An AI-based Adaptive Perception Early Warning System and Method

Abstract

The present invention discloses an adaptive perception early warning system and method based on artificial intelligence, belonging to the field of intelligent early warning technology. Data files are compiled through the block clouds of different users; the block clouds of each user are updated in a periodic cycle; based on the periodic cycle update method, an alarm behavior set is generated to represent the linkage alarm relationship between sensor devices; the alarm behavior relationship is quantified, and the linkage probability of the linkage alarm relationship is analyzed; through an iterative method, discrete analysis is performed on the linkage probability to judge the response deviation effect of the sensor device under the behavior of the linkage alarm relationship, and an artificial intelligence perception output is made and sent to the backend; the present invention can protect the safety of the user's home environment through the alarm linkage behavior of the sensor without studying the user's home habit characteristics, can solve the problem of the user's personalized home habit differences, break the isolated island alarms of the sensors, and improve the adaptive intelligent level of the sensors.

Description

Technical Field

[0001] The present invention relates to the technical field of intelligent early warning, and specifically provides an adaptive perception early warning system and method based on artificial intelligence. Background Art

[0002] Intelligent sensor devices can perceive the user's home environment and provide monitoring and alarm feedback; with the popularization of intelligent sensor devices in the home environment, the quality of life of users has been improved.

[0003] However, due to differences in the manufacturers, models, and functions of various intelligent sensor devices, most of the current intelligent sensor devices on the market are still in an "isolated and closed" state and are difficult to form the ability to cooperate in alarms; at the same time, due to the large individual differences in users' home habits, it is even more difficult to achieve this cooperative alarm ability.

[0004] Artificial intelligence is the key to realizing intelligent management in smart homes. However, the applications of most current artificial intelligence products on the market are limited to understanding and analyzing users' instructions and behaviors through technologies such as deep learning, natural language processing, and computer vision to achieve the function of automatically adjusting the working state of devices. They still target solving the operating state problems of individual sensor devices and are still unable to solve the problem of cooperative alarms when facing multi-sensor scenarios and personalized home habits to ensure the safety of the user's home environment. Summary of the Invention

[0005] The purpose of the present invention is to provide an adaptive perception early warning system and method based on artificial intelligence to solve the problems raised in the above background art.

[0006] To solve the above technical problems, the present invention provides the following technical solutions:

[0007] An adaptive perception early warning system based on artificial intelligence, the system includes: a block cloud module, a data classification and recording module, and an artificial intelligence algorithm operation module;

[0008] The block cloud module compiles data files through the block clouds of different users and updates the block cloud of each user in a periodic cycle;

[0009] The data classification and recording module generates an alarm behavior set based on the periodic cycle update method and represents the linkage alarm relationship between sensor devices based on the alarm behavior;

[0010] The artificial intelligence algorithm operation module is used to quantify the alarm behavior relationship and analyze the linkage probability of the linkage alarm relationship; in an iterative manner, discrete analysis is performed on the linkage probability to judge the response deviation effect of the sensor device under the behavior of the linkage alarm relationship, and an artificial intelligence perception output is made and sent to the backend.

[0011] Furthermore, the block cloud module includes a user cloud unit and a cloud update unit;

[0012] The user cloud unit, based on user authorization, uniformly numbers the smart home sensors used by the user and constructs a smart home sensor database set, denoted as SD = {SD i |i ∈ [1, I]}, where SD i represents the i-th smart home sensor, and I is the total number of smart home sensors;

[0013] The cloud update unit, with a day as the cycle unit, divides the time range within a day into X time segments of equal scale, and denotes the t-th time segment as T t , and adds a cycle mark to the time segment T t , denoted as the cycle time segment T t (k), where k represents the cycle number, and k ∈ [1, Y], and Y represents the total number of cycles.

[0014] Furthermore, the data classification and recording module includes a cycle classification unit and a data recording unit;

[0015] The cycle classification unit, based on the cycle time segment T t (k), collects the alarm behavior set F[T t (k)], obtains the instruction upload time when each smart home sensor triggers an alarm feedback recorded in real time by the backend every day. If the instruction upload time of the smart home sensor SD i belongs to the cycle time segment T t (k), then the smart home sensor SD i is recorded in the alarm behavior set F[T t (k)];

[0016] The data recording unit is used to arbitrarily select two smart home sensors from the smart home sensor database set SD to form a linkage alarm relationship pair, denoted as SD i |SD j , where SD j represents the j-th smart home sensor, i ≠ j; when the smart home sensor SD i and the smart home sensor SD j both exist in the alarm behavior set F[T t(k)], it indicates within the periodic time segment T t (k), the linkage alarm relationship pair SD i |SD j holds. Otherwise, it indicates that within the periodic time segment T t (k), the linkage alarm relationship pair SD i |SD j does not hold.

