

[0070] S200, the first behavior information is predicted according to the historical behavior information. Wherein, the first behavior information includes a first time period of next unlocking of the user and a corresponding first unlocking mode.

[0071] Exemplarily, the time characteristics (such as the time period in which hour, whether it belongs to weekdays / weekends, etc.) and the unlocking mode characteristics (such as the frequency of various unlocking modes in a period of time) can be extracted according to the historical behavior information, and a time sequence model (such as LSTM) is trained according to the historical behavior information: input the time characteristics of unlocking in a period of time (such as the time period sequence of unlocking of each user in 30 days), output the time period of next unlocking (regression task); a classification model (such as random forest) is trained according to the historical behavior information: input the characteristics of the last N times of unlocking (such as the proportion of various unlocking modes of each user in 30 days), output: the category of next unlocking mode (classification task) (wherein, the time period prediction uses MSE (mean square error), and the unlocking mode prediction uses cross-entropy loss function), the time characteristics and the unlocking mode characteristics of the last N times of unlocking information in the historical behavior information are respectively input into the time sequence model (such as LSTM) and the classification model (such as random forest) to predict the first time period and the first unlocking mode of next unlocking of each user.

[0072] S300, in the first time period, when it is determined that the user approaches the smart lock, a first operation is performed according to the first unlocking mode. Wherein, the first operation refers to starting the sensor module and the corresponding identification module of the smart lock.

[0073] Exemplarily, the smart lock can be communicatively connected with an electronic device carried by a user, and the smart lock includes an environment sensor (such as an infrared sensor, a Doppler microwave sensor, a Bluetooth sensor, etc.). A position signal of the user from the electronic device is received (which can be calculated according to a low-power Bluetooth (BLE) connection established between the smart lock and the electronic device carried by the user through a received signal strength indication (RSSI) and a time difference of arrival (TDOA) of a Bluetooth signal), and a user ID can be determined according to the electronic device carried by the user, wherein the electronic device can be a mobile phone, an electronic watch, a tablet computer, etc.; or a position signal of the user can be determined according to the environment sensor (such as detecting the duration of an infrared (wavelength 9-10 μm) signal radiated by a human body, detecting the speed and direction of a moving object through Doppler shift), and whether the user is close to the smart lock is determined according to the position signal. For example, it is judged that the distance between the user and the smart lock is ≤2 meters through a Bluetooth RSSI value or UWB ranging, and the current time ∈ 08:10-08:20, and the semiconductor capacitive sensor and the fingerprint recognition module are started according to the first unlocking mode in the first behavior information.

[0074] S400, in the case where it is determined that the user is close to the smart lock, a second operation is performed in the first time period, and a third operation of the smart lock after the second operation is performed by the user is received. The second operation refers to starting only the sensor module, and the third operation refers to an input signal of the user.

[0075] Exemplarily, when it is not in the first time period, only the environment sensor in the first device is activated to monitor the proximity event, all sensor modules are started after it is determined that the user is close, and the recognition module is manually triggered by the user, if there is no operation within 30 seconds, the sleep state is entered.

[0076] S500, the corresponding recognition module is started based on the third operation.

[0077] Exemplarily, the unlocking mode of the user can be determined according to the third operation received by the sensor module, and the corresponding recognition module is started.

[0078] In the prior art, the control method of the smart lock usually adopts a fixed wake-up interval or a sensor trigger, all sensors (such as a fingerprint sensor, an infrared sensor, a password keyboard) are in a low-power monitoring state for a long time, and all recognition modules (such as a fingerprint recognition chip, a password decoder, an NFC reader) are started simultaneously as soon as a user operation (such as touch, key pressing) is detected. At this time, the user may only use one unlocking mode (such as fingerprint), but all sensors and recognition modules are awakened, resulting in a large amount of invalid power consumption. And it cannot be dynamically adjusted according to user behavior, resulting in unnecessary power consumption. At present, the energy consumption optimization of the smart lock usually considers the overall use of the smart lock by the user, and does not consider the individual differences of different user behaviors (such as the homecoming time of different users, the commonly used unlocking mode).

