Home environment remote monitoring and early warning system and method based on artificial intelligence

Abstract

The present invention discloses a home environment remote monitoring and early warning system and method based on artificial intelligence, relates to the technical field of home environment monitoring and early warning, solves the technical problems that it is difficult to give early warning to ordinary home data in the monitoring and early warning of ordinary home data, and that in the monitoring and early warning of special home data, the personalized early warning threshold used is obtained based on a large amount of historical data of the family, and the reasonable setting of the early warning threshold for the family lacking or less historical data is ignored; the present invention obtains the judgment factor of the next time period based on the characteristic value of the ordinary home data and the environmental data of the next time period, and issues early warning information of the ordinary home data based on the judgment factor; obtains the personalized index of the current family based on reference data, and extracts the personalized threshold from the regional historical data based on the personalized index; and issues early warning information of the special home data based on the personalized threshold; the present invention can improve the comfort of family living and the accuracy of early warning.

Description

Technical Field

[0001] The present invention belongs to the field of smart home and relates to home environment monitoring and early warning technology, specifically to a home environment remote monitoring and early warning system and method based on artificial intelligence. Background Art

[0002] The comfort and safety of the home environment directly affects the health and quality of life of residents, especially for families with elderly, children and pets. With the continuous advancement of technology, the smart home market is developing rapidly. More and more families are beginning to adopt smart home devices to improve the quality of life and convenience. Smart home systems can not only realize remote control of home appliances, but also monitor and manage the home environment in real time through sensors and intelligent algorithms.

[0003] The prior art (invention patent application with publication number CN118644952A) discloses a remote monitoring and early warning system for home environment based on artificial intelligence. Its technical solution is: by receiving real-time sensor monitoring data and real-time monitoring video of the target residence, on the one hand, the current sensor monitoring data is sorted out and combined with the water monitoring unit, electricity monitoring unit, gas monitoring unit and smoke monitoring unit to carry out the home environment safety assessment of the target residence, on the other hand, the most recent door sensing data is extracted to perceive the living status of the residents in the current target residence, and the home behavior safety assessment unit or the away behavior safety assessment unit is used to identify the corresponding abnormal activities of the residents of the current target residence, and the alarm window of the home environment safety risk hidden dangers and abnormal activities of the residents of the current target residence is generated and pushed to the application terminal.

[0004] When the above scheme remotely monitors and warns the home environment, it ignores the problem of delayed effect of temperature and humidity regulation in the adjustment of the home environment, and ignores the early warning and early adjustment of ordinary home data such as temperature and humidity, reducing the health and safety protection of people who are easily affected by temperature and humidity changes; at the same time, the above scheme sets the warning threshold based on a large amount of historical data of the family, ignores the reasonable setting of the warning threshold for families with lack of or less historical data, and cannot quickly set personalized thresholds for families that first access the system, reducing the practicality of the system.

[0005] Therefore, the present invention discloses a home environment remote monitoring and early warning system and method based on artificial intelligence to solve the above technical problems. Summary of the invention

[0006] The present invention aims to solve at least one of the technical problems existing in the prior art; to this end, the present invention proposes a remote monitoring and early warning system and method for a home environment based on artificial intelligence, which is used to solve the technical problems that it is difficult to provide early warning for ordinary home data in the monitoring and early warning of ordinary home data, and that the personalized early warning threshold used in the monitoring and early warning of special home data is obtained based on a large amount of historical data of the family, ignoring the reasonable setting of the early warning threshold for families with lacking or less historical data. The present invention obtains characteristic values ​​of ordinary home data based on an artificial intelligence model, obtains judgment factors for the next time period based on the characteristic values ​​and environmental data of the next time period, issues early warning information for the ordinary home data based on the judgment factors and controls smart home devices; obtains reference data of the current family, obtains personalized indicators of the current family based on the reference data, extracts personalized thresholds from regional historical data based on the personalized indicators; and issues early warning information for special home data based on the personalized thresholds to solve the above problems.

[0007] To achieve the above-mentioned object, the first aspect of the present invention provides a home environment remote monitoring and early warning system based on artificial intelligence, comprising: a general early warning module, and a data collection module, a special early warning module and a database connected thereto;

[0008] The data collection module is used to plan a day into several time periods and obtain basic data and environmental data of the current time period; wherein the basic data includes ordinary home data and special home data, the ordinary home data includes the internal temperature and internal humidity of the home, the special home data includes the water consumption, gas consumption and electricity consumption, and the environmental data includes the external temperature and external humidity outside the home;

[0009] The general warning module is used to obtain the characteristic value of the general household data based on the artificial intelligence model, obtain the judgment factor of the next time period based on the characteristic value and the environmental data of the next time period, issue the warning information of the general household data based on the judgment factor and control the smart home equipment; wherein the smart home equipment includes a smart air conditioner and a smart constant humidity machine, and the judgment factor includes a temperature judgment factor WP and a humidity judgment factor SP;

[0010] The special warning module is used to extract the existing special household data of the current household from the database and mark the special household data as reference data; obtain the personalized indicators of the current household based on the reference data, and extract the personalized thresholds from the regional historical data based on the personalized indicators; and issue warning information of the special household data based on the personalized thresholds; wherein the personalized indicators include water consumption indicators, gas consumption indicators and electricity consumption indicators.

[0011] Preferably, the planning of a day into several time periods includes:

[0012] Extract the number of warnings YC issued by the current family the day before, and determine whether the number of warnings YC exceeds the number threshold YZ; if yes, obtain the dynamic duration DS based on the formula DS=BS×δ1×(2-YC / YZ); if no, obtain the dynamic duration DS based on the formula DS=BS×δ2×YZ / YC; where BS is the length of the manually set standard time period, and δ1 and δ2 are manually set proportional adjustment coefficients, both of which have a value range of (0,2);

[0013] Based on the dynamic duration, DS divides a day into several time periods starting from 0:00.

[0014] It should be noted that if the current family is connected to the system for the first time today, and the system was not running in the current family the day before, there is no warning information. At this time, the manually set standard time period is used to plan today into several time periods; similarly, if the current family has been connected to the system before and disconnected before today, the data of the last full day of the previous connection will be used as the data of the previous day for analysis.