[0017] Furthermore, the artificial intelligence algorithm operation module includes a linkage probability algorithm unit and a perception iterative analysis algorithm unit;

[0018] The linkage probability algorithm unit is used to construct a counting function CF{·}:

[0019] If SD i ∈F[T t (k)], then let CF{if: SD i ∈F[T t (k)]} = 1. If then let CF{if: SD i ∈F[T t (k)]} = 0;

[0020] If SD j ∈F[T t (k)], then let CF{if: SD j ∈F[T t (k)]} = 1. If then let CF{if: SD j ∈F[T t (k)]} = 0;

[0021] If the linkage alarm relationship pair SD i |SD j holds within the periodic time segment T t (k), then let CF{if: SD i ∈F[T t (k)], SD j ∈F[T t (k)]} = 1. If the linkage alarm relationship pair SD i |SD j does not hold within the periodic time segment T t (k), then let CF{if: SD i ∈F[T t (k)], SD j ∈F[T t (k)]} = 0;

[0022] Based on the alarm behavior set F[T t(k)], analyze and calculate the linkage alarm relationship pair SD i |SD j for the linkage probability value. The formula is: In the formula, P(SD i |SD j ) represents the linkage probability value of the linkage alarm relationship pair SD i |SD j ;

[0023] The described perception iterative analysis algorithm unit is used to collect each linkage probability value with the smart home sensor SD i as the relationship index reference object, and generate a linkage alarm sample cluster, denoted as RI(SD i ) = {P(SD i |SD j )|j ∈ [1, I], i ≠ j};

[0024] Construct an adaptive judgment perception iterative model:

[0025] Denote the input item of the (h + 1)-th iteration as ING h+1 , and ING h+1 ∈ RI(SD i ) - ∪ h∈[1，h+1) {ING h}, where ING h represents the input item of the h-th iteration;

[0026] At the (h + 1)-th iteration, if the linkage probability value P(SD i |SD r ) is selected as the input item, then ING h+1 = P(SD i |SD r ), where SD r represents the r-th smart home sensor, and i ≠ j ≠ r;

[0027] The judgment process of the (h + 1)-th iteration is as follows:

[0028] In the formula, LPD(ING h+1 ) represents the linkage probability dispersion of the input item ING h+1 , NUM[RI(SD i )] represents the total number of linkage probability values included in the linkage alarm sample cluster RI(SD i ), max[RI(SD i )] and min[RI(SD i )] respectively represent taking the maximum value and the minimum value in the linkage alarm sample cluster RI(SD i );

[0029] A preset discrete threshold. If the linkage probability dispersion LPD(ING h+1 ) is less than or equal to the discrete threshold, then the input item ING is extracted h+1 =P(SD i |SD r ) corresponding to the smart home sensor SD r ; otherwise, it is not extracted and the next iteration is entered;

[0030] When h + 1 > NUM[RI(SD i )], the iteration stops;

[0031] Collect and count the smart home sensors extracted in each iteration, and generate a judgment set of linkage alarm relationships denoted as RS(Y|SD i ), and send it to the backend.

[0032] An artificial intelligence-based adaptive perception and early warning method, which includes the following steps:

[0033] Step S1: Compile a data file through the block clouds of different users; update the block cloud of each user in a periodic loop manner;

[0034] Step S2: Generate an alarm behavior set based on the periodic loop update method; characterize the linkage alarm relationship between sensor devices based on the alarm behavior;

[0035] Step S3: Quantify the alarm behavior relationship and analyze the linkage probability of the linkage alarm relationship;

[0036] Step S4: Through iteration, perform discrete analysis on the linkage probability, judge the response deviation effect of the sensor device under the linkage alarm relationship behavior, and make an artificial intelligence perception output and send it to the backend.

[0037] Further, the specific implementation process of the step S1 includes:

[0038] Step S1.1: After obtaining user authorization, uniformly number the smart home sensors used by the user and construct a smart home sensor database set denoted as SD = {SD i |i ∈ [1, I]}, where SD i represents the i-th smart home sensor and I is the total number of smart home sensors;

[0039] Step S1.2: Taking a day as the periodic loop unit, divide the time range within a day into X time segments of equal scale, denote the t-th time segment as T t , and add a periodic mark to the time segment T t , denoted as the periodic time segment T t(k), where k represents the cycle number and k ∈ [1, Y], and Y represents the total number of cycles;

[0040] According to the above method, different from the existing periodic data induction methods, the home habits of users are involved in the present invention. Taking a day as the cycle unit, the time range within a day is divided into time segments, making the later data collection more targeted and more in line with the application conditions of sensors in the user's home scenario.

[0041] Furthermore, the specific implementation process of step S2 includes:

[0042] Step S2.1: Based on the periodic time segment T t (k), collect the alarm behavior set F[T t (k)], obtain the instruction upload time of each smart home sensor triggering an alarm feedback recorded in real time every day at the back end. If the instruction upload time of the smart home sensor SD i belongs to the periodic time segment T t (k), then record the smart home sensor SD i into the alarm behavior set F[T t (k)];

[0043] Step S2.2: Arbitrarily select two smart home sensors from the smart home sensor database set SD to form a linkage alarm relationship pair, denoted as SD i |SD j , where SD j represents the j-th smart home sensor, i ≠ j; when both the smart home sensor SD i and the smart home sensor SD j exist in the alarm behavior set F[T t (k)], it means that within the periodic time segment T t (t), the linkage alarm relationship pair SD i |SD j is established, otherwise it means that within the periodic time segment T t (k), the linkage alarm relationship pair SD i |SD j is not established;

