Intelligent lock with intelligent acquisition function and control method thereof

Abstract

The invention relates to the technical field of intelligent locks, in particular to an intelligent lock with an intelligent acquisition function and a control method thereof.Importance levels of corresponding acquisition functions are represented through attribute weights, then the attribute weights are corrected through data calling records, and personalized weights better conforming to user habits are obtained; then a fitness function is determined based on the personalized weight, so that an optimal acquisition function calling strategy can be found through an optimization algorithm, and the calling strategy not only can select and reject a plurality of acquisition functions according to the residual electric quantity so as to prolong standby time, but also can match user habits and reserve common and important acquisition functions of a user; the purpose of balancing energy consumption and functions of the intelligent lock is perfectly achieved, it is ensured that safety and convenience are provided, high efficiency and energy efficiency are kept, user experience is greatly improved, and good prospects are achieved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of intelligent locks, and in particular to an intelligent lock with intelligent collection function and a control method thereof. BACKGROUND

[0002] Modern intelligent locks integrate a variety of data collection functions, such as multi-modal authentication, which allows users to unlock through multiple ways such as fingerprint, facial recognition, etc., improving security and convenience. In addition, the intelligent lock can also monitor abnormal behavior in front of the door through loitering identification, issue timely alarms, and through temperature environment monitoring function, help users understand indoor and outdoor temperature, even link with smart home system, automatically adjust indoor temperature, and improve living comfort.

[0003] However, these various collection functions also bring energy consumption problems. The advanced functions of the intelligent lock require more energy support, which may cause the battery life to be shortened, and need to be replaced or charged more frequently. If energy consumption is not properly managed, it may also affect normal use due to insufficient power at critical moments, causing inconvenience to users.

[0004] Therefore, there is a need for an intelligent lock with intelligent collection function and a control method thereof that can balance energy consumption and function. SUMMARY

[0005] Therefore, the present application provides an intelligent lock with intelligent collection function and a control method thereof to solve the problem that the intelligent lock in the prior art cannot balance energy consumption and multiple collection functions.

[0006] The present application provides an intelligent lock with intelligent collection function, comprising: a data monitoring module for obtaining the remaining power of the intelligent lock, the attribute weight of the multiple collection functions, and the data call record of each collection function, wherein the attribute weight represents the importance level of the corresponding collection function; a personalized analysis module for correcting the attribute weight of each collection function according to the data call record to obtain the personalized weight of each collection function; a first data preparation module for establishing a plurality of initial collection strategy vectors with the remaining power as a constraint, wherein the collection strategy vector is used to describe the calling strategy of the multiple collection functions before the remaining power is exhausted; a second data preparation module for establishing a fitness function based on the format of the personalized weight and the collection strategy vector, wherein the fitness function is used to describe the rationality of a collection strategy vector; an optimization module for optimizing the plurality of initial collection strategy vectors based on the fitness function using a preset optimization algorithm to obtain an optimal collection strategy vector; an execution module for calling the multiple collection functions according to the optimal collection strategy vector.

[0007] The application further provides a preferred scheme: the attribute weight comprises a first-level weight and a second-level weight, and the first-level weight represents a higher importance level than the second-level weight; according to the data calling record, the attribute weight of each collection function is corrected to obtain the personalized weight of each collection function, comprising: obtaining the use frequency of each collection function according to the data calling record; correcting the first-level weight according to the following formula to obtain the personalized weight of a collection function corresponding to the first-level weight: ; wherein, the personalized weight of a collection function corresponding to the first-level weight, the first-level weight, a natural constant, the use frequency of the collection function corresponding to the first-level weight, a first preset unit adjustment coefficient; correcting the second-level weight according to the following formula to obtain the personalized weight of a collection function corresponding to the second-level weight: ; wherein, the personalized weight of a collection function corresponding to the second-level weight, the second-level weight, a preset linear increasing function, the use frequency of the collection function corresponding to the second-level weight, a second preset unit adjustment coefficient.

[0008] The application further provides a preferred scheme: the attribute weight further comprises a third-level weight, and the third-level weight represents a lower importance level than the second-level weight; according to the data calling record, the attribute weight of each collection function is corrected to obtain the personalized weight of each collection function, further comprising: correcting the third-level weight according to the following formula to obtain the personalized weight of a collection function corresponding to the third-level weight: ; wherein, the personalized weight of a collection function corresponding to the third-level weight, the third-level weight, a natural logarithmic function, the use frequency of the collection function corresponding to the third-level weight, a third preset unit adjustment coefficient.