[0015] Preferably, the obtaining of basic data and environmental data for the current time period includes:

[0016] After authorization by the user, the real-time temperature inside the home is obtained through the temperature sensor, and the real-time temperature is marked as the internal temperature; the real-time humidity inside the home is obtained through the humidity sensor, and the real-time humidity is marked as the internal humidity;

[0017] The water consumption of the household in the current time period is obtained through the smart water meter, the gas consumption of the household in the current time period is obtained through the smart gas meter, and the electricity consumption of the household in the current time period is obtained through the smart electricity meter; the regional real-time temperature and regional real-time humidity of the area where the household is located are obtained through a computer, and the regional real-time temperature is marked as the external temperature, and the regional real-time humidity is marked as the external humidity.

[0018] Preferably, the method of obtaining the characteristic values ​​of common household data based on the artificial intelligence model includes:

[0019] The internal temperature obtained in the current time period is divided into group A1, and the internal humidity is divided into group A2; the data of group A1 and group A2 in the regional historical data and the characteristic values ​​corresponding to the data of group A1 and group A2 are obtained through the database;

[0020] Integrate the A1 group data and the corresponding characteristic values ​​into a number of training data and test data, use the training data to train the artificial intelligence model; use the test data to test the trained artificial intelligence model, and adjust the artificial intelligence model according to the test results; finally obtain a temperature characteristic model whose input is the A1 group data and output is the internal temperature characteristic value; wherein the artificial intelligence model includes a BP neural network model or an RBF neural network model;

[0021] The A2 group data and the corresponding characteristic values ​​are integrated into a number of training data and test data, and the artificial intelligence model is trained using the training data; the trained artificial intelligence model is tested using the test data, and the artificial intelligence model is adjusted according to the test results; finally, a humidity characteristic model is obtained with the input of the A2 group data and the output of the internal humidity characteristic value;

[0022] The A1 group data of the current time period is input into the temperature characteristic model to obtain the internal temperature characteristic value of the current time period; the A2 group data of the current time period is input into the humidity characteristic model to obtain the internal humidity characteristic value of the current time period.

[0023] It is worth noting that the present invention uses a temperature characteristic model and a humidity characteristic model when obtaining the characteristic values ​​of ordinary household data. The collected temperature data and humidity data are input into the corresponding models to obtain characteristic values. The use of artificial intelligence models can quickly extract characteristic values ​​from complex data, thereby improving the operating efficiency of the system. At the same time, the output characteristic values ​​provide data support for subsequent analysis to obtain the judgment factors for the next time period.

[0024] Preferably, obtaining the judgment factor for the next time period based on the characteristic value and the environmental data for the next time period includes:

[0025] Extract the internal temperature characteristic value NWZ, the internal humidity characteristic value NSZ, the external temperature WW and the external humidity WS of the current time period, and the external temperature AWW and the external humidity AWS of the next time period;

[0026] Based on the formula WP=NWZ+NWZ×α1×ln((AWW-WW) / WW+1), the temperature judgment factor WP for the next time period is obtained; based on the formula SP=NSZ+NSZ×α2×ln((AWS-WS) / WS+1), the humidity judgment factor SP for the next time period is obtained; wherein, the judgment factors include the temperature judgment factor WP and the humidity judgment factor SP; α1 and α2 are the amplitude adjustment factors of the ln() function, and the value range is (0,2).

[0027] It is worth noting that the present invention adds an analysis of the impact of environmental data on the temperature and humidity inside the home when obtaining the judgment factor for the next time period, and obtains a judgment factor that can indicate the temperature and humidity of the next stage through the influence of environmental data. Through this judgment factor, the temperature and humidity of the next stage can be dynamically adjusted in advance according to the current data, avoiding the problem of delayed effect of temperature and humidity adjustment, and providing health and safety protection for people who are easily affected by temperature and humidity changes.

[0028] Preferably, the method of issuing warning information of common home data based on the judgment factor and controlling the smart home devices includes:

[0029] C1: Mark the time of the next time period as the target time, extract the temperature judgment factor WP, and determine whether the temperature judgment factor WP is greater than the maximum value of the standard temperature range; if yes, issue an over-temperature warning message for the target time, record the number of warnings, and turn on the smart air conditioner at the beginning of the next time period to cool the room to the optimal temperature; if no, jump to C2; the optimal temperature is the middle value of the standard temperature range, and the standard temperature range is obtained by manual setting;

[0030] C2: Determine whether the temperature judgment factor WP is less than the minimum value of the standard temperature range; if yes, issue a warning message of low temperature at the target time, record the number of warnings, and turn on the smart air conditioner at the beginning of the next time period to raise the indoor temperature to the optimal temperature; if no, do nothing;

[0031] C3: Extract humidity judgment factor SP, and judge whether the humidity judgment factor SP is greater than the maximum value of the standard humidity range; if yes, issue a warning message of high humidity at the target time, record the number of warnings, and start the intelligent constant humidity machine to dehumidify the room to the optimum humidity at the beginning of the next time period; if no, jump to C4; wherein the optimum humidity is the middle value of the standard humidity range, and the standard humidity range is obtained by manual setting;

[0032] C4: Determine whether the humidity judgment factor SP is less than the minimum value of the standard humidity range; if yes, issue a low humidity warning message at the target time, record the number of warnings, and turn on the intelligent constant humidity machine at the beginning of the next time period to humidify the room to the optimal humidity; if no, do nothing.

[0033] Preferably, the obtaining of the personalized index of the current family based on the reference data includes:

[0034] Extract reference data, and classify water consumption in the reference data into B1, gas consumption into B2, and electricity consumption into B3;

[0035] Obtain the variance of each type of data in turn, and determine whether the variance is less than the corresponding defined variance; if yes, retain the data of this type; if no, remove the data in this type that differs the most from the mode of this type, and re-determine the variance until the variance of this type is less than the corresponding defined threshold, and retain the remaining data of this type; wherein the defined threshold is obtained through experience;

[0036] The average values ​​of the various types of retained data are obtained, and the average value of the B1 category is marked as the water consumption index Z1, the average value of the B2 category is marked as the gas consumption index Z2, and the average value of the B3 category is marked as the electricity consumption index Z3; wherein the personalized indicators include the water consumption index, the gas consumption index, and the electricity consumption index.

[0037] It is worth noting that when the present invention obtains the water consumption index, gas consumption index and electricity consumption index, it obtains them by performing variance judgment on various types of data. The variance judgment can remove data with large differences in various types of data, and then the average value of each type of data is extracted to obtain indicators representing the characteristics of each type of data, which provides data support for the subsequent extraction of personalized thresholds.