[0044] According to the above method, the application of sensors in the user's home scene is also reflected in the collaborative alarm between different sensors. Specifically, the home environment has spatial characteristics and scene modal characteristics. The spatial characteristics are reflected in the kitchen, bathroom and bedroom, and the scene modal characteristics are reflected in the home environment scene within the kitchen space and the home environment scene within the bedroom space. Multi-space and multi-scene reflect the personalization of the user's home habits. In the home environment, the functions of sensors are different and the alarm tasks are also different, such as temperature and humidity sensors (which can accurately measure the ambient temperature and humidity, help the smart home system to monitor the environmental conditions in real time, and intelligently adjust home appliances such as air conditioners and humidifiers based on this, so that the home temperature and humidity are always maintained in the most comfortable state. They are mainly used in baby rooms, living rooms, bedrooms and other areas to provide a comfortable living environment), infrared sensors (using infrared technology to achieve Monitoring and tracking of targets, real-time monitoring of indoor and outdoor personnel activities, and intelligent control of the switch status of lights, curtains and other equipment based on activity information. It is mainly used in areas such as entrances and corridors to realize automatic control of lights turning on when people come and turning off when people leave), PM2.5 sensors (specially used to detect fine particulate matter in the indoor air of the home to ensure indoor air quality), as well as door magnetic sensors, gas concentration sensors, light sensors and water intrusion sensors, etc. In view of this, the use of sensor equipment in smart homes is mainly to realize the function of automatically adjusting the working status of the equipment. There is still a linkage island state between sensors, and it is difficult for alarms to coordinate; the present invention assumes that there is linkage between sensors within a micro-quantized time segment, and constitutes a linkage alarm relationship pair. The sensors can generate alarm tasks within the same micro-quantized time segment, which means that there is a linkage alarm relationship pair between them with a high probability.

[0045] Furthermore, the specific implementation process of step S3 includes:

[0046] Step S3.1: Construct counting function CF{·}:

[0047] If SD i ∈F[T t (k)], then let CF{if: SD i ∈F[T t (k)]}=1, if Then let CF{if:SD i ∈F[T t (k)]}=0;

[0048] If SD j ∈F[T t (k)], then let CF{if: SD j ∈F[T t (k)]}=1, if Then let CF{if:SDj ∈F[T t (k)]} = 0;

[0049] If the linkage alarm relationship pair SD i |SD j is established within the periodic time segment T t (k), then let CF{if: SD i ∈F[T t (k)], SD j ∈F[T t (k)]} = 1. If the linkage alarm relationship pair SD i |SD j is not established within the periodic time segment T t (k), then let CF{if: SD i ∈F[T t (k)], SD j ∈F[T t (k)]} = 0;

[0050] Step S3.2: Based on the alarm behavior set F[T t (k)], analyze and calculate the linkage probability value of the linkage alarm relationship pair SD i |SD j . The formula is: Where P(SD i |SD j ) represents the linkage probability value of the linkage alarm relationship pair SD i |SD j ;

[0051] According to the above method, the analysis of the linkage probability value is to verify the periodic stability of the linkage alarm relationship pair. The larger the linkage probability value, the higher the periodic stability of the linkage alarm relationship pair.

[0052] Furthermore, the specific implementation process of step S4 includes:

[0053] Step S4.1: Using the smart home sensor SD i as the relationship index reference object, collect each linkage probability value and generate a linkage alarm sample cluster, denoted as RI(SD i ) = {P(SD i |SD j )|j ∈ [1, I], i ≠ j};

[0054] Construct an adaptive judgment and perception iterative model:

[0055] Denote the input item of the (h + 1)-th iteration as ING h+1 , and ING h+1 ∈RI(SDi ) - ∪ h∈[1，h+1) {ING h}, where ING h represents the input item of the h-th iteration;

[0056] At the (h + 1)-th iteration, if the selected linkage probability value P(SD i |SD r ) is used as the input item, then ING h+1 = P(SD i |SD r ), where SD r represents the r-th smart home sensor, and i ≠ j ≠ r;

[0057] The judgment process of the (h + 1)-th iteration is as follows:

[0058] In the formula, LPD(ING h+1 ) represents the linkage probability dispersion of the input item ING h+1 , NUM[RI(SD i )] represents the total number of linkage probability values included in the linkage alarm sample cluster RI(SD i ), max[RI(SD i )] and min[RI(SD i )] respectively represent taking the maximum value and the minimum value in the linkage alarm sample cluster RI(SD i );

[0059] Preset a dispersion threshold. If the linkage probability dispersion LPD(ING h+1 ) is less than or equal to the dispersion threshold, then extract the smart home sensor SD h+1 = P(SD i |SD r ) corresponding to it; otherwise, do not extract and enter the next iteration; r ;

[0060] When h + 1 > NUM[RI(SD i )], the iteration stops;

[0061] According to the above method, after analyzing the linkage probability values, only the similarity of alarm behaviors between sensors can be quantitatively characterized. However, the performance of this similarity under the global linkage behavior is still unclear. Therefore, the present invention uses an adaptive judgment and perception iteration model to analyze the discreteness of the linkage probability, so as to reflect the response deviation effect of the sensor device under the linkage alarm relationship behavior. In particular, the above analysis takes a certain sensor as the reference object for the relationship index, and then outputs all the sensors that have a linkage alarm behavior with a certain sensor. Thus, each sensor can analyze all the sensors that have a linkage alarm behavior with it, making the accuracy of the intelligent level higher and the pertinence stronger.