[0009] The application further provides a preferred scheme: in the collection strategy vector, each element corresponds to a collection function, and the element value of each element is used to represent the opening duration of the collection function corresponding to the element before the remaining power of the smart lock is consumed.

[0010] The application further provides a preferred scheme: the fitness function is: ; wherein, represents the fitness of a collection strategy vector, is the maximum value in a plurality of elements corresponding to a plurality of collection functions corresponding to the attribute weight with the highest importance level in the collection strategy vector, is an element value in the collection strategy vector, is the personalized weight of the collection function corresponding to the element value, and are different preset influence weights, respectively.

[0011] The application further provides a preferred scheme: the preset optimization algorithm is a particle swarm optimization algorithm; based on the fitness function, the optimization algorithm is used to optimize a plurality of initial collection strategy vectors to obtain an optimal collection strategy vector, including: obtaining the position and speed of a target particle, wherein the target particle is a particle corresponding to a collection strategy vector with a current position and speed to be updated in the particle swarm optimization algorithm; obtaining the historical optimal value and the global optimal value of a target element in the target particle, wherein the target element is an element in one dimension of the collection strategy vector with the current position and speed to be updated in the target particle; obtaining a first learning rate corresponding to the target element according to the personalized weight of the collection function corresponding to the target element, wherein the first learning rate is inversely proportional to the personalized weight; updating the position and speed of the target element in the target particle according to the first learning rate.

[0012] The application further provides a preferred scheme: the position and speed of the target element in the target particle are updated according to the first learning rate, and the method further includes: obtaining a second learning rate according to the numerical size relationship between the target element and the historical optimal value; obtaining a third learning rate according to the numerical size relationship between the target element and the global optimal value; updating the position and speed of the target element in the target particle according to the first learning rate, the second learning rate and the third learning rate.

[0013] The application further provides a preferred scheme: if the attribute weight of the collection function corresponding to the target element represents the highest importance level, then: When the difference between the historical best value and the target element value is positive, the second learning rate is proportional to the difference between the historical best value and the target element value; when the difference between the historical best value and the target element value is negative, the second learning rate is zero. When the difference between the global optimum and the target element value is positive, the third learning rate is proportional to the difference between the global optimum and the target element value; when the difference between the global optimum and the target element value is negative, the third learning rate is zero. If the attribute weight representing the importance level of the collection function corresponding to the target element is the lowest, then: When the difference between the historical best value and the target element value is negative, the second learning rate is proportional to the absolute value of the difference between the historical best value and the target element value; when the difference between the historical best value and the target element value is positive, the second learning rate is zero. When the difference between the global optimum and the target element is negative, the third learning rate is proportional to the absolute value of the difference between the global optimum and the target element; when the difference between the global optimum and the target element is positive, the third learning rate is zero.

[0014] The present invention also provides a preferred embodiment: updating the position and velocity of the target element in the target particle according to a first learning rate, a second learning rate, and a third learning rate, including: The velocity of the target element in the target particle is updated using the following formula: in, Indicates the target element is in the th position. Speed ​​at the next iteration Indicates the target element is in the th position. Speed ​​at the next iteration and Each has a different preset weight ratio. , and These are the first learning rate, the second learning rate, and the third learning rate, respectively. It is a random number. Indicates the target element is in the th position. Position at the next iteration Indicates the target element is in the th position. The historical best value at the next iteration Indicates the target element is in the th position. The global optimum value at the next iteration.

[0015] This invention also provides a control method for a smart lock with intelligent data acquisition function, comprising: The system obtains the remaining battery power of the smart lock, the attribute weights of various data collection functions, and the data call records of each data collection function. The attribute weights represent the importance level of the corresponding data collection function. According to the data call record, the attribute weight of each collection function is corrected to obtain the personalized weight of each collection function; A plurality of initial collection strategy vectors are established with the residual power as a constraint condition, wherein the collection strategy vector is used to describe the calling strategy of the plurality of collection functions before the residual power is consumed; Based on the personalized weight and the format of the collection strategy vector, a fitness function is established to describe the rationality of a collection strategy vector; Based on the fitness function, a plurality of initial collection strategy vectors are optimized by using a preset optimization algorithm to obtain an optimal collection strategy vector; According to the optimal collection strategy vector, the plurality of collection functions are called.