[0038] Preferably, the step of extracting a personalized threshold from regional historical data based on a personalized indicator includes:

[0039] Extract the water consumption index, gas consumption index and electricity consumption index of each household in the region using the system for n time periods; mark the mode of the water consumption index as the regional water consumption index QZ1, mark the mode of the gas consumption index as the regional gas consumption index QZ2, and mark the mode of the electricity consumption index as the regional electricity consumption index QZ3;

[0040] Based on the formula HZi=βi×DZi×Zi / QZi, the personalized threshold HZi of the personalized indicator corresponding to the current household is obtained; wherein, i represents a number, and when it is 1, it represents the parameter data of the water consumption indicator; when it is 2, it represents the parameter data of the gas consumption indicator; when it is 3, it represents the parameter data of the electricity consumption indicator; DZi represents a fixed threshold; βi is an artificially set proportional adjustment coefficient, and the value range of βi is (0,2).

[0041] It is worth noting that the present invention first obtains the water consumption index, gas consumption index and electricity consumption index of the current family using the system for n time periods, and then obtains the regional water consumption index, regional gas consumption index and regional electricity consumption index of each family in the region using the system for n time periods, and obtains the personalized threshold value that can indicate the current family's water, gas and electricity consumption habits through comparative analysis between the corresponding indicators; the present invention does not require a large amount of historical data of the current family to obtain the personalized threshold value, but only needs a small amount of data in several time periods. This not only reduces the amount of calculation of the system, but also can quickly set personalized threshold values ​​for families that first access the system, avoiding the problem of analyzing a large amount of historical data resulting in the inability to set personalized threshold values ​​for the using family for a long time, thereby improving the practicality of the system.

[0042] It is worth noting that the present invention can dynamically adjust the threshold according to the living habits and family preferences of family members, so that the early warning system is closer to actual needs, can reduce false alarms and missed alarms, and improve the reliability of the system and user trust.

[0043] Preferably, the issuing of warning information of special household data based on the personalized threshold includes:

[0044] Extract the current household’s water, gas and electricity consumption in the current time period;

[0045] Determine whether the water consumption exceeds the personalized threshold value HZ1; if yes, issue a warning message of excessive water consumption and record the number of warnings; if no, do nothing;

[0046] Determine whether the gas usage exceeds the personalized threshold HZ2; if yes, issue a warning message indicating that the gas usage exceeds the limit, and record the number of warnings; if no, do nothing;

[0047] Determine whether the water consumption exceeds the personalized threshold HZ3; if yes, issue an alarm message indicating that the power consumption exceeds the limit and record the number of alarms; if no, do nothing.

[0048] A second aspect of the present invention provides a home environment remote monitoring and early warning method based on artificial intelligence, comprising the following steps:

[0049] S1: Plan a day into several time periods and obtain basic data and environmental data for the current time period;

[0050] S2: Obtaining characteristic values ​​of common household data based on the artificial intelligence model, obtaining judgment factors for the next time period based on the characteristic values ​​and environmental data for the next time period, issuing warning information of common household data based on the judgment factors and controlling smart home devices;

[0051] S3: extracting existing special household data of the current household from the database, and marking the special household data as reference data; obtaining personalized indicators of the current household based on the reference data, and extracting personalized thresholds from regional historical data based on the personalized indicators; and issuing early warning information of the special household data based on the personalized thresholds.

[0052] Compared with the prior art, the present invention has the following beneficial effects:

[0053] 1. The present invention obtains characteristic values ​​of ordinary household data based on an artificial intelligence model, obtains judgment factors of the next time period based on the characteristic values ​​and environmental data of the next time period, issues early warning information of ordinary household data based on the judgment factors and controls smart home devices; obtains reference data of the current household, obtains personalized indicators of the current household based on the reference data, and extracts personalized thresholds from regional historical data based on the personalized indicators; issues early warning information of special household data based on personalized thresholds, thereby solving the technical problems that it is difficult to give early warnings to ordinary household data in the monitoring and early warning of ordinary household data, and that the personalized early warning thresholds used in the monitoring and early warning of special household data are obtained based on a large amount of historical data of the household, ignoring the reasonable setting of the early warning thresholds for households with less or no historical data; the present invention can improve the safety and comfort of family living while improving the accuracy and reliability of the early warning system.

[0054] 2. When acquiring the environmental factors for the next time period, the present invention introduces the analysis of the impact of environmental data on the temperature and humidity inside the home, and extracts the judgment factor that can predict the temperature and humidity changes in the next stage. The judgment factor is used to pre-regulate the temperature and humidity in the next stage, avoiding the effect delay problem existing in the traditional temperature and humidity adjustment method. This intelligent control method provides a healthier and safer protection for those who are easily affected by temperature and humidity changes. The system can more accurately predict and adjust the indoor environment, thereby effectively avoiding the discomfort caused to family members due to environmental changes.

[0055] 3. The present invention adopts an efficient method to obtain personalized thresholds for household water, gas and electricity usage habits. First, the system collects the water consumption, gas consumption and electricity consumption data of the current household during n time periods of using the system. Then, the corresponding indicator data of other households in the same area in the same time period are integrated. By comparing and analyzing these indicators, the system can extract personalized thresholds that reflect the unique usage habits of the current household. When determining personalized thresholds, the present invention does not need to rely on a large amount of historical data, but can be completed with brief data of a few time periods. This approach significantly reduces the computational burden of the system and enables new users to quickly set personalized thresholds when they first access the system, thereby avoiding the need for massive historical data in traditional methods, solving the problem of long-term personalized threshold setting, and greatly improving the practical application value and convenience of the system. BRIEF DESCRIPTION OF THE DRAWINGS

[0056] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0057] Figure 1 It is a schematic diagram of the operation steps of the present invention;

[0058] Figure 2 It is a schematic diagram of the system module of the present invention;

[0059] Figure 3 A schematic diagram of the operation steps of obtaining the judgment factor for the next time period according to the present invention;

[0060] Figure 4 A schematic diagram of the operation steps for extracting a personalized threshold value according to the present invention. DETAILED DESCRIPTION