[0062] Step S4.2: Collect and count the smart home sensors extracted during each iteration, and generate a judgment set of linkage alarm relationships denoted as RS(Y|SD i ), and send it to the backend.

[0063] An electronic device includes a memory, a processor, and a computer program stored on the memory and executable on the processor. By means of computer-executable instructions, when the processor executes the computer program, the above-mentioned artificial intelligence-based adaptive perception and early warning method is implemented.

[0064] Compared with the prior art, the beneficial effects achieved by the present invention are as follows: In the artificial intelligence-based adaptive perception and early warning system and method provided by the present invention, data files are compiled through the block clouds of different users; each user's block cloud is updated in a periodic cycle; based on the periodic cycle update method, an alarm behavior set is generated to characterize the linkage alarm relationship between sensor devices; the alarm behavior relationship is quantified, and the linkage probability of the linkage alarm relationship is analyzed; through iteration, the linkage probability is discretely analyzed, the response deviation effect of the sensor device under the linkage alarm relationship behavior is judged, and an artificial intelligence perception output is made and sent to the backend. The present invention can protect the safety of the user's home environment through the alarm linkage behavior of the sensor without studying the characteristics of the user's home habits. While solving the problem of the user's personalized home habit differences, it breaks the isolated alarms of the sensors and improves the adaptive intelligent level of the sensors. BRIEF DESCRIPTION OF THE DRAWINGS

[0065] The drawings are used to provide a further understanding of the present invention, and constitute a part of the specification. They are used together with the embodiments of the present invention to explain the present invention, and do not constitute a limitation to the present invention. In the drawings:

[0066] Figure 1 is a schematic structural diagram of an artificial intelligence-based adaptive perception and early warning system of the present invention;

[0067] Figure 2 It is a schematic diagram of the steps of an adaptive perception and early warning method based on artificial intelligence according to the present invention. Specific implementation manners

[0068] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0069] Please refer to Figure 1 , in the first embodiment: Provide an adaptive perception and early warning system based on artificial intelligence, and the system includes: a block cloud module, a data classification and recording module, and an artificial intelligence algorithm operation module;

[0070] The block cloud module prepares data files through the block clouds of different users; and updates the block cloud of each user in a periodic cycle manner;

[0071] Preferably, the block cloud module includes a user cloud unit and a cloud update unit;

[0072] The user cloud unit, based on user authorization, uniformly numbers the smart home sensors used by the user, and constructs a smart home sensor database set, denoted as SD = {SD i |i ∈ [1, I]}, where SD i represents the i-th smart home sensor, and I is the total number of smart home sensors;

[0073] The cloud update unit takes a day as a periodic cycle unit, divides the time range within a day into X time segments with equal scales, and denotes the t-th time segment as T t , and adds a cycle mark to the time segment T t , denoted as the periodic time segment T t (k), k represents the cycle number, and k ∈ [1, Y], where Y represents the total number of cycles;

[0074] The data classification and recording module generates an alarm behavior set based on the periodic cycle update method; and represents the linkage alarm relationship between sensor devices based on the alarm behavior;

[0075] Preferably, the data classification and recording module includes a cycle classification unit and a data recording unit;

[0076] The cycle classification unit collects the alarm behavior set F[T t (k) based on the periodic time segment T t(k)], obtain the instruction upload time of each smart home sensor triggering an alarm feedback recorded in real time by the backend every day. If the instruction upload time of the smart home sensor SD i belongs to the periodic time segment T t (k), then record the smart home sensor SD i in the alarm behavior set F[T t (k)];

[0077] Data recording unit, used to arbitrarily select two smart home sensors from the smart home sensor database set SD to form a linked alarm relationship pair, denoted as SD i |SD j , where SD j represents the j-th smart home sensor, i≠j; when the smart home sensor SD i and the smart home sensor SD j both exist in the alarm behavior set F[T t (k)], it means that within the periodic time segment T t (k), the linked alarm relationship pair SD i |SD j is established, otherwise it means that within the periodic time segment T t (k), the linked alarm relationship pair SD i |SD j is not established;

[0078] Artificial intelligence algorithm operation module, used to quantify the alarm behavior relationship and analyze the linkage probability of the linked alarm relationship; through an iterative method, discretely analyze the linkage probability, judge the response deviation effect of the sensor device under the linked alarm relationship behavior, and make an artificial intelligence perception output and send it to the backend;

[0079] Preferably, the artificial intelligence algorithm operation module includes a linkage probability algorithm unit and a perception iterative analysis algorithm unit;

[0080] Linkage probability algorithm unit, used to construct a counting function CF{·}:

[0081] If SD i ∈F[T t (k)], then let CF{if: SD i ∈F[T t (k)]} = 1, if then let CF{if: SD i ∈F[T t (k)]} = 0;