[0079] To this end, each sensor module of the present application is connected with a unique identification module through a dedicated interface (such as UART, SPI) to form a "one-to-one" channel, each sensor module can be independently switched between low-power mode and high-response mode, reducing the number of invalid wake-up of the sensor module and the identification module of the smart lock, and reducing the power consumption. In the historical behavior information, the data is aggregated according to the user ID, and the detailed behavior data of each user using the smart lock is comprehensively collected, which can more accurately understand the user behavior pattern. According to the historical behavior information, the first time period and the corresponding first unlocking mode of the next unlocking of the user are predicted, which can dynamically determine the possible unlocking time and mode of the user according to the historical behavior of the user, instead of using a fixed wake-up interval, thereby avoiding unnecessary wake-up and reducing power consumption. When obtaining the historical behavior information, the data is aggregated according to the user ID, so that the behavior data of each user is relatively independent and complete. In the subsequent processing and analysis process, individual differences between different users are considered, such as the home-coming time and the commonly used unlocking mode of different users, so that the prediction model can better adapt to the behavior characteristics of different users, improve the generalization ability of the prediction model, and more accurately predict the unlocking behavior of different users.

[0080] In a possible implementation, referring to Figure 2 , the method further includes:

[0081] S600, adjusting the working parameters of the smart lock according to the first behavior information. The working parameters include the working period of the identification module of the smart lock and the sampling frequency of the sensor module.

[0082] Exemplarily, the working parameters of the identification module and the sensor module can be dynamically adjusted based on the predicted first behavior information (first time period, first unlocking mode), and the working parameters of the environmental sensor in the first device can be adjusted, for example, in the predicted first time period: the environmental sensor and the sensor module corresponding to the first unlocking mode maintain a high sampling rate, and the identification module (such as a face recognition camera) corresponding to the first unlocking mode only wakes up at a high frequency in the predicted first time period; in the non-first time period: the environmental sensor and each sensor module maintain a low sampling rate to save power (such as reducing the infrared sensor from 10Hz to 5Hz), and the identification period of each identification module is extended to reduce invalid wake-up.

[0083] Through the above step S600, in the predicted first time period, the system assumes that the user will arrive soon, therefore, the sampling rate is increased and the identification period is shortened, the "high-response mode" is entered in advance, so that the user is quickly responded when approaching. The high-frequency sampling of the sensor can detect the approaching behavior earlier. The high-frequency wake-up of the identification module can shorten the total time from approaching to unlocking. In the non-predicted time period, the system defaults to enter the "low-power mode", and the identification period is extended and the sampling rate is reduced to reduce invalid calculation.

[0084] In a possible implementation, referring to Figure 2 , the historical behavior information includes a plurality of pieces of unlocking information, and each piece of unlocking information includes a time and a manner of unlocking by the user each time. S200, first behavior information is predicted according to the historical behavior information, including:

[0085] S210, the associated features and the user features of each piece of unlocking information are determined according to the historical behavior information. The associated features include time features and corresponding behavior features of the user using the smart lock, and the behavior features are used to reflect the unlocking manner of the user, and the user features are used to reflect the behavior mode of the user.

[0086] It can be understood that the historical behavior information includes the time and the unlocking manner of a plurality of users using the smart lock.

[0087] Exemplarily, the time stamp can be decomposed into hours, days of the week, and whether it is a weekday to obtain the time features according to the historical behavior information, and the unlocking manner is one-hot encoded to obtain the associated features. The historical behavior mode of a plurality of users is obtained according to the historical behavior information to obtain the user features, for example, the unlocking manner frequency: P(fingerprint) = 0.7, P(password) = 0.3, and the unlocking time period: peak period = 7:00-9:00 (accounting for 60%).

[0088] S220, the first behavior information is predicted based on the associated features and the user features.

[0089] Exemplarily, a prediction model can be trained according to the associated features of the data in each period (such as 30 days) and the user features of different users in the historical behavior information, for example, a time series classification model (such as LSTM) or a gradient boosting tree (such as XGBoost) (the label (Label) is the time period (such as 08:00-09:00) and the manner (such as fingerprint) of the next unlocking, and a cross-entropy loss function is used to optimize a multi-classification task: time period classification, unlocking manner classification), the associated features of the most recent period (such as 30 days) and the user features of different users in the historical behavior information are input, the prediction model outputs a probability distribution, and the highest probability is taken as a prediction result to obtain the first behavior information.