[0061] The technical solution of the present invention will be clearly and completely described below in conjunction with the embodiments. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0062] See also Figure 1-2 , the first aspect of the present invention provides a home environment remote monitoring and early warning system based on artificial intelligence, including: a general early warning module, and a data collection module, a special early warning module and a database connected thereto;

[0063] Data collection module: used to plan a day into several time periods and obtain basic data and environmental data of the current time period; wherein the basic data includes ordinary home data and special home data, ordinary home data includes internal temperature and internal humidity inside the home, special home data includes water consumption, gas consumption and electricity consumption, and environmental data includes external temperature and external humidity outside the home;

[0064] General warning module: used to obtain the characteristic value of general household data based on the artificial intelligence model, obtain the judgment factor of the next time period based on the characteristic value and the environmental data of the next time period, issue warning information of general household data based on the judgment factor and control the smart home equipment; wherein, the smart home equipment includes smart air conditioners and smart constant humidity machines, and the judgment factors include temperature judgment factor WP and humidity judgment factor SP;

[0065] Special warning module: used to extract the existing special household data of the current household from the database and mark the special household data as reference data; obtain the personalized indicators of the current household based on the reference data, and extract personalized thresholds from the regional historical data based on the personalized indicators; issue warning information of special household data based on the personalized thresholds; among which, the personalized indicators include water consumption indicators, gas consumption indicators and electricity consumption indicators.

[0066] It should be noted that the regional historical data refers to the historical data of each household in the area where the current household is located.

[0067] It should be noted that the regions are obtained through manual settings, such as a community is a region, and a village is a region.

[0068] It should be noted that the existing special home data of the current family is extracted from the database and marked as reference data. If the current family accesses the system for the first time today and there is no special home data, a personalized threshold is set after the current family uses the system for n time periods. During this period, a fixed threshold is used instead of the personalized threshold.

[0069] For example, in this embodiment:

[0070] The number of warnings issued by the current family the day before is YC=6, which does not exceed the number threshold YZ=10. Based on the formula DS=BS×δ2×YZ / YC=2×0.9×10 / 6=3min, the dynamic duration DS=3min is obtained; wherein, in this embodiment, the standard time period BS is set to 2min, and the proportional adjustment coefficient δ2 is set to 0.9;

[0071] Based on the dynamic duration DS=3min, a day is divided into several time periods starting from 0:00;

[0072] After authorization by the user, the real-time temperature inside the home is obtained through the temperature sensor, and the real-time temperature is marked as the internal temperature; the real-time humidity inside the home is obtained through the humidity sensor, and the real-time humidity is marked as the internal humidity;

[0073] The water consumption of the household in the current time period is obtained through the smart water meter, the gas consumption of the household in the current time period is obtained through the smart gas meter, and the electricity consumption of the household in the current time period is obtained through the smart electricity meter; the regional real-time temperature and regional real-time humidity of the area where the household is located are obtained through the computer, and the regional real-time temperature is marked as the external temperature, and the regional real-time humidity is marked as the external humidity;

[0074] The internal temperature obtained in the current time period is divided into group A1, and the internal humidity is divided into group A2; the data of group A1 and group A2 in the regional historical data and the characteristic values ​​corresponding to the data of group A1 and group A2 are obtained through the database;

[0075] Integrate the A1 group data and the corresponding characteristic values ​​into a number of training data and test data, use the training data to train the artificial intelligence model; use the test data to test the trained artificial intelligence model, and adjust the artificial intelligence model according to the test results; finally obtain a temperature characteristic model whose input is the A1 group data and output is the internal temperature characteristic value; wherein the artificial intelligence model includes a BP neural network model or an RBF neural network model;

[0076] The A2 group data and the corresponding characteristic values ​​are integrated into a number of training data and test data, and the artificial intelligence model is trained using the training data; the trained artificial intelligence model is tested using the test data, and the artificial intelligence model is adjusted according to the test results; finally, a humidity characteristic model is obtained with the input of the A2 group data and the output of the internal humidity characteristic value;

[0077] Input the A1 group data of the current time period into the temperature characteristic model to obtain the internal temperature characteristic value of the current time period; input the A2 group data of the current time period into the humidity characteristic model to obtain the internal humidity characteristic value of the current time period;

[0078] Extract the internal temperature characteristic value NWZ=26℃, the internal humidity characteristic value NSZ=40%, the external temperature WW=30℃, the external humidity WS=60% in the current time period, and the external temperature AWW=31℃, the external humidity AWS=62% in the next time period;

[0079] Based on the formula WP=NWZ+NWZ×α1×ln((AWW-WW) / WW+1)=26+26×0.8×ln((31-30) / 30+1)=26.7°C, the temperature judgment factor WP=26.7°C is obtained for the next time period; based on the formula SP=NSZ+NSZ×α2×ln((AWS-WS) / WS+1)=40%+40%×0.95×ln((62%-60%) / 60%+1)=41.2%, the humidity judgment factor SP=41.2% is obtained for the next time period; wherein, in this embodiment, the amplitude adjustment factor α1 is taken as 0.8, and the amplitude adjustment factor α2 is taken as 0.95;

[0080] Since the temperature judgment factor WP=26.7℃ is within the standard temperature range, no operation is performed;

[0081] Since the humidity judgment factor SP=41.2% is within the standard humidity range, no operation is performed;

[0082] Extracting existing special household data of the current household from the database, and marking the special household data as reference data;

[0083] Extract reference data, and classify water consumption in the reference data into B1, gas consumption into B2, and electricity consumption into B3;

[0084] Obtain the variance of each type of data in turn, and determine whether the variance is less than the corresponding defined variance; if yes, retain the data of this type; if no, remove the data in this type that differs the most from the mode of this type, and re-judge the variance until the variance of this type is less than the corresponding defined threshold, and retain the remaining data of this type; the defined threshold is obtained through experience;

[0085] Obtain the average values ​​of the various types of retained data, mark the average value of category B1 as the water consumption index Z1, mark the average value of category B2 as the gas consumption index Z2, and mark the average value of category B3 as the electricity consumption index Z3; wherein the personalized index includes the water consumption index, the gas consumption index, and the electricity consumption index;

[0086] Extract the water consumption index, gas consumption index and electricity consumption index of each household in the region using the system for 200 time periods; mark the mode of the water consumption index as the regional water consumption index QZ1, mark the mode of the gas consumption index as the regional gas consumption index QZ2, and mark the mode of the electricity consumption index as the regional electricity consumption index QZ3;