[0082] If SD j ∈F[T t (k)], then let CF{if: SDj ∈F[T t (k)]} = 1, if then let CF{if: SD j ∈F[T t (k)]} = 0;

[0083] If the linkage alarm relationship pair SD i |SD j is established within the periodic time segment T t (k), then let CF{if: SD i ∈F[T t (k)], SD j ∈[T t (k)]} = 1, if the linkage alarm relationship pair SD i |SD j is not established within the periodic time segment T t (k), then let CF{if: SD i ∈F[T t (k)], SD j ∈F[T t (k)]} = 0;

[0084] Based on the alarm behavior set F[T t (k)], analyze and calculate the linkage probability value of the linkage alarm relationship pair SD i |SD j . The formula is: In the formula, P(SD i |SD j ) represents the linkage probability value of the linkage alarm relationship pair SD i |SD j ;

[0085] The perception iteration analysis algorithm unit is used to collect each linkage probability value with the smart home sensor SD i as the relationship index reference object, and generate a linkage alarm sample cluster, denoted as RI(SD i ) = {P(SD i |SD j )|j ∈ [1, I], i ≠ j};

[0086] Construct an adaptive judgment and perception iteration model:

[0087] Denote the input item of the (h + 1)-th iteration as ING h+1 , and ING h+1 ∈ RI(SD i ) - ∪ h∈[1，h+1) {ING h}, where ING h represents the input item of the h-th iteration;

[0088] At the (h + 1)-th iteration, if the selected linkage probability value P(SD i |SD r ) is used as the input item, then ING h+1 = P(SD i |SD r ), where SD r represents the r-th smart home sensor, and i ≠ j ≠ r;

[0089] The judgment process of the (h + 1)-th iteration is as follows:

[0090] In the formula, LPD(ING h+1 ) represents the linkage probability dispersion degree of the input item ING h+1 , NUM[RI(SD i )] represents the total number of linkage probability values included in the linkage alarm sample cluster RI(SD i ), max[RI(SD i )] and min[RI(SD i )] respectively represent taking the maximum value and the minimum value in the linkage alarm sample cluster RI(SD i );

[0091] Preset a dispersion threshold. If the linkage probability dispersion degree LPD(ING h+1 ) is less than or equal to the dispersion threshold, then extract the smart home sensor SD h+1 corresponding to when the input item ING i = P(SD r |SD r ); otherwise, do not extract and enter the next iteration;

[0092] When h + 1 > NUM[RI(SD i )], the iteration stops;

[0093] Collect and count the smart home sensors extracted in each iteration, and generate a judgment set of linkage alarm relationships denoted as RS(Y|SD i ), and send it to the backend.

[0094] Please refer to Figure 2 , in the second embodiment: Provide an artificial intelligence-based adaptive perception and early warning method, and this method includes the following steps:

[0095] Step S1: Compile a data file through the block clouds of different users; update the block cloud of each user in a periodic cycle manner;

[0096] Exemplarily, after obtaining user authorization, uniformly number the smart home sensors used by the user, and construct a smart home sensor database set, denoted as SD = {SD i | i ∈ [1, I]}, where SD i represents the i-th smart home sensor, and I is the total number of smart home sensors;

[0097] Taking a day as a cycle unit, divide the time range within a day into X time segments of equal scale, and denote the t-th time segment as T t , and attach a cycle mark to the time segment T t , denoted as the periodic time segment T t (k), where k represents the cycle number, and k ∈ [1, Y], and Y represents the total number of cycles;

[0098] Step S2: Generate an alarm behavior set based on a cycle-based update method; based on the alarm behavior, characterize the linkage alarm relationship between sensor devices;

[0099] Exemplarily, based on the periodic time segment T t (k), collect the alarm behavior set F[T t (k)], obtain the instruction upload time when each smart home sensor triggers an alarm feedback recorded in real time every day at the backend. If the instruction upload time of the smart home sensor SD i belongs to the periodic time segment T t (k), then record the smart home sensor SD i into the alarm behavior set F[T t (k)];

[0100] Arbitrarily select two smart home sensors from the smart home sensor database set SD to form a linkage alarm relationship pair, denoted as SD i | SD j , where SD j represents the j-th smart home sensor, i ≠ j; when both the smart home sensor SD i and the smart home sensor SD j exist in the alarm behavior set F[T t (k)], it means that within the periodic time segment T t (k), the linkage alarm relationship pair SD i | SD j is established, otherwise it means that within the periodic time segment T t (k), the linkage alarm relationship pair SD i | SD j is not established;

[0101] Step S3: Quantify the alarm behavior relationship and analyze the linkage probability of the linkage alarm relationship;

[0102] Exemplarily, construct a counting function CF{·}:

[0103] If SD i ∈ F[T t (k)], then let CF{if: SD i ∈ F[T t (k)]} = 1. If then let CF{if: SD i ∈ F[T t (k)]} = 0;

[0104] If SD j ∈ F[T t (k)], then let CF{if: SD j ∈ F[T t (k)]} = 1. If then let CF{if: SD j ∈ F[T t (k)]} = 0;