[0016] The beneficial effects of the above embodiments are: The present application provides a smart lock with intelligent collection function and a control method thereof, which obtains the residual power of the smart lock, the attribute weight of the plurality of collection functions and the data call record of each collection function through the data monitoring module, then corrects the attribute weight of each collection function according to the data call record through the personalized analysis module to obtain the personalized weight of each collection function, establishes a plurality of initial collection strategy vectors with the residual power as a constraint condition through the first data preparation module, establishes a fitness function based on the personalized weight and the format of the collection strategy vector through the second data preparation module, then optimizes a plurality of initial collection strategy vectors by using a preset optimization algorithm based on the fitness function through the optimization module to obtain an optimal collection strategy vector, and finally, the execution module is used to call the plurality of collection functions according to the optimal collection strategy vector. Compared with the prior art, the present application represents the importance level of the corresponding collection function by the attribute weight, then corrects the attribute weight by using the data call record to obtain the personalized weight more in line with the user's habits, and then determines the fitness function based on the personalized weight, so that the optimal collection function calling strategy can be found by using the optimization algorithm. The calling strategy not only can select the plurality of collection functions according to the residual power to prolong the standby time, but also matches the user's habits and retains the collection functions commonly used and important by the user, perfectly achieving the goal of balancing the energy consumption and the function of the smart lock, ensuring safety and convenience, maintaining high energy efficiency, greatly improving the user experience, and having a good prospect. BRIEF DESCRIPTION OF DRAWINGS

[0017] Figure 1 A module architecture diagram of an embodiment of the smart lock with intelligent collection function provided by the present application is provided. Figure 2 A method flowchart of the control method of the smart lock with intelligent collection function provided by the present application is provided. DETAILED DESCRIPTION

[0018] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0019] Combination Figure 1 As shown, a specific embodiment of the present invention discloses a smart lock with intelligent data acquisition function, comprising: The data monitoring module 110 is used to obtain the remaining power of the smart lock, the attribute weights of various data collection functions, and the data call records of each data collection function. The attribute weights represent the importance level of the corresponding data collection function. The personalized analysis module 120 is used to adjust the attribute weights of each collection function based on the data retrieval records, and obtain the personalized weights of each collection function. The first data preparation module 130 is used to establish multiple initial acquisition strategy vectors with the remaining power as a constraint. The acquisition strategy vectors are used to describe the calling strategies of various acquisition functions of the smart lock before the remaining power is exhausted. The second data preparation module 140 is used to establish a fitness function based on the format of personalized weights and collection strategy vectors. The fitness function is used to describe the rationality of a collection strategy vector. The optimization module 150 is used to optimize multiple initial acquisition strategy vectors based on the fitness function and a preset optimization algorithm to obtain the optimal acquisition strategy vector. The execution module 160 is used to invoke various acquisition functions based on the optimal acquisition strategy vector.

[0020] In the above process, the data call record refers to the number of times the data collected by each collection function is used, rather than the number of times it is collected. For example, in the password verification function, the number of times the entered password is successfully verified; the number of times the collected image data is called by the loitering monitoring function or the number of times it is called by the face recognition function; the number of times the collected sound, temperature, humidity and other data are used by other intelligent algorithms (such as fault early warning algorithms, smart home linkage algorithms, etc.).

[0021] Furthermore, the various modules mentioned above can be integrated into the smart lock depending on their specific computing power. For example, the optimization module needs to run a preset optimization algorithm. However, due to factors such as cost and stability, the smart lock may not be able to integrate expensive hardware with high computing power. In this case, only the hardware in the smart lock with data transmission and reception functions can be used as the optimization module, and the specific calculation process of the optimization module described above can be handed over to a remote server.

[0022] Compared to existing technologies, this invention characterizes the importance level of corresponding data collection functions through attribute weights, then uses data retrieval records to correct the attribute weights, resulting in personalized weights that better suit user habits. Based on these personalized weights, a fitness function is determined. This allows for the optimization algorithm to find the optimal data collection function retrieval strategy. This strategy not only allows for the selection of multiple data collection functions based on remaining battery power to extend standby time, but also matches user habits, retaining frequently used and important data collection functions. It perfectly achieves the goal of balancing the energy consumption and functionality of smart locks, ensuring both security and convenience while maintaining high energy efficiency, greatly improving the user experience and demonstrating great promise.