[0090] Through the above steps S210 to S220, the user behavior has periodicity (such as high-frequency use of fingerprint unlocking in the morning of weekdays), and by decomposing the time stamp into hours, days of the week and the like, the model can learn the pattern of "fingerprint is more likely to be used at 8:00 on Monday". The behavior features (such as unlocking mode) directly reflect single behavior, and need to be one-hot encoded to avoid numerical bias (such as password = 2 should not be considered as twice of fingerprint = 1). The user features (such as frequency of commonly used unlocking mode) reflect long-term behavior patterns, and are used to correct the prediction (such as when the user uses fingerprint 90% of the time, the model should be more inclined to predict fingerprint). Combining time, behavior and user features into a vector enables the model to capture the complex relationship of "particular time + particular user → particular unlocking mode". Combining static historical data with dynamic sensor data (such as Bluetooth RSSI) avoids prediction rigidity.

[0091] Optionally, referring to Figure 3 , S210, determining the associated features and user features of each unlocking information according to historical behavior information, including:

[0092] S211, determining the short-term features of each unlocking information based on each unlocking information in the historical behavior information. The short-term features are used to reflect the local correlation of each unlocking information.

[0093] Exemplarily, for the unlocking time sequence in the historical behavior information, a sliding window (such as the last 5 unlocking records) can be set, and the local behavior pattern, i.e. the short-term features, include: behavior continuity: the number of consecutive same unlocking modes (such as 3 consecutive fingerprints) is counted; time interval pattern: the mean and variance of the time interval between adjacent unlockings are calculated (such as μ = 36900 s, σ = 500 s); local entropy: the randomness of the unlocking mode is measured (such as 80% of the fingerprints and 20% of the passwords, and a low entropy value indicates strong regularity).

[0094] S212, determining the long-term features based on the short-term features of each unlocking information. The long-term features are used to reflect the periodicity of each unlocking information.

[0095] Exemplarily, Fourier transform or seasonal decomposition (STL) can be used to extract the long-term features. For example, FFT is performed on the unlocking time (hours) of User_A in the past 30 days, and it is found that the significant periods are 24 hours (daily period) and 168 hours (weekly period), and the long-term feature extraction includes: peak period frequency: the unlocking number ratio of each day at a fixed time period (such as 7:00-9:00) is counted; period intensity: the autocorrelation coefficient is calculated (such as the autocorrelation value of the 24-hour period is 0.85); unlocking mode periodicity: the proportion of fingerprints in the morning from Monday to Friday (such as 90% of the fingerprints on weekdays and 50% of the fingerprints on weekends).

[0096] S213, gate fusion of each short-term feature and each long-term feature to obtain the associated features of the smart lock.

[0097] Exemplarily, a gated recurrent unit (GRU) or attention mechanism can be used to dynamically adjust the weights of short-term and long-term features, and the weighted fusion obtains the associated features of the smart lock. If the current time is in the user historical peak period (e.g., 8:00), the weight of the long-term feature is increased; if an abnormal time interval is detected (e.g., an interval of 20 minutes instead of the usual 9 hours), the weight of the short-term feature is increased.

[0098] Through the above steps S211 to S213, the core of the short-term feature is to capture the local dynamics of user behavior, and the behavior pattern of the last few unlocking is analyzed through the sliding window. By quantifying the short-term feature, the model can distinguish between "daily habits" and "temporary behavior", for example, when the user suddenly changes the unlocking method, the short-term feature (such as high interval variance) will trigger abnormal detection. The long-term feature aims to reveal the deep periodicity of user behavior, and the stability of the long-term feature makes it a "anchor point" for prediction, especially when the data is sparse (such as a new user), the periodicity of the group can be relied on to initialize the prediction. When an abnormal time interval is detected (e.g., the user usually unlocks with an interval of 12 hours, but the current time interval from the last unlocking is only 1 hour), the weight of the short-term feature will be increased to quickly respond to the behavior change. The long-term feature dominates the scene: in the user historical peak period (e.g., Monday 8:00), the long-term feature will dominate the prediction to avoid short-term noise interference. The attention score of the Transformer can be further introduced to dynamically adjust the weight according to the similarity between the unlocking time and the historical pattern (e.g., the matching degree between the current time and the historical Monday afternoon is 0.85, then the weight of the corresponding long-term feature is increased). Through gated fusion, the model can not only utilize the historical periodicity, but also flexibly adapt to real-time behavior changes, for example, when the user temporarily adjusts the work and rest, the weight of the long-term feature is gradually reduced to avoid prediction rigidity.