[0105] If the linkage alarm relation pair SD i |SD j is established within the periodic time segment T t (k), then let CF{if: SD i ∈ F[T t (k)], SD j ∈ F[T t (k)]} = 1. If the linkage alarm relation pair SD i |SD j is not established within the periodic time segment T t (k), then let CF{if: SD i ∈ F[T t (k)], SD j ∈ F[T t (k)]} = 0;

[0106] Based on the alarm behavior set F[T t (k)], analyze and calculate the linkage probability value of the linkage alarm relation pair SD i |SD j . The formula is: In the formula, P(SD i |SD j ) represents the linkage probability value of the linkage alarm relation pair SD i |SD j ;

[0107] Step S4: Through an iterative approach, perform a discrete analysis of the linkage probability, judge the response deviation effect of the sensor device under the linkage alarm relationship behavior, and make an artificial intelligence perception output, which is sent to the backend;

[0108] Exemplarily, taking the smart home sensor SD i as the relationship index reference object, collect each linkage probability value, and generate a linkage alarm sample cluster, denoted as RI(SD i ) = {P(SD i |SD j )|j ∈ [1, I), i ≠ j};

[0109] Construct an adaptive judgment perception iterative model:

[0110] Denote the input item of the (h + 1)-th iteration as ING h+1 , and ING h+1 ∈ RI(SD i ) - ∪ h∈[1，h+1) {ING h}, where ING h represents the input item of the h-th iteration;

[0111] At the (h + 1)-th iteration, if the linkage probability value P(SD i |SD r ) is selected as the input item, then ING h+1 = P(SD i |SD r ), where SD r represents the r-th smart home sensor, and i ≠ j ≠ r;

[0112] The judgment process of the (h + 1)-th iteration is as follows:

[0113] In the formula, LPD(ING h+1 ) represents the linkage probability dispersion of the input item ING h+1 , NUM[RI(SD i )] represents the total number of linkage probability values included in the linkage alarm sample cluster RI(SD i ), max[RI(SD i )] and min[RI(SD i )] respectively represent taking the maximum value and the minimum value in the linkage alarm sample cluster RI(SD i );

[0114] Preset a discrete threshold. If the linkage probability dispersion LPD(ING h+1 ) is less than or equal to the discrete threshold, then extract the input item ING h+1 = P(SD i |SDr ) corresponding to the smart home sensor SD when r ; otherwise, do not extract and enter the next iteration;

[0115] When h + 1 > NUM[RI(SD i )], the iteration stops;

[0116] Collect and count the smart home sensors extracted in each iteration, and generate a linkage alarm relationship research and judgment set denoted as RS(Y|SD i ), and send it to the backend.

[0117] An electronic device includes a memory, a processor, and a computer program stored on the memory and executable on the processor. By means of computer-executable instructions, when the processor executes the computer program, the above-mentioned artificial intelligence-based adaptive perception and early warning method is implemented;

[0118] Exemplarily, the computer storage medium can be any combination of one or more computer-readable media. The computer-readable media can be a computer-readable signal medium or a computer-readable storage medium. The computer-readable storage medium can, for example, but is not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination of the above. More specific examples (non-exhaustive list) of the computer-readable storage medium include: an electrical connection having one or more wires, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above. In this document, the computer-readable storage medium can be any tangible medium that contains or stores a program, and this program can be used by or in combination with an instruction execution system, apparatus, or device.

[0119] The computer-readable signal medium can include a data signal propagated in a baseband or as part of a carrier wave, in which the computer-readable program code is carried. Such a propagated data signal can take various forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination of the above. The computer-readable signal medium can also be any computer-readable medium other than the computer-readable storage medium, and this computer-readable medium can send, propagate, or transmit a program for use by or in combination with an instruction execution system, apparatus, or device.

[0120] The program code contained on a computer-readable medium can be transmitted by any suitable medium, including but not limited to wireless, wire, optical fiber cable, etc., or any suitable combination of the above. The computer program code for performing the operations of the present invention can be written in one or more programming languages or combinations thereof. The programming languages include object-oriented programming languages such as Java, Smalltalk, C++, and also include conventional procedural programming languages such as the "C" language or similar programming languages. The program code can be executed entirely on the user's computer, partially on the user's computer, executed as a stand-alone software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In the case of a remote computer, the remote computer can be connected to the user's computer through any type of network, including a local area network (LAN) or a wide area network (WAN), or can be connected to an external computer (e.g., by using an Internet service provider to connect through the Internet).

[0121] It should be noted that, in this article, relational terms such as first and second are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any actual relationship or order between these entities or operations. Moreover, the term "comprising", "including" or any other variation thereof is intended to cover non-exclusive inclusion, so that a process, method, article or device comprising a series of elements not only includes those elements, but also includes other elements not expressly listed, or further includes elements inherent to such process, method, article or device.

[0122] Finally, it should be noted that the above are only preferred embodiments of the present invention and are not used to limit the present invention. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art can still modify the technical solutions described in the foregoing embodiments, or perform equivalent replacements for some of the technical features. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principle of the present invention shall be included within the protection scope of the present invention.