[0087] Based on the formula HZ1=β1×DZ1×Z1 / QZ1, the personalized threshold HZ1 of the current household water consumption index is obtained; based on the formula HZ2=β2×DZ2×Z2 / QZ2, the personalized threshold HZ2 of the current household gas consumption index is obtained; based on the formula HZ3=β3×DZ3×Z3 / QZ3, the personalized threshold HZ3 of the current household electricity consumption index is obtained; where DZi represents a fixed threshold; β1, β2 and β3 are manually set proportional adjustment coefficients, and the value range of β1, β2 and β3 is (0,2);

[0088] Extract the current household’s water, gas and electricity consumption in the current time period;

[0089] Determine whether the water consumption exceeds the personalized threshold HZ1; if yes, issue an alarm message indicating that the water consumption exceeds the limit and record the number of alarms; if no, do nothing;

[0090] Determine whether the gas usage exceeds the personalized threshold HZ2; if yes, issue an alarm message indicating that the gas usage exceeds the limit and record the number of alarms; if no, do nothing;

[0091] Determine whether the water consumption exceeds the personalized threshold HZ3; if yes, issue an alarm message indicating that the power consumption exceeds the limit and record the number of alarms; if no, do nothing.

[0092] In this application, a day is divided into several time periods, including:

[0093] Extract the number of warnings YC issued by the current family the day before, and determine whether the number of warnings YC exceeds the number threshold YZ; if yes, obtain the dynamic duration DS based on the formula DS=BS×δ1×(2-YC / YZ); if no, obtain the dynamic duration DS based on the formula DS=BS×δ2×YZ / YC; where BS is the length of the manually set standard time period, and δ1 and δ2 are manually set proportional adjustment coefficients, both of which have a value range of (0,2);

[0094] Based on the dynamic duration, DS divides a day into several time periods starting from 0:00.

[0095] It should be noted that if the current family is connected to the system for the first time today, and the system was not running in the current family the day before, there is no warning information. At this time, the manually set standard time period is used to plan today into several time periods; similarly, if the current family has been connected to the system before and disconnected before today, the data of the last full day of the previous connection will be used as the data of the previous day for analysis.

[0096] It should be noted that the planning of several time periods each day starts at the current 0:00.

[0097] It should be noted that, in the dynamic duration DS obtained based on the formula DS=BS×δ1×(2-YC / YZ), the formula DS=BS×δ1×(2-YC / YZ) is obtained by changing DS=BS×δ1×(1-(YC-YZ) / YZ).

[0098] It should be noted that when the value of YC / YZ is not less than 2 or the value of the dynamic duration DS is less than the duration protection value, the dynamic duration DS is set to the duration protection value; wherein, the duration protection value is used to prevent the value of the dynamic duration DS from being too small, resulting in an unreasonable time period.

[0099] It should be noted that, when a day is divided into several time periods starting from 0:00 based on the dynamic time period DS, if there is excess time less than the dynamic time period DS after the last complete time period, the excess time is regarded as one time period.

[0100] It should be noted that a complete time period is a complete dynamic duration DS.

[0101] This application obtains basic data and environmental data for the current time period, including:

[0102] After user authorization, the real-time temperature inside the home is obtained through the temperature sensor and marked as the internal temperature; the real-time humidity inside the home is obtained through the humidity sensor and marked as the internal humidity;

[0103] The water consumption of the household in the current time period is obtained through the smart water meter, the gas consumption of the household in the current time period is obtained through the smart gas meter, and the electricity consumption of the household in the current time period is obtained through the smart electricity meter; the regional real-time temperature and regional real-time humidity of the area where the household is located are obtained through the computer, and the regional real-time temperature is marked as the external temperature, and the regional real-time humidity is marked as the external humidity.

[0104] It should be noted that the current time period is the time period in which the current time is located.

[0105] In this application, the characteristic values ​​of common household data are obtained based on the artificial intelligence model, including:

[0106] The internal temperature obtained in the current time period is divided into group A1, and the internal humidity is divided into group A2; the data of group A1 and group A2 in the regional historical data and the characteristic values ​​corresponding to the data of group A1 and group A2 are obtained through the database;

[0107] Integrate the A1 group data and the corresponding characteristic values ​​into a number of training data and test data, use the training data to train the artificial intelligence model; use the test data to test the trained artificial intelligence model, and adjust the artificial intelligence model according to the test results; finally obtain a temperature characteristic model whose input is the A1 group data and output is the internal temperature characteristic value; wherein the artificial intelligence model includes a BP neural network model or an RBF neural network model;

[0108] The A2 group data and the corresponding characteristic values ​​are integrated into a number of training data and test data, and the artificial intelligence model is trained using the training data; the trained artificial intelligence model is tested using the test data, and the artificial intelligence model is adjusted according to the test results; finally, a humidity characteristic model is obtained with the input of the A2 group data and the output of the internal humidity characteristic value;

[0109] The A1 group data of the current time period is input into the temperature characteristic model to obtain the internal temperature characteristic value of the current time period; the A2 group data of the current time period is input into the humidity characteristic model to obtain the internal humidity characteristic value of the current time period.

[0110] Specifically, the trained artificial intelligence model is tested using the test data, and the specific steps for adjusting the artificial intelligence model according to the test results are as follows:

[0111] For example, group A1: input the A1 group data in the test data into the trained artificial intelligence model to obtain the corresponding internal temperature characteristic value, and compare the corresponding internal temperature characteristic value with the internal temperature characteristic value in the test data. When the difference between the two is within the threshold, the threshold is obtained based on experience, and there is no need to adjust the parameters, and proceed to the next group of test data; if it is not within the threshold, adjust the corresponding parameters until the corresponding difference is within the threshold before proceeding to the next group of test data. When the number of test data whose output internal temperature characteristic values ​​of all test data have a difference within the threshold accounts for 90% or more of the total test data, a temperature characteristic model with input as group A1 data and output as internal temperature characteristic value is obtained.

[0112] It should be noted that the internal temperature characteristic value used in the test data is obtained from the maximum value, minimum value and mode of the group of data. The specific calculation formula is:

[0113] For example, group A1: extract the maximum value ZD, minimum value ZX and mode ZS of group A1 in the test data, and obtain the internal temperature characteristic value NWZ corresponding to the current group A1 based on the formula NWZ=(BL1×ZD+BL2×ZX+ZS) / 2; wherein BL1 and BL2 are both proportional adjustment coefficients greater than 0, and BL1+BL2=1, BL1>BL2.