[0099] Exemplarily, please refer to Figure 3 , S211, determining the short-term feature of each unlocking information based on the historical behavior information, including:

[0100] S2111, calculating the first correlation between each unlocking information to obtain a feature matrix corresponding to each unlocking information. The first correlation is used to reflect the correlation between the time interval and the corresponding unlocking method of each unlocking information in the second time period.

[0101] Exemplarily, the first correlation is a joint correlation between the time interval and the unlocking manner between each pair of unlocking information in a second time period (such as the last 1 hour or the last 5 unlocking operations). For each unlocking information, the other unlocking information (such as the last 5 unlocking operations) in the second time period is traced back to form a subsequence, and the joint correlation between each pair of unlocking information in the subsequence is calculated to obtain the time interval similarity and the unlocking manner similarity. For each unlocking information, the feature matrix is an N*N symmetric matrix, which represents the joint correlation of all unlocking pairs in the window.

[0102] S2112, obtaining the short-term feature of each unlocking information according to each feature matrix.

[0103] Exemplarily, for the feature matrix of each unlocking information, a statistical quantity (such as the trace, average value, proportion of elements exceeding a preset threshold, etc. of the feature matrix) can be extracted as the short-term feature.

[0104] Through the above steps S2111 to S2112, the importance of time and manner can be balanced by weighted combination. If the user is used to using fingerprints at fixed intervals, the time interval similarity will dominate; if the user frequently switches the manner, the unlocking manner similarity weight will be improved. By aggregating the statistical quantity of the feature matrix, the short-term feature can compress the high-dimensional correlation information into a low-dimensional vector, while retaining the key patterns (such as regularity, mutation point) of local correlation, providing a basis for subsequent long-term feature fusion or anomaly detection.

[0105] Exemplarily, please refer to Figure 3 S212, determining the long-term feature based on the short-term feature of each unlocking information, including:

[0106] S2121, calculating the mutual information value between each unlocking information based on each short-term feature to obtain the feature vector corresponding to each unlocking information.

[0107] Exemplarily, the short-term feature can be discretized by binning and converted into discrete symbols. For each pair of unlocking information, the frequency of the joint distribution of the short-term features is counted, and the mutual information value is calculated according to the marginal probability and joint probability of each pair of unlocking information to obtain the feature vector of the mutual information value corresponding to each unlocking information.

[0108] S2122, performing feature enhancement on each feature vector according to the similarity between the feature vectors to obtain the long-term feature of each unlocking information.

[0109] Exemplarily, the cosine similarity or Euclidean distance can be used to calculate the similarity between the feature vectors. The unlocking information is regarded as a graph node, and the similarity is regarded as an edge weight. The Top-K similar edges (such as K=3) are retained, the feature vectors of the neighbor nodes of the unlocking information are aggregated according to the similarity or the dynamic weight is allocated using an attention mechanism, and the aggregated vector is spliced with the short-term feature to obtain the long-term feature.

[0110] By the above steps S2121-S2122, the unlocking information is regarded as a graph node, and the similarity is regarded as an edge weight, allowing information to flow from similar nodes. Splicing the original short-term features and the enhanced features retains the original information while introducing global context.

[0111] Optionally, referring to Figure 3 , S210, determining the associated features and user features of each unlocking information according to the historical behavior information, further includes:

[0112] S214, constructing a directed graph based on the time sequence of each unlocking information in the historical behavior information.

[0113] It can be understood that the directed graph includes a directed graph corresponding to each user ID in the historical behavior information.