[0023] Furthermore, in a preferred embodiment, the attribute weights include primary weights and secondary weights, with the primary weights representing a higher level of importance than the secondary weights; the steps performed by the personalized analysis module 120 are as follows: based on the data retrieval records, the attribute weights of each acquisition function are corrected to obtain the personalized weights for each acquisition function, including: Based on the data retrieval records, the usage frequency of each acquisition function can be obtained; The primary weights are adjusted according to the following formula to obtain the personalized weights for a particular data collection function corresponding to the primary weights: ; in, This refers to a personalized weight for a data collection function corresponding to the primary weight. As a first-level weight, It is a natural constant. This refers to the frequency of use of the data collection function corresponding to this primary weight. Adjust the coefficient for the first preset unit; The secondary weights are adjusted according to the following formula to obtain the personalized weights for a data collection function corresponding to the secondary weights: ; in, This refers to a personalized weight for a data collection function corresponding to the secondary weight. As a secondary weight, As a pre-defined linearly increasing function, This refers to the frequency of use of the data collection function corresponding to this secondary weight. Adjust the coefficient for the second preset unit.

[0024] In the above process, the primary weight corresponds to the core functions of the smart lock related to unlocking, such as fingerprint recognition, voiceprint recognition, NFC recognition, and password recognition, while the secondary weight corresponds to relatively less important data collection functions. In this embodiment, the primary and secondary weights are adjusted based on usage frequency to make the subsequent data collection function invocation strategy more in line with user habits. For example, if the user uses voiceprint recognition to unlock less frequently over a period of time, the personalized weight of voiceprint recognition after adjustment will also be lower. Thus, in the subsequent optimization process, this method will be more inclined to find invocation methods that shorten the activation time of the voiceprint recognition function.

[0025] Furthermore, it is worth noting that the primary and secondary weights corrected by the method in this embodiment are more sensitive to frequency than the secondary weights. For example, when the usage frequency is the same, the personalized weight corresponding to the primary weight obtained by this method will be significantly higher than the personalized weight corresponding to the secondary weight. This allows this method to preserve the runtime of the data collection function corresponding to the primary weight as much as possible during subsequent optimization, and tends to obtain the optimal calling strategy by adjusting the duration of the data collection function corresponding to the secondary weight. This ensures that even if all other data collection functions are turned off in extreme cases, the smart lock can still maintain the normal operation of the critical function of unlocking, avoiding a situation where unlocking is impossible due to the need to extend battery life, which contradicts the functional purpose of the smart lock itself.

[0026] Furthermore, in a preferred embodiment, the attribute weights also include third-level weights, where the importance level represented by the third-level weights is lower than that of the second-level weights; the steps performed by the personalized analysis module 120 above—correcting the attribute weights of each acquisition function according to the data retrieval records to obtain the personalized weights of each acquisition function—also include: The three-level weights are adjusted according to the following formula to obtain the personalized weights for a data collection function corresponding to the three-level weights: ; in, This refers to a personalized weighting for a data collection function, corresponding to a three-tier weighting system. It has three levels of weight. It is the natural logarithm function. The frequency of use of the data collection function corresponding to these three weight levels. Adjust the coefficient for the third preset unit.

[0027] The above process adds a third level of weight on top of the second level of weight. The data collection functions corresponding to the third level of weight are some functions that are less important than the second level of weight. For example, if the second level of weight is a security function such as loitering monitoring, then the third level of weight can be other icing on the cake functions that are unrelated to unlocking and security, such as temperature and environmental monitoring functions for linkage with smart homes.

[0028] Compared to the correction formula for the second-level weights, the formula provided in this embodiment further reduces the sensitivity of the third-level weights to usage frequency. This results in a clear gradient when the first-level, second-level, and third-level weights are corrected. After correction, the personalized weights corresponding to the first-level, second-level, and third-level weights will exhibit a clear three-level division. This makes the system more inclined to adjust the duration of the collection function corresponding to the least important third-level weight when searching for the optimal calling strategy, followed by the duration of the collection function corresponding to the second-level weight, and finally the duration of the collection function corresponding to the first-level weight. This is to maintain the operation of the core unlocking function of the smart lock as much as possible and ensure user experience.