[0114] See also Figure 3In this application, the judgment factor for the next time period is obtained based on the characteristic value and the environmental data of the next time period, including:

[0115] Extract the internal temperature characteristic value NWZ, the internal humidity characteristic value NSZ, the external temperature WW and the external humidity WS of the current time period, and the external temperature AWW and the external humidity AWS of the next time period;

[0116] Based on the formula WP=NWZ+NWZ×α1×ln((AWW-WW) / WW+1), the temperature judgment factor WP for the next time period is obtained; based on the formula SP=NSZ+NSZ×α2×ln((AWS-WS) / WS+1), the humidity judgment factor SP for the next time period is obtained; wherein, the judgment factors include the temperature judgment factor WP and the humidity judgment factor SP; α1 and α2 are the amplitude adjustment factors of the ln() function, and the value range is (0,2).

[0117] It should be noted that the temperature judgment factor can be understood as the characteristic temperature of the next stage obtained through analysis; the humidity judgment factor can be understood as the characteristic humidity of the next stage obtained through analysis.

[0118] In this application, warning information of common home data is issued based on the judgment factor and smart home devices are controlled, including:

[0119] C1: Mark the time of the next time period as the target time, extract the temperature judgment factor WP, and determine whether the temperature judgment factor WP is greater than the maximum value of the standard temperature range; if yes, issue an over-temperature warning message for the target time, record the number of warnings, and turn on the smart air conditioner at the beginning of the next time period to cool the room to the optimal temperature; if no, jump to C2; the optimal temperature is the middle value of the standard temperature range, and the standard temperature range is obtained by manual setting;

[0120] C2: Determine whether the temperature judgment factor WP is less than the minimum value of the standard temperature range; if yes, issue a warning message of low temperature at the target time, record the number of warnings, and turn on the smart air conditioner at the beginning of the next time period to raise the indoor temperature to the optimal temperature; if no, do nothing;

[0121] C3: Extract humidity judgment factor SP, and judge whether humidity judgment factor SP is greater than the maximum value of standard humidity range; if yes, issue warning information of high humidity at target time, record the number of warnings, and start intelligent constant humidity machine to dehumidify the room to the optimum humidity at the beginning of next time period; if no, jump to C4; the optimum humidity is the middle value of standard humidity range, and standard humidity range is obtained by manual setting;

[0122] C4: Determine whether the humidity judgment factor SP is less than the minimum value of the standard humidity range; if yes, issue a low humidity warning message at the target time, record the number of warnings, and turn on the intelligent constant humidity machine at the beginning of the next time period to humidify the room to the optimal humidity; if no, do nothing.

[0123] In this application, personalized indicators of the current family are obtained based on reference data, including:

[0124] Extract reference data, and classify water consumption in the reference data into B1, gas consumption into B2, and electricity consumption into B3;

[0125] Obtain the variance of each type of data in turn, and determine whether the variance is less than the corresponding defined variance; if yes, retain the data of this type; if no, remove the data in this type that differs the most from the mode of this type, and re-judge the variance until the variance of this type is less than the corresponding defined threshold, and retain the remaining data of this type; the defined threshold is obtained through experience;

[0126] The average values ​​of the various types of retained data are obtained, and the average value of the B1 category is marked as the water consumption index Z1, the average value of the B2 category is marked as the gas consumption index Z2, and the average value of the B3 category is marked as the electricity consumption index Z3; wherein the personalized indicators include the water consumption index, the gas consumption index, and the electricity consumption index.

[0127] See also Figure 4 In this application, personalized thresholds are extracted from regional historical data based on personalized indicators, including:

[0128] Extract the water consumption index, gas consumption index and electricity consumption index of each household in the region using this system for n time periods; mark the mode of the water consumption index as the regional water consumption index QZ1, mark the mode of the gas consumption index as the regional gas consumption index QZ2, and mark the mode of the electricity consumption index as the regional electricity consumption index QZ3;

[0129] Based on the formula HZi=βi×DZi×Zi / QZi, the personalized threshold HZi of the personalized indicator corresponding to the current household is obtained; wherein, i represents a number, and when it is 1, it represents the parameter data of the water consumption indicator; when it is 2, it represents the parameter data of the gas consumption indicator; when it is 3, it represents the parameter data of the electricity consumption indicator; DZi represents a fixed threshold; βi is an artificially set proportional adjustment coefficient, and the value range of βi is (0,2).

[0130] It should be noted that the specific steps of obtaining the personalized threshold HZi of the personalized indicator corresponding to the current family based on the formula HZi=βi×DZi×Zi / QZi are as follows:

[0131] Based on the formula HZ1=β1×DZ1×Z1 / QZ1, the personalized threshold value HZ1 of the current household water consumption index is obtained;

[0132] Based on the formula HZ2=β2×DZ2×Z2 / QZ2, the personalized threshold value HZ2 of the current household gas usage indicator is obtained;

[0133] Based on the formula HZ3=β3×DZ3×Z3 / QZ3, the personalized threshold HZ3 of the current household electricity consumption indicator is obtained.

[0134] It should be noted that in this application, i represents a number. When i is 1, it represents the parameter data of the water consumption index; when i is 2, it represents the parameter data of the gas consumption index; when i is 3, it represents the parameter data of the electricity consumption index. The specific explanation is:

[0135] When i is 1, Z1 represents the water consumption index, QZ1 represents the regional water consumption index, DZ1 represents the fixed threshold of water consumption, HZ1 represents the personalized threshold of water consumption index, and β1 represents the proportional adjustment coefficient of water consumption index;

[0136] When i is 2, Z2 represents the gas usage index, QZ2 represents the regional gas usage index, DZ2 represents the fixed threshold of gas usage, HZ2 represents the personalized threshold of gas usage index, and β2 represents the proportional adjustment coefficient of gas usage index;

[0137] When i is 3, Z3 represents the electricity consumption index, QZ3 represents the regional electricity consumption index, DZ3 represents the fixed threshold of electricity consumption, HZ3 represents the personalized threshold of electricity consumption index, and β3 represents the proportional adjustment coefficient of electricity consumption index.