[0114] Illustratively, the directed graph can be constructed according to the time sequence of the plurality of unlocking information corresponding to each user ID in the historical behavior information.

[0115] S215, calculating the attention weight between the nodes with directed edges in the directed graph, and determining the attention feature of each unlocking information according to the attention weight.

[0116] Illustratively, it can be a node pair (U_i, U_j) with a directed edge in the directed graph, for the edge U_i→U_j, using scaled dot-product attention to calculate the unnormalized attention weight, for all outgoing edges of U_i, using Softmax to normalize the attention weight, and according to the attention weight, aggregating the neighbor features of each unlocking information to obtain the attention feature.

[0117] S216, aggregating each attention feature to obtain the user feature.

[0118] It can be understood that the user feature includes a user feature corresponding to each user ID.

[0119] Illustratively, each attention feature corresponding to a certain user ID can be globally pooled (the mean or maximum value can be taken for all attention features), if it is necessary to retain the timing information, an LSTM or a Transformer can be used for further encoding, and a LayerNorm or a BatchNorm can be performed to obtain the user feature corresponding to the user ID.

[0120] By the above steps S214 to S216, the directed graph is encoded with the time dependence of the unlocking behavior through the edges. Compared with the unordered set, the graph structure can preserve the temporal context, which is convenient for subsequent analysis of behavior patterns. Mean pooling is simple and efficient, suitable for short sequences; long sequences can be modeled using LSTM to model temporal dependencies. By modeling the time sequence through the directed graph, capturing key dependencies through attention mechanisms, and aggregating to generate user-level features, the system can efficiently utilize historical behavior information and adapt to dynamic scenarios.

[0121] For example, refer to Figure 3 S215, calculate the attention weight between nodes with directed edges in the directed graph, and determine the attention feature of each unlocking information according to the attention weight, including:

[0122] S2151, initialize each node in the directed graph based on each unlocking information, to obtain the initial feature of each node.

[0123] For example, for each unlocking information, the computable features (such as statistics, type encoding of unlocking method) can be extracted, and the computable features projected through a linear layer are assigned to the corresponding nodes to form the initial node feature matrix of the directed graph.

[0124] S2152, calculate the attention weight between nodes with directed edges in the directed graph, and update the initial feature of the node according to the attention weight to obtain the attention feature of each unlocking information.

[0125] For example, the attention weight of the node can be calculated and normalized, the neighbor information is aggregated according to the attention weight, and the initial feature of the node is updated to obtain the attention feature.

[0126] Through the above steps S2151 to S2152, by initializing the semantic explicit node feature and dynamically aggregating neighbor information using the attention mechanism, the temporal dependence and key patterns in the unlocking behavior can be efficiently captured.

[0127] Optionally, refer to Figure 3 S220, based on the association feature and the user feature, the first behavior information is predicted, including:

[0128] S221, according to the association feature, the user feature and the latest unlocking information, the probability distribution information is predicted. The probability distribution information includes the probability distribution of the time period and the unlocking method of the next unlocking of the user.

[0129] It can be understood that the latest unlocking information can include the time and unlocking method of the last N unlocking.

[0130] Exemplarily, the three types of features, the association features, the user features of each user, and the latest unlocking information, can be spliced and compressed into a unified dimension through a fully connected layer, and then input into a prediction model to obtain the first behavior information. For example, the prediction model includes: a time prediction branch that uses Softmax to output the probability of a 24-hour time period (such as one probability value per hour), and a unlocking mode prediction branch that uses Softmax to output the probability of the unlocking mode (such as fingerprint, password, facial recognition, etc.), and the prediction model is obtained by jointly training the association features, the user features of each user, and the latest unlocking information in the data of each time period (such as 30 days) in the historical behavior information: the two branches share the underlying features, and the total loss is the cross-entropy weighted sum.

[0131] S222, obtaining the first behavior information according to the probability distribution information.

[0132] It can be understood that the first behavior information can include the time period and the unlocking mode of the next unlocking of different users.