[0029] Furthermore, in a preferred embodiment, during the operation of the first data preparation module 130, each element in the acquisition strategy vector corresponds to a specific acquisition function. The value of each element represents the duration for which the corresponding acquisition function will be active before the smart lock's remaining battery power is depleted. This acquisition strategy vector digitally represents the operation strategy of the acquisition function using the remaining runtime, allowing subsequent preset optimization algorithms to perform optimization.

[0030] Specifically, in a preferred embodiment, in the steps of the second data preparation module 140, the fitness function is: ; in, This represents the fitness of a collection strategy vector. In the data collection strategy vector, the maximum value among the multiple elements corresponding to the highest importance attribute weight for each data collection function is considered. To collect the element values ​​in the strategy vector, This assigns a personalized weight to the data collection function corresponding to that element value. and Each has a different preset influence weight.

[0031] The significance of the fitness function mentioned above is that the first term measures the duration of the core function in a data collection strategy, and the weighted summation of the second term calculates the degree of matching between a data collection strategy and user behavior. This allows the fitness function to characterize both the security and practicality of a data collection strategy itself, as well as the degree of personalization of the strategy, making the final optimal data collection strategy more scientific and reasonable.

[0032] Furthermore, in a preferred embodiment, the preset optimization algorithm is the particle swarm optimization algorithm; the steps performed by the optimization module 150 are as follows: based on the fitness function, the optimization algorithm is used to optimize multiple initial acquisition strategy vectors to obtain the optimal acquisition strategy vector, specifically including: Obtain the position and velocity of the target particle, where the target particle is the particle corresponding to a sampling strategy vector whose position and velocity are currently to be updated in the particle swarm algorithm; Obtain the historical optimal value and global optimal value of the target element in the target particle. The target element is an element in the target particle whose current position and velocity are to be updated in one dimension. Based on the personalized weights of the collection functions corresponding to the target element, the first learning rate corresponding to the target element is obtained. The first learning rate is inversely proportional to the personalized weights. Update the position and velocity of the target element in the target particle based on the first learning rate.

[0033] In practice, after determining the format and fitness function of the acquisition strategy vector, any existing preset optimization algorithm can be used for optimization, such as a genetic algorithm. This embodiment uses a particle swarm optimization (PSO) algorithm. In this PSO algorithm, the value range of each element in the acquisition strategy vector is considered as one dimension, and the resulting high-dimensional space is the solution space. Each specific acquisition strategy vector can be viewed as a particle existing at a specific position in the solution space. The PSO algorithm finds the optimal position in the solution space through the continuous movement of the particle swarm; the "coordinates" of this position are the optimal acquisition strategy vector. The learning rate can be understood as the step size of each particle's movement. It is understood that the concepts of particle swarm optimization and learning rate are existing technologies that are understandable to those skilled in the art, and will not be elaborated upon further in this paper.

[0034] The significance of this embodiment lies in further improving the learning rate based on the existing particle swarm optimization algorithm. By setting a first learning rate that is inversely proportional to the personalized weights, the convergence speed and optimization tendency of the entire optimization algorithm are changed, leading to a faster attainment of the optimal solution. Specifically, in this embodiment, if the personalized weight of a certain data collection function is larger, it indicates that the data collection function is more important to the user. In this case, reducing the learning rate of that data collection function can force the preset optimization algorithm to be more inclined to find the optimal data collection strategy vector faster and more comprehensively by changing the data collection functions with lower personalized weights. This allows for finding a data collection strategy that better matches the user's habits under the condition of equal remaining battery power.

[0035] Furthermore, in a preferred embodiment, updating the position and velocity of the target element in the target particle according to the first learning rate further includes: The second learning rate is obtained based on the numerical relationship between the target element and the historical best value; The third learning rate is obtained based on the numerical relationship between the target element and the global optimum. Update the position and velocity of the target element in the target particle based on the first learning rate, the second learning rate, and the third learning rate.

[0036] In this embodiment, a second learning rate and a third learning rate are added based on the first learning rate. The second and third learning rates will change dynamically according to the movement of the particles in the solution space, so as to further control the movement of the particles during the optimization process and achieve convergence faster and more accurately.