Claims

1. An adaptive perception warning method based on artificial intelligence, characterized in that: The method comprises the following steps: Step S1: compile data archives through the block clouds of different users; update the block cloud of each user through a periodic cycle; Step S2: Generate an alarm behavior set based on a periodic update method; characterize the linkage alarm relationship between sensor devices based on the alarm behavior; Step S3: quantify the alarm behavior relationship and analyze the linkage probability of the linkage alarm relationship; Step S4: Through an iterative method, discrete analysis is performed on the linkage probability to determine the response deviation effect of the sensor device under the linkage alarm relationship behavior, and artificial intelligence perception output is made and sent to the back end; The specific implementation process of step S1 includes: Step S1.1: After the user's authorization, the smart home sensors used by the user are uniformly numbered, and a smart home sensor database set is constructed, which is recorded as SD = {SD i |i∈[1,I]}, where SD i represents the i-th smart home sensor, and I is the total number of smart home sensors; Step S1.2: Taking the day as the cycle unit, divide the time range within a day into X time segments of equal size, and record the tth time segment as T t , and for the time segment T t Additional periodic mark, recorded as periodic time segment T t (k), k represents the cycle number, and k∈[1,Y], Y represents the total number of cycles; The specific implementation process of step S2 includes: Step S2.1: Based on the periodic time segment T t (k), collect the alarm behavior set F[T t (k)], obtain the upload time of each smart home sensor triggering alarm feedback recorded in real time by the backend every day. If the smart home sensor SD i The instruction upload time belongs to the periodic time segment T t (k) then the smart home sensor SD i Recorded in the alarm behavior set F[T t (k)]; Step S2.2: Randomly select two smart home sensors from the smart home sensor database set SD to form a linkage alarm relationship pair, denoted as SD i |SD j , where SD j represents the jth smart home sensor, i≠j; when the smart home sensor SD i and Smart Home Sensor SD j All exist in the alarm behavior set F[T t (k)], it means that in the periodic time segment T t (k) In the linkage alarm relationship, the SD i |SD j If true, it means that in the periodic time segment T t (k) In the linkage alarm relationship, the SD i |SD j Not true.

2. The adaptive perception warning method based on artificial intelligence according to claim 1 is characterized in that: The specific implementation process of step S3 includes: Step S3.1: Construct counting function CF{·}: If SD i ∈F[T t (k)], then let CF{if: SD i ∈F[T t (k)]}=1, if Then let CF{if:SD i ∈F[T t (k)]}=0; If SD j ∈F[T t (k)], then let CF{if: SD j ∈F[T t (k)]}=1, if Then let CF{if:SD j ∈F[T t (k)]}=0; If the linkage alarm relationship is SD i |SD j In the cycle time segment T t (k) holds, then let CF{if: SD i ∈F[T t (k)], SD j ∈F[T t (k)]}=1, if the linkage alarm relationship is SD i |SD j In the cycle time segment T t (k) is not true, then let CF{if: SD i ∈F[T t (k)], SD j ∈F[T t (k)]}=0; Step S3.2: Based on the alarm behavior set F[T t (k)], analyze and calculate the linkage alarm relationship to SD i |SD j The linkage probability value is: In the formula, P(SD i |SD j ) indicates the linkage alarm relationship to SD i |SD j The linkage probability value.

3. The adaptive perception warning method based on artificial intelligence according to claim 2 is characterized in that: The specific implementation process of step S4 includes: Step S4.1: Smart Home Sensor SD i The reference object of the relation index is collected, and each linkage probability value is generated, and a linkage alarm sample cluster is generated, which is recorded as RI (SD i )={P(SD i |SD j )|j∈[1,I],i≠j}; Constructing an adaptive judgment and perception iteration model: The input item of the h+1th iteration is recorded as ING h+1 , and ING h+1 ∈RI(SD i )-∪ h∈[1，h+1) {ING h }, among which, ING h represents the input item of the hth iteration; In the h+1th iteration, if the linkage probability value P(SD i |SD r ) as input, then ING h+1 =P(SD i |SD r ), where SD r represents the rth smart home sensor, and i≠j≠r; The judgment process of the h+1th iteration is as follows: Where, LPD(ING h+1 ) indicates the input item ING h+1 The linkage probability dispersion, NUM[RI(SD i )] represents the linkage alarm sample cluster RI (SD i ) contains the total number of linkage probability values, max[RI(SD i )] and min[RI(SD i )] respectively represent the linkage alarm sample cluster RI (SD i ) to take the maximum and minimum values; Preset discrete threshold, if linkage probability dispersion LPD(ING h+1 ) is less than or equal to the discrete threshold, then the input item ING is extracted h+1 =P(SD i |SD r ) corresponding smart home sensor SD r ; Otherwise, do not extract and enter the next iteration; When h+1>NUM[RI(SD i )], the iteration stops; Step S4.2: Collect and count the smart home sensors extracted in each iteration, and generate a linkage alarm relationship analysis set denoted as RS(Y|SD i ) and sent to the backend.