[0133] Exemplarily, the time period and the unlocking mode combination with the highest probability of future unlocking of different users (such as “15:00-16:00 + fingerprint unlocking”) can be obtained according to the probability distribution information, if the user ID has been determined in advance according to the electronic device carried by the user, then the user is matched with the corresponding first behavior information predicted by the user to determine whether it is in the first time period; if the user ID cannot be determined in advance, then whether the time when the user approaches the intelligent lock is in the first time period is determined according to the first behavior information of all users predicted. Top-K behavior candidates (such as the top 3 high-probability user combinations) can also be generated to obtain the first behavior information. And if the highest probability is lower than a preset threshold, it is marked as “low confidence prediction”.

[0134] Through the above steps S221 to S222, in the fully trained model, the highest probability combination usually corresponds to the most likely behavior, and should avoid misprediction in the sparse data area (such as 3 am). For example, if the unlocking at midnight in the historical data is less than N times (such as 10 times), the model may be forced to output a low confidence prediction, at which time the prediction should be rejected instead of outputting an incorrect result. Multiple candidate behaviors can be generated and used for subsequent energy consumption planning of the intelligent lock.

[0135] It should be understood that the size of the serial number of each step in the above embodiments does not mean the order of execution, and the execution order of each process should be determined according to its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present application.

[0136] Corresponding to the intelligent lock energy consumption optimization control method described in the above embodiments, the embodiments of the present application also provide an intelligent lock energy consumption optimization control device, and each module of the device can realize each step of the intelligent lock energy consumption optimization control method. Figure 5A structural block diagram of the intelligent lock energy consumption optimization control device provided by the embodiment of the present application is shown. For ease of illustration, only the parts related to the embodiment of the present application are shown.

[0137] With reference to Figure 4 The device comprises:

[0138] An acquisition module, configured to acquire historical behavior information of a user using the intelligent lock in real time;

[0139] A prediction module, configured to obtain first behavior information according to the historical behavior information; wherein the first behavior information comprises a first time period of next unlocking of the user and a corresponding first unlocking mode;

[0140] A first operation module, configured to, in the first time period, perform a first operation according to the first unlocking mode when it is determined that the user approaches the intelligent lock; wherein the first operation refers to starting a sensor module and a corresponding identification module of the intelligent lock;

[0141] A receiving module, configured to, not in the first time period, perform a second operation when it is determined that the user approaches the intelligent lock, and receive a third operation of the intelligent lock after the second operation is performed by the user; wherein the second operation refers to starting only the sensor module, and the third operation refers to an input signal of the user;

[0142] A starting module, configured to start the corresponding identification module based on the third operation.

[0143] It should be noted that the information interaction, execution process and the like between the above modules are based on the same concept as the method embodiments of the present application, and the specific functions and the technical effects brought by the same can be referred to the method embodiments part, which will not be described herein.

[0144] Those skilled in the art can clearly understand that, for the convenience and brevity of description, only the division of the above functional units and modules is exemplified, and in actual application, the above functions can be completed by different functional units and modules according to needs, that is, the internal structure of the device is divided into different functional units or modules to complete all or part of the functions described above. The functional units and modules in the embodiments can be integrated in one processing unit, or each unit can exist physically, or two or more units can be integrated in one unit. The integrated unit can be realized in the form of hardware or in the form of software functional unit. In addition, the specific names of the functional units and modules are only for easy distinction, and do not limit the protection scope of the present application. The specific working process of the units and modules in the device can refer to the corresponding process in the foregoing method embodiments, which will not be described herein.

[0145] This application also provides a smart lock. Figure 5 This is a schematic diagram of the structure of a smart lock provided in one embodiment of this application. Figure 5 As shown, the smart lock 5 in this embodiment includes: at least one processor 50 ( Figure 5 Only one is shown in the image), at least one memory 51 ( Figure 5 (Only one is shown in the image) and a computer program 52 stored in the at least one memory 51 and executable on the at least one processor 50. When the processor 50 executes the computer program 52, it causes the smart lock 5 to implement the steps in any of the above-described smart lock energy consumption optimization control method embodiments, or causes the smart lock 5 to implement the functions of each module / unit in the above-described device embodiments.