[0037] Specifically, in a preferred embodiment, if the attribute weight representing the importance level of the acquisition function corresponding to the target element is the highest, then: When the difference between the historical best value and the target element value is positive, the second learning rate is proportional to the difference between the historical best value and the target element value; when the difference between the historical best value and the target element value is negative, the second learning rate is zero. When the difference between the global optimum and the target element value is positive, the third learning rate is proportional to the difference between the global optimum and the target element value; when the difference between the global optimum and the target element value is negative, the third learning rate is zero. If the attribute weight representing the importance level of the collection function corresponding to the target element is the lowest, then: When the difference between the historical best value and the target element value is negative, the second learning rate is proportional to the absolute value of the difference between the historical best value and the target element value; when the difference between the historical best value and the target element value is positive, the second learning rate is zero. When the difference between the global optimum and the target element is negative, the third learning rate is proportional to the absolute value of the difference between the global optimum and the target element; when the difference between the global optimum and the target element is positive, the third learning rate is zero.

[0038] The above process is an optimal rule for the changes of the second and third learning rates. It mainly uses the attribute weights and the direction of particle movement to dynamically influence the optimization trend, reduce meaningless computation, and improve convergence speed and accuracy.

[0039] For example, if a certain acquisition function has the highest attribute weight, it indicates that this acquisition function is the most important. Therefore, when optimizing, we should try to find an acquisition strategy vector that maintains this acquisition function for a longer period of time. In this case, we can increase the step size of the particle's positive movement in the dimension of this acquisition function (i.e., when the difference between the historical best value and the target element value is positive, increase the value of the second learning rate, and the same applies to the third learning rate), and decrease the step size of the particle's negative movement in the dimension of this acquisition function (i.e., when the difference between the historical best value and the target element value is negative, directly ignore the second learning rate and set it to zero, and the same applies to the third learning rate).

[0040] Similarly, if the attribute weight of a collection function is the lowest, it indicates that the collection function is the least important. Therefore, when optimizing, we should try to find a collection strategy vector with a shorter duration of the collection function. In this case, we can increase the step size of the particle's negative movement in the dimension of the collection function (i.e., when the difference between the historical best value and the target element value is negative, increase the value of the second learning rate, and the same applies to the third learning rate), and decrease the step size of the particle's positive movement in the dimension of the collection function (i.e., when the difference between the historical best value and the target element value is positive, directly ignore the second learning rate and set it to zero, and the same applies to the third learning rate).

[0041] Furthermore, based on the above ideas, the specific improvement of the existing particle swarm optimization algorithm in this invention is manifested in: updating the position and velocity of the target element in the target particle according to the first learning rate, the second learning rate, and the third learning rate, including: The velocity of the target element in the target particle is updated using the following formula: in, Indicates the target element is in the th position. Speed ​​at the next iteration Indicates the target element is in the th position. Speed ​​at the next iteration and Each has a different preset weight ratio. , and These are the first learning rate, the second learning rate, and the third learning rate, respectively. It is a random number. Indicates the target element is in the th position. Position at the next iteration Indicates the target element is in the th position. The historical best value at the next iteration Indicates the target element is in the th position. The global optimum value at the next iteration.

[0042] The above formula makes the particle swarm algorithm in this embodiment more inclined to find a collection strategy that conforms to the user's habits, so as to achieve a balance between energy consumption, functionality and personalization.

[0043] Combination Figure 2 As shown, the present invention also provides a control method for a smart lock with intelligent data acquisition function, comprising: S201. Obtain the remaining battery power of the smart lock, the attribute weights of various data collection functions, and the data call records of each data collection function. The attribute weights represent the importance level of the corresponding data collection function. S202. Based on the data call records, adjust the attribute weights of each collection function to obtain the personalized weights of each collection function; S203. Using the remaining power as a constraint, establish multiple initial acquisition strategy vectors, wherein the acquisition strategy vectors are used to describe the calling strategies of various acquisition functions of the smart lock before the remaining power is exhausted. S204. Based on the format of personalized weights and acquisition strategy vectors, establish a fitness function. The fitness function is used to describe the rationality of an acquisition strategy vector. S205. Based on the fitness function, a preset optimization algorithm is used to optimize multiple initial acquisition strategy vectors to obtain the optimal acquisition strategy vector. S206. Based on the optimal acquisition strategy vector, call up multiple acquisition functions.

[0044] It should be noted that the corresponding methods provided in the above embodiments can implement the technical solutions described in the above method embodiments. The specific implementation principles of each step can be found in the corresponding content in the above method embodiments, and will not be repeated here.