[0138] In this application, warning information of special household data is issued based on personalized thresholds, including:

[0139] Extract the current household’s water, gas and electricity consumption in the current time period;

[0140] Determine whether the water consumption exceeds the personalized threshold HZ1; if yes, issue an alarm message indicating that the water consumption exceeds the limit and record the number of alarms; if no, do nothing;

[0141] Determine whether the gas usage exceeds the personalized threshold HZ2; if yes, issue an alarm message indicating that the gas usage exceeds the limit and record the number of alarms; if no, do nothing;

[0142] Determine whether the water consumption exceeds the personalized threshold HZ3; if yes, issue an alarm message indicating that the power consumption exceeds the limit and record the number of alarms; if no, do nothing.

[0143] The second aspect of the present invention provides a method for remote monitoring and early warning of a home environment based on artificial intelligence, comprising the following steps:

[0144] S1: Plan a day into several time periods and obtain basic data and environmental data for the current time period;

[0145] S2: Obtaining characteristic values ​​of common household data based on the artificial intelligence model, obtaining judgment factors for the next time period based on the characteristic values ​​and environmental data for the next time period, issuing warning information of common household data based on the judgment factors and controlling smart home devices;

[0146] S3: extracting existing special household data of the current household from the database, and marking the special household data as reference data; obtaining personalized indicators of the current household based on the reference data, and extracting personalized thresholds from regional historical data based on the personalized indicators; and issuing early warning information of the special household data based on the personalized thresholds.

[0147] Part of the data in the above formula is calculated by removing the dimension and taking its numerical value. The formula is a formula closest to the actual situation obtained by software simulation of a large amount of collected data; the preset parameters and preset thresholds in the formula are set by technical personnel in this field according to actual conditions or obtained through simulation of a large amount of data.

[0148] Working principle of the present invention:

[0149] Plan a day into several time periods, and obtain the basic data and environmental data of the current time period; obtain the characteristic values ​​of ordinary household data based on the artificial intelligence model. In this step, the artificial intelligence model can quickly extract characteristic values ​​from complex data, which improves the operating efficiency of the system. At the same time, the output characteristic values ​​provide data support for the subsequent analysis to obtain the judgment factor of the next time period; obtain the judgment factor of the next time period based on the characteristic value and the environmental data of the next time period. In this step, the judgment factor that can indicate the temperature and humidity of the next stage is obtained. Through this judgment factor, the temperature and humidity of the next stage can be dynamically adjusted in advance according to the current data, avoiding the problem of delayed effect of temperature and humidity adjustment; Based on the judgment factors, early warning information of ordinary home data is issued and the smart home devices are controlled; the special home data that already exists in the current family is extracted from the database, and the special home data is marked as reference data; the personalized indicators of the current family are obtained based on the reference data. This step can obtain indicators representing the characteristics of various types of data, which provides data support for the subsequent extraction of personalized thresholds; personalized thresholds are extracted from regional historical data based on personalized indicators. This step enables the system to reduce the amount of system calculation when obtaining personalized thresholds, and can avoid the problem of analyzing a large amount of historical data and causing the user family to be unable to set personalized thresholds for a long time; early warning information of special home data is issued based on the personalized threshold.

[0150] The above embodiments are only used to illustrate the technical method of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical method of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical method of the present invention.

Claims

1. A remote monitoring and early warning system for home environment based on artificial intelligence, characterized in that: include: General warning module, and the data collection module, special warning module and database connected thereto; The data collection module is used to plan a day into several time periods and obtain basic data and environmental data of the current time period; wherein the basic data includes ordinary home data and special home data, the ordinary home data includes the internal temperature and internal humidity of the home, the special home data includes the water consumption, gas consumption and electricity consumption, and the environmental data includes the external temperature and external humidity outside the home; The general warning module is used to obtain the characteristic value of the general household data based on the artificial intelligence model, obtain the judgment factor of the next time period based on the characteristic value and the environmental data of the next time period, issue the warning information of the general household data based on the judgment factor and control the smart home equipment; wherein the smart home equipment includes a smart air conditioner and a smart constant humidity machine, and the judgment factor includes a temperature judgment factor WP and a humidity judgment factor SP; The special warning module is used to extract the existing special household data of the current household from the database, mark the special household data as reference data; obtain the personalized index of the current household based on the reference data, extract the personalized threshold from the regional historical data based on the personalized index; issue the warning information of the special household data based on the personalized threshold; wherein the personalized index includes the water consumption index, the gas consumption index and the electricity consumption index; The characteristic values ​​of common household data obtained based on the artificial intelligence model include: The internal temperature obtained in the current time period is divided into group A1, and the internal humidity is divided into group A2; the data of group A1 and group A2 in the regional historical data and the characteristic values ​​corresponding to the data of group A1 and group A2 are obtained through the database; Integrate the A1 group data and the corresponding characteristic values ​​into a number of training data and test data, use the training data to train the artificial intelligence model; use the test data to test the trained artificial intelligence model, and adjust the artificial intelligence model according to the test results; finally obtain a temperature characteristic model whose input is the A1 group data and output is the internal temperature characteristic value; wherein the artificial intelligence model includes a BP neural network model or an RBF neural network model; The A2 group data and the corresponding characteristic values ​​are integrated into a number of training data and test data, and the artificial intelligence model is trained using the training data; the trained artificial intelligence model is tested using the test data, and the artificial intelligence model is adjusted according to the test results; finally, a humidity characteristic model is obtained with the input of the A2 group data and the output of the internal humidity characteristic value; Input the A1 group data of the current time period into the temperature characteristic model to obtain the internal temperature characteristic value of the current time period; input the A2 group data of the current time period into the humidity characteristic model to obtain the internal humidity characteristic value of the current time period; The step of obtaining the judgment factor for the next time period based on the characteristic value and the environmental data for the next time period includes: Extract the internal temperature characteristic value NWZ, the internal humidity characteristic value NSZ, the external temperature WW and the external humidity WS of the current time period, and the external temperature AWW and the external humidity AWS of the next time period; Based on the formula WP=NWZ+NWZ×α1×ln((AWW-WW) / WW+1), the temperature judgment factor WP for the next time period is obtained; based on the formula SP=NSZ+NSZ×α2×ln((AWS-WS) / WS+1), the humidity judgment factor SP for the next time period is obtained; wherein, the judgment factors include the temperature judgment factor WP and the humidity judgment factor SP; α1 and α2 are the amplitude adjustment factors of the ln() function, and the value range is (0,2).