4. An adaptive perception warning system based on artificial intelligence, characterized in that: The system includes: a block cloud module, a data classification and recording module, and an artificial intelligence algorithm operation module; The block cloud module compiles data archives through the block clouds of different users; and updates the block cloud of each user in a periodic cycle; The data classification and recording module generates an alarm behavior set based on a periodic update method; based on the alarm behavior, it characterizes the linkage alarm relationship between sensor devices; The artificial intelligence algorithm operation module is used to quantify the alarm behavior relationship and analyze the linkage probability of the linkage alarm relationship; through an iterative method, the linkage probability is discretely analyzed to determine the response deviation effect of the sensor equipment under the linkage alarm relationship behavior, and make artificial intelligence perception output and send it to the back end; The block cloud module includes a user cloud unit and a cloud update unit; The user cloud unit, based on user authorization, uniformly numbers the smart home sensors used by the user and constructs a smart home sensor database set, denoted as SD = {SD i |i∈[1,I]}, where SD i represents the i-th smart home sensor, and I is the total number of smart home sensors; The cloud update unit takes a day as a cycle unit and divides the time range of a day into X time segments of equal size, and the tth time segment is recorded as T t , and for the time segment T t Additional periodic mark, recorded as periodic time segment T t (k), k represents the cycle number, and k∈[1,Y], Y represents the total number of cycles; The data classification and recording module includes a period classification unit and a data recording unit; The cycle classification unit is based on the cycle time segment T t (k), collect the alarm behavior set F[T t (k)], obtain the upload time of each smart home sensor triggering alarm feedback recorded in real time by the backend every day. If the smart home sensor SD i The instruction upload time belongs to the periodic time segment T t (k) then the smart home sensor SD i Recorded in the alarm behavior set F[T t (k)]; The data recording unit is used to select any two smart home sensors from the smart home sensor database set SD to form a linkage alarm relationship pair, which is recorded as SD i |SD j , where SD j represents the jth smart home sensor, i≠j; when the smart home sensor SD i and Smart Home Sensor SD j All exist in the alarm behavior set F[T t (k)], it means that in the periodic time segment T t (k) In the linkage alarm relationship, the SD i |SD j If true, it means that in the periodic time segment T t (k) In the linkage alarm relationship, the SD i |SD j Not true.

5. The adaptive perception warning system based on artificial intelligence according to claim 4 is characterized by: The artificial intelligence algorithm operation module includes a linkage probability algorithm unit and a perception iteration analysis algorithm unit; The linkage probability algorithm unit is used to construct a counting function CF{·}: If SD i ∈F[T t (k)], then let CF{if: SD i ∈F[T t (k)]}=1, if Then let CF{if:SD i ∈F[T t (k)]}=0: If SD j ∈F[T t (k)], then let CF{if: SD j ∈F[T t (k)]}=1, if Then let CF{if:SD j ∈F[T t (k)]}=0; If the linkage alarm relationship is SD i |SD j In the cycle time segment T t (k) holds, then let CF{if: SD i ∈F[T t (k)], SD j ∈F[T t (k)]}=1, if the linkage alarm relationship is SD i |SD j In the cycle time segment T t (k) is not true, then let CF{if: SD i ∈F[T t (k)], SD j ∈F[Tt(k)]}=0; Based on the alarm behavior set F[T t (k)], analyze and calculate the linkage alarm relationship to SD i |SD j The linkage probability value is: In the formula, P(SD i |SD j ) indicates the linkage alarm relationship to SD i |SD j The linkage probability value of The perception iteration analysis algorithm unit is used to generate a smart home sensor SD i The reference object of the relation index is collected, and each linkage probability value is generated, and a linkage alarm sample cluster is generated, which is recorded as RI (SD i )={P(SD i |SD j )|j∈[1,I],i≠j}; Constructing an adaptive judgment and perception iteration model: The input item of the h+1th iteration is recorded as ING h+1 , and ING h+1 ∈RI(SD i )-∪ h∈[1，h+1) {ING h }, among which, ING h represents the input item of the hth iteration; In the h+1th iteration, if the linkage probability value P(SD i |SD r ) as input, then ING h+1 =P(SD i |SD r ), where SD r represents the rth smart home sensor, and i≠j≠r; The judgment process of the h+1th iteration is as follows: Where, LPD(ING h+1 ) indicates the input item ING h+1 The linkage probability dispersion, NUM[RI(SD i )] represents the linkage alarm sample cluster RI (SD i ) contains the total number of linkage probability values, max[RI(SD i )] and min[RI(SD i )] respectively represent the linkage alarm sample cluster RI (SD i ) to take the maximum and minimum values; Preset discrete threshold, if linkage probability dispersion LPD(ING h+1 ) is less than or equal to the discrete threshold, then the input item ING is extracted h+1 =P(SD i |SD r ) corresponding smart home sensor SD r ; Otherwise, do not extract and enter the next iteration; When h+1>NUM[RI(SD i )], the iteration stops; Collect and count the smart home sensors extracted in each iteration, and generate a linkage alarm relationship analysis set denoted as RS(Y|SD i ) and sent to the backend.

6. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: By means of computer executable instructions, when the processor executes the computer program, an adaptive perception warning method based on artificial intelligence as described in any one of claims 1 to 3 is implemented.

Images