[0045] This invention provides a smart lock with intelligent data acquisition function and its control method. The method acquires the remaining battery power, attribute weights of various data acquisition functions, and data call records for each function through a data monitoring module. Then, a personalized analysis module adjusts the attribute weights of each data acquisition function based on the data call records to obtain personalized weights for each function. A first data preparation module establishes multiple initial data acquisition strategy vectors with the remaining battery power as a constraint. A second data preparation module establishes a fitness function based on the personalized weights and the format of the data acquisition strategy vectors. An optimization module then optimizes the multiple initial data acquisition strategy vectors using a preset optimization algorithm based on the fitness function to obtain the optimal data acquisition strategy vector. Finally, an execution module invokes various data acquisition functions according to the optimal data acquisition strategy vector. Compared to existing technologies, this invention characterizes the importance level of corresponding data collection functions through attribute weights, then uses data retrieval records to correct the attribute weights, resulting in personalized weights that better suit user habits. Based on these personalized weights, a fitness function is determined. This allows for the optimization algorithm to find the optimal data collection function retrieval strategy. This strategy not only allows for the selection of multiple data collection functions based on remaining battery power to extend standby time, but also matches user habits, retaining frequently used and important data collection functions. It perfectly achieves the goal of balancing the energy consumption and functionality of smart locks, ensuring both security and convenience while maintaining high energy efficiency, greatly improving the user experience and demonstrating great promise.

[0046] It should be noted that the various embodiments in this specification are described in a progressive manner, with each embodiment focusing on the differences from other embodiments. The same or similar parts between the various embodiments can be referred to each other.

[0047] The above description of the disclosed embodiments enables those skilled in the art to make or use the invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the invention. Therefore, the invention is not to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A smart lock with intelligent data acquisition function, characterized in that, include: The data monitoring module is used to obtain the remaining battery power of the smart lock, the attribute weights of various data collection functions, and the data call records of each data collection function. The attribute weights represent the importance level of the corresponding data collection function. The personalized analysis module is used to adjust the attribute weights of each collection function based on the data retrieval records, thereby obtaining the personalized weights of each collection function. The first data preparation module is used to establish multiple initial acquisition strategy vectors with the remaining power as a constraint. The acquisition strategy vectors are used to describe the calling strategies of various acquisition functions of the smart lock before the remaining power is exhausted. The second data preparation module is used to establish a fitness function based on the format of personalized weights and collection strategy vectors. The fitness function is used to describe the rationality of a collection strategy vector. The optimization module is used to optimize multiple initial acquisition strategy vectors based on the fitness function and a preset optimization algorithm to obtain the optimal acquisition strategy vector. The execution module is used to invoke various acquisition functions based on the optimal acquisition strategy vector.

2. The smart lock with intelligent data acquisition function according to claim 1, characterized in that, Attribute weights include primary weights and secondary weights, with primary weights representing a higher level of importance than secondary weights; Based on the data retrieval records, the attribute weights of each data collection function are adjusted to obtain personalized weights for each function, including: Based on the data retrieval records, the usage frequency of each acquisition function can be obtained; The primary weights are adjusted according to the following formula to obtain the personalized weights for a particular data collection function corresponding to the primary weights: ； in, This refers to a personalized weight for a data collection function corresponding to the primary weight. As a first-level weight, It is a natural constant. This refers to the frequency of use of the data collection function corresponding to this primary weight. Adjust the coefficient for the first preset unit; The secondary weights are adjusted according to the following formula to obtain the personalized weights for a data collection function corresponding to the secondary weights: ； in, This refers to a personalized weight for a data collection function corresponding to the secondary weight. It is a secondary weight. As a pre-defined linearly increasing function, This refers to the frequency of use of the data collection function corresponding to this secondary weight. Adjust the coefficient for the second preset unit.

3. The smart lock with intelligent data acquisition function according to claim 2, characterized in that, Attribute weights also include third-level weights, where the importance level is lower than that of the second-level weights; Based on the data retrieval records, the attribute weights of each data collection function are adjusted to obtain personalized weights for each function, including: The three-level weights are adjusted according to the following formula to obtain the personalized weights for a data collection function corresponding to the three-level weights: ； in, This refers to a personalized weighting for a data collection function, corresponding to a three-tier weighting system. It has three levels of weight. It is the natural logarithm function. The frequency of use of the data collection function corresponding to these three weight levels. Adjust the coefficient for the third preset unit.