[0146] For example, the computer program 52 may be divided into one or more modules / units, which are stored in the memory 51 and executed by the processor 50 to complete this application. The one or more modules / units may be a series of computer program instruction segments capable of performing a specific function, which describe the execution process of the computer program 52 in the smart lock 5.

[0147] The smart lock 5 includes sensors and an analysis device, wherein the analysis device can be a computing device such as a desktop computer, laptop, handheld computer, or cloud server. The smart lock may include, but is not limited to, a processor 50 and a memory 51. Those skilled in the art will understand that... Figure 5 Figure 5 This is merely an example of smart lock 5 and does not constitute a limitation on smart lock 5. It may include more or fewer components than shown in the figure, or combine certain components, or different components, such as input / output devices, network access devices, buses, etc.

[0148] The processor 50 can be a Central Processing Unit (CPU), or it can be other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor can be a microprocessor or any conventional processor.

[0149] The memory 51 can be an internal storage unit of the smart lock 5 in some embodiments, such as a hard disk or a memory of the smart lock 5. The memory 51 can also be an external storage device of the smart lock 5 in other embodiments, such as a plug-in hard disk, a smart media card (SMC), a secure digital (SD) card, a flash card, etc. equipped on the smart lock 5. Further, the memory 51 can include both the internal storage unit and the external storage device of the smart lock 5. The memory 51 is used to store an operating system, an application program, a boot loader, data, and other programs, such as program codes of the computer program, etc. The memory 51 can also be used to temporarily store data that has been output or will be output.

[0150] The computer readable storage medium stores a computer program, and the computer program is executed by a processor to implement the steps in any of the above method embodiments.

[0151] The computer program product, when executed on the smart lock, causes the smart lock to implement the steps in any of the above method embodiments.

[0152] The integrated unit, if implemented in the form of a software function unit and sold or used as an independent product, can be stored in a computer readable storage medium. Based on such understanding, the present application can implement all or part of the processes in the above embodiments by a computer program to instruct related hardware to complete, and the computer program can be stored in a computer readable storage medium. The computer program is executed by a processor to implement the steps of each method embodiment. The computer program includes computer program code, which can be in the form of source code, object code, executable files, or some intermediate forms. The computer readable medium at least includes any entity or device capable of carrying the computer program code to the smart lock, a recording medium, a computer memory, a read-only memory (ROM), a random access memory (RAM), an electrical carrier signal, a telecommunications signal, and a software distribution medium. For example, a U disk, a mobile hard disk, a magnetic disk or an optical disk, etc.

[0153] In the above embodiments, the description of each embodiment has its own focus, and the parts not described or recorded in detail in a certain embodiment can be referred to the relevant description of other embodiments.

[0154] Those skilled in the art can understand that the units and algorithm steps of each example described in combination with the embodiments disclosed herein can be realized in electronic hardware or a combination of computer software and electronic hardware. Whether the functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of the present application.

[0155] In the embodiments provided in the present application, it should be understood that the disclosed intelligent lock and method can be implemented in other ways. For example, the above-described intelligent lock embodiments are merely illustrative, for example, the division of the modules or units is only a logical function division, and actual implementation can have another division manner, for example, a plurality of units or components can be combined or integrated into another system, or some features can be ignored or not executed. In addition, the coupling or direct coupling or communication connection between the displayed or discussed each other can be indirect coupling or communication connection through some interfaces, devices or units, and can be electrical, mechanical or other forms.

[0156] The units described as separate components can or can not be physically separated, and the components shown as units can or can not be physical units, that is, they can be located in one place, or can be distributed on a plurality of network units. Part or all of the units can be selected according to actual needs to achieve the purpose of the embodiments of the present application.

[0157] The above-described embodiments are only used to illustrate the technical solutions of the present application, but not to limit them; although the present application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that the technical solutions recorded in the foregoing embodiments can be modified, or some technical features can be replaced by equivalents; and these modifications or replacements do not make the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present application, and should be included in the protection scope of the present application.