2. The artificial intelligence-based home environment remote monitoring and early warning system according to claim 1 is characterized in that: The day is planned into several time periods, including: Extract the number of warnings YC issued by the current family the day before, and determine whether the number of warnings YC exceeds the number threshold YZ; if yes, obtain the dynamic duration DS based on the formula DS=BS×δ1×(2-YC / YZ); if no, obtain the dynamic duration DS based on the formula DS=BS×δ2×YZ / YC; where BS is the standard time period, δ1 and δ2 are proportional adjustment coefficients, both in the range of (0,2); Based on the dynamic duration, DS divides a day into several time periods starting from 0:

00.

3. The artificial intelligence-based home environment remote monitoring and early warning system according to claim 1 is characterized in that: The obtaining of basic data and environmental data for the current time period includes: After authorization by the user, the real-time temperature inside the home is obtained through the temperature sensor, and the real-time temperature is marked as the internal temperature; the real-time humidity inside the home is obtained through the humidity sensor, and the real-time humidity is marked as the internal humidity; The water consumption of the household in the current time period is obtained through the smart water meter, the gas consumption of the household in the current time period is obtained through the smart gas meter, and the electricity consumption of the household in the current time period is obtained through the smart electricity meter; the regional real-time temperature and regional real-time humidity of the area where the household is located are obtained through a computer, and the regional real-time temperature is marked as the external temperature, and the regional real-time humidity is marked as the external humidity.

4. The artificial intelligence-based home environment remote monitoring and early warning system according to claim 1 is characterized in that: The method of issuing warning information of common home data based on the judgment factor and controlling the smart home devices includes: C1: Mark the time of the next time period as the target time, extract the temperature judgment factor WP, and determine whether the temperature judgment factor WP is greater than the maximum value of the standard temperature range; if yes, issue an over-temperature warning message for the target time, record the number of warnings, and turn on the smart air conditioner at the beginning of the next time period to cool the room to the optimal temperature; if no, jump to C2; the optimal temperature is the middle value of the standard temperature range; C2: Determine whether the temperature judgment factor WP is less than the minimum value of the standard temperature range; if yes, issue a warning message of low temperature at the target time, record the number of warnings, and turn on the smart air conditioner at the beginning of the next time period to raise the indoor temperature to the optimal temperature; if no, do nothing; C3: Extract humidity judgment factor SP, and judge whether the humidity judgment factor SP is greater than the maximum value of the standard humidity range; if yes, issue a warning message of high humidity at the target time, record the number of warnings, and start the intelligent constant humidity machine to dehumidify the room to the optimum humidity at the beginning of the next time period; if no, jump to C4; wherein the optimum humidity is the middle value of the standard humidity range; C4: Determine whether the humidity judgment factor SP is less than the minimum value of the standard humidity range; if yes, issue a low humidity warning message at the target time, record the number of warnings, and turn on the intelligent constant humidity machine at the beginning of the next time period to humidify the room to the optimal humidity; if no, do nothing.

5. The artificial intelligence-based home environment remote monitoring and early warning system according to claim 1 is characterized in that: The obtaining of the personalized index of the current family based on the reference data includes: Extract reference data, and classify water consumption in the reference data into B1, gas consumption into B2, and electricity consumption into B3; Obtain the variance of each type of data in turn, and determine whether the variance is less than the corresponding defined variance; if yes, retain the data of this type; if no, remove the data in this type that differs the most from the mode of this type, and re-determine the variance until the variance of this type is less than the corresponding defined threshold, and retain the remaining data of this type; The average values ​​of the various types of retained data are obtained, and the average value of the B1 category is marked as the water consumption index Z1, the average value of the B2 category is marked as the gas consumption index Z2, and the average value of the B3 category is marked as the electricity consumption index Z3; wherein the personalized indicators include the water consumption index, the gas consumption index, and the electricity consumption index.

6. The artificial intelligence-based home environment remote monitoring and early warning system according to claim 5 is characterized in that: The step of extracting a personalized threshold from the regional historical data based on the personalized index includes: Extract the water consumption index, gas consumption index and electricity consumption index of each household in the region using the system for n time periods; mark the mode of the water consumption index as the regional water consumption index QZ1, mark the mode of the gas consumption index as the regional gas consumption index QZ2, and mark the mode of the electricity consumption index as the regional electricity consumption index QZ3; Based on the formula HZi=βi×DZi×Zi / QZi, the personalized threshold HZi of the personalized indicator corresponding to the current household is obtained; wherein, i represents a number, and when it is 1, it represents the parameter data of the water consumption indicator; when it is 2, it represents the parameter data of the gas consumption indicator; when it is 3, it represents the parameter data of the electricity consumption indicator; DZi represents a fixed threshold; βi is a proportional adjustment coefficient, and the value range of βi is (0,2).

7. The artificial intelligence-based home environment remote monitoring and early warning system according to claim 6 is characterized in that: The issuing of early warning information of special household data based on personalized thresholds includes: Extract the current household’s water, gas and electricity consumption in the current time period; Determine whether the water consumption exceeds the personalized threshold value HZ1; if yes, issue a warning message of excessive water consumption and record the number of warnings; if no, do nothing; Determine whether the gas usage exceeds the personalized threshold HZ2; if yes, issue a warning message indicating that the gas usage exceeds the limit, and record the number of warnings; if no, do nothing; Determine whether the water consumption exceeds the personalized threshold value HZ3; if yes, issue an alarm message indicating that the power consumption exceeds the standard and record the number of alarms; if no, do nothing.

8. A method for remote monitoring and early warning of a home environment based on artificial intelligence, based on the operation of a remote monitoring and early warning system for a home environment based on artificial intelligence according to any one of claims 1 to 7, characterized in that: S1: Plan a day into several time periods and obtain basic data and environmental data for the current time period; S2: Obtaining characteristic values ​​of common household data based on the artificial intelligence model, obtaining judgment factors for the next time period based on the characteristic values ​​and environmental data for the next time period, issuing warning information of common household data based on the judgment factors and controlling smart home devices; S3: extracting existing special household data of the current household from the database, and marking the special household data as reference data; Obtain personalized indicators for the current family based on reference data, and extract personalized thresholds from regional historical data based on the personalized indicators; Issue early warning information on special home data based on personalized thresholds.

Images