4. The smart lock with intelligent data acquisition function according to claim 1, characterized in that, In the data acquisition strategy vector, each element corresponds to a data acquisition function. The value of each element is used to represent the duration of the data acquisition function corresponding to that element before the smart lock runs out of power.

5. The smart lock with intelligent data acquisition function according to claim 4, characterized in that, The fitness function is: ； in, This represents the fitness of a collection strategy vector. In the data collection strategy vector, the maximum value among the multiple elements corresponding to the highest importance attribute weight for each data collection function is considered. To collect element values ​​from the strategy vector, This assigns a personalized weight to the data collection function corresponding to that element value. and Each has a different preset influence weight.

6. The smart lock with intelligent data acquisition function according to claim 1, characterized in that, The default optimization algorithm is particle swarm optimization. Based on the fitness function, an optimization algorithm is used to optimize multiple initial acquisition strategy vectors to obtain the optimal acquisition strategy vector, including: Obtain the position and velocity of the target particle, where the target particle is the particle corresponding to a sampling strategy vector whose position and velocity are currently to be updated in the particle swarm algorithm; Obtain the historical optimal value and global optimal value of the target element in the target particle. The target element is an element in the target particle whose current position and velocity are to be updated in one dimension. Based on the personalized weights of the collection functions corresponding to the target element, the first learning rate corresponding to the target element is obtained. The first learning rate is inversely proportional to the personalized weights. Update the position and velocity of the target element in the target particle based on the first learning rate.

7. The smart lock with intelligent data acquisition function according to claim 6, characterized in that, Based on the first learning rate, the position and velocity of the target elements in the target particle are updated, including: The second learning rate is obtained based on the numerical relationship between the target element and the historical best value; The third learning rate is obtained based on the numerical relationship between the target element and the global optimum. Update the position and velocity of the target element in the target particle based on the first learning rate, the second learning rate, and the third learning rate.

8. The smart lock with intelligent data acquisition function according to claim 7, characterized in that, If the attribute weight representing the importance level of the collection function corresponding to the target element is the highest, then: When the difference between the historical best value and the target element value is positive, the second learning rate is proportional to the difference between the historical best value and the target element value; when the difference between the historical best value and the target element value is negative, the second learning rate is zero. When the difference between the global optimum and the target element value is positive, the third learning rate is proportional to the difference between the global optimum and the target element value; when the difference between the global optimum and the target element value is negative, the third learning rate is zero. If the attribute weight representing the importance level of the collection function corresponding to the target element is the lowest, then: When the difference between the historical best value and the target element value is negative, the second learning rate is proportional to the absolute value of the difference between the historical best value and the target element value; when the difference between the historical best value and the target element value is positive, the second learning rate is zero. When the difference between the global optimum and the target element is negative, the third learning rate is proportional to the absolute value of the difference between the global optimum and the target element; when the difference between the global optimum and the target element is positive, the third learning rate is zero.

9. The smart lock with intelligent data acquisition function according to claim 7, characterized in that, Based on the first learning rate, the second learning rate, and the third learning rate, update the position and velocity of the target element in the target particle, including: The velocity of the target element in the target particle is updated using the following formula: in, Indicates the target element is in the th position. Speed ​​at the next iteration Indicates the target element is in the th position. Speed ​​at the next iteration and Each has a different preset weight ratio. , and These are the first learning rate, the second learning rate, and the third learning rate, respectively. It is a random number. Indicates the target element is in the th position. Position at the next iteration Indicates the target element is in the th position. The historical best value at the next iteration Indicates the target element is in the th position. The global optimum value at the next iteration.

10. A control method for a smart lock with intelligent data acquisition function, characterized in that, include: The system obtains the remaining battery power of the smart lock, the attribute weights of various data collection functions, and the data call records of each data collection function. The attribute weights represent the importance level of the corresponding data collection function. Based on the data call records, the attribute weights of each collection function are adjusted to obtain personalized weights for each collection function. Using the remaining power as a constraint, multiple initial acquisition strategy vectors are established. These acquisition strategy vectors describe the invocation strategies of various acquisition functions of the smart lock before the remaining power is exhausted. Based on the format of personalized weights and acquisition strategy vectors, a fitness function is established. The fitness function is used to describe the rationality of an acquisition strategy vector. Based on the fitness function, a preset optimization algorithm is used to optimize multiple initial acquisition strategy vectors to obtain the optimal acquisition strategy vector. Based on the optimal acquisition strategy vector, multiple acquisition functions are invoked.

Images