Claims

1. A method for optimizing energy consumption control in smart locks, characterized in that, Applied to smart locks; the method includes: Real-time acquisition of historical user behavior information regarding the smart lock; First behavior information is predicted based on the historical behavior information; wherein, the first behavior information includes the first time period of the user's next unlock and the corresponding first unlock method; During the first time period, if it is determined that a user is approaching the smart lock, a first operation is performed according to the first unlocking method; wherein, the first operation refers to activating the sensor module and the corresponding recognition module of the smart lock; If the user is not within the first time period, and the user is determined to be approaching the smart lock, a second operation is performed, and a third operation by the user on the smart lock after the second operation is performed is received; wherein, the second operation refers to activating only the sensor module, and the third operation refers to the user's input signal; The corresponding identification module is activated based on the third operation; The historical behavior information includes multiple unlocking information entries, each including the time and method of unlocking each time the user unlocks; the step of predicting the first behavior information based on the historical behavior information includes: Based on the historical behavior information, the association features and user features of each unlocking information are determined; wherein, the association features include the time features of the user using the smart lock and the corresponding behavior features, the behavior features are used to reflect the user's unlocking method, and the user features are used to reflect the user's behavior pattern; The first behavioral information is obtained by prediction based on the associated features and the user features; The step of determining the association features and user features of each unlocking information based on the historical behavior information includes: Based on each unlocking record in the historical behavior information, short-term features of each unlocking record are determined; wherein, the short-term features are used to reflect the local correlation of each unlocking record. Long-term features are determined based on the short-term features of each unlocking information; wherein, the long-term features are used to reflect the periodicity of each unlocking information. The smart lock's associated features are obtained by gating and fusing the short-term and long-term features, including: determining the weight of each long-term feature if the current time is determined to be during a user's historical peak period based on the historical behavior information; determining the weight of each short-term feature if an abnormal time interval is detected based on the historical behavior information; and gating and fusing the short-term and long-term features based on the weights of the long-term and short-term features to obtain the smart lock's associated features.

2. The smart lock energy consumption optimization control method as described in claim 1, characterized in that, The method further includes: The operating parameters of the smart lock are adjusted based on the first behavioral information; wherein the operating parameters include the working cycle of the smart lock's identification module and the sampling frequency of the sensor module.

3. The smart lock energy consumption optimization control method as described in claim 1, characterized in that, The determination of short-term characteristics of each unlocking record based on the historical behavior information includes: The first correlation between each unlocking information is calculated to obtain the feature matrix corresponding to each unlocking information; wherein, the first correlation is used to reflect the time interval of each unlocking information and the correlation of the corresponding unlocking method within the second time period; The short-term features of each unlocking message are obtained based on the respective feature matrices.

4. The smart lock energy consumption optimization control method as described in claim 1, characterized in that, The determination of long-term features based on the short-term features of each unlocking information includes: Based on the aforementioned short-term features, the mutual information value between each unlocking information is calculated to obtain the feature vector corresponding to each unlocking information; Based on the similarity between the feature vectors, feature enhancement is performed on each feature vector to obtain the long-term feature of each unlocking information.

5. The smart lock energy consumption optimization control method as described in claim 1, characterized in that, The step of determining the association features and user features of each unlocking information based on the historical behavior information also includes: A directed graph is constructed based on the temporal order of each unlocking information in the historical behavior information; Calculate the attention weights between nodes with directed edges in the directed graph, and determine the attention features of each unlocking message based on the attention weights; The user features are obtained by aggregating the attention features.

6. The smart lock energy consumption optimization control method as described in claim 5, characterized in that, The calculation of attention weights between nodes with directed edges in the directed graph, and the determination of attention features for each unlocking message based on these attention weights, includes: Initialize each node in the directed graph based on the unlocking information to obtain the initial features of each node; Calculate the attention weights between nodes with directed edges in the directed graph, and update the initial features of the nodes according to the attention weights to obtain the attention features of each unlocking message.

7. The smart lock energy consumption optimization control method as described in claim 1, characterized in that, The process of predicting the first behavioral information based on the associated features and the user features includes: The probability distribution information is predicted based on the associated features, the user features, and the most recent unlocking information; wherein, the probability distribution information includes the probability distribution of the time period and unlocking method for the user's next unlock; The first behavioral information is obtained based on the probability distribution information.

8. A smart lock, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the computer program, it implements the method as described in any one of claims 1 to 7.

Images