Smart home control panel and recommendation method thereof
By using time decay algorithms and calendar type adjustments in the smart home control panel, combined with real-time environmental data, the problem of fixed and inaccurate recommendation results in the existing technology is solved, and more accurate and diverse equipment and scenario recommendations are achieved.
Patent Information
- Application Number
- CN202310623564.X
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2023-05-26
- Publication Date
- 2024-11-26
AI Technical Summary
The recommendation method of existing smart home control panels fails to finely divide time and area, resulting in fixed and inaccurate recommendation results, which cannot meet users' intelligent and diversified needs.
By obtaining the operation records of home equipment and scenarios in each time period within the most recent set date period of the current time, using the time attenuation algorithm to calculate the preliminary recommendation results, and adjust it based on working days and rest days, and further adjust it in combination with real-time environmental data. To provide more accurate recommendations.
Comprehensive recommendations based on time periods, regions and calendar types are realized, which improves the accuracy and diversity of recommendations, ensures the real-time and accuracy of recommendation results, and meets the actual needs of users.
Smart Images

Figure CN119024731A_ABST
Abstract
Description
Technical Field
[0001] The invention relates to a smart home control panel and a recommendation method thereof, belonging to the technical field of smart homes. Background Art
[0002] With the advent of the smart era, smart homes have developed rapidly. The control panel recommendation method for smart homes mainly uses big data algorithms to calculate the number of home device usages per day and makes recommendations based on the calculation results. The existing smart home control panel recommendation method mainly uses the time decay algorithm for calculation. The calculation data used is the number of times the user has used each device and scene every day in the past few days. However, this algorithm is more suitable for small-volume recommendations. The calculation data used is one day's data, and the division accuracy of time data is not fine enough, resulting in the same content recommended at every moment of the day in the recommendation results obtained by the algorithm. Recommendations are not divided into regions, and the current recommendations do not distinguish whether they are recommendations for weekdays or weekends, resulting in the recommended devices or scenes being relatively fixed and inaccurate, and unable to meet the user's intelligent and diversified recommendation needs. Summary of the invention
[0003] The purpose of the present invention is to provide a smart home control panel and a recommendation method thereof, so as to solve the problem that the existing control panel does not divide the area and time into fine divisions to recommend devices and scenes, resulting in fixed and inaccurate recommendations.
[0004] To achieve the above object, the solution of the present invention includes:
[0005] A method for recommending a smart home control panel of the present invention comprises the following steps:
[0006] 1) According to the host, user, device and scene area, obtain the operation records of home devices and scenes in each time period within the set date period closest to the current time, form device data and scene data respectively, and use the time decay algorithm to calculate each time period based on the device data and scene data to obtain the preliminary recommendation results of home devices and / or scenes in each area in each time period;
[0007] 2) The preliminary recommendation results are adjusted according to weekdays and weekends, and the weekend type devices or scenes in the weekday recommendation results are deleted, and the weekday type devices or scenes in the weekend recommendation results are deleted to obtain the adjusted recommendation results.
[0008] Beneficial effect: The recommendation method of the smart home control panel of the present invention calculates the preliminary recommendation results of home appliances and / or scenes in each area in each time period according to the operation records of home appliances and scenes in each time period within the set date period closest to the current time, and considers working days and weekends to classify the recommendation results, and adjusts the preliminary recommendation results according to working days and weekends to obtain the adjusted recommendation results. This method considers each time period, each area, working days and weekends, and makes comprehensive recommendations. The recommendation results are more accurate and meet actual needs. The recommendations in each time period are not exactly the same, which ensures the real-time and accuracy of the recommendations.
[0009] Furthermore, the weekday type devices and scenes are the devices and scenes for which the number of operations in the first set time period on weekdays is greater than the set number of operations, and there is no operation in this time period on weekends; the holiday type devices and scenes are the devices and scenes for which the number of operations in the second set time period on weekends is greater than the set number of operations, and there is no operation in this time period on weekdays.
[0010] Beneficial effect: When the number of times a device or scene is operated during a certain time period on a weekday exceeds the set number of operations, and there is no operation during this time period on a weekend, it means that the device or scene is used on a weekday. Similarly, devices or scenes of the weekend type are obtained. By classifying the devices or scenes in each time period, the accuracy of the control panel recommendation can be improved.
[0011] Furthermore, the method also includes the step of adjusting the recommendation result obtained in step 2) according to the real-time environmental data.
[0012] Furthermore, the adjustment step according to the real-time environmental data is: comparing the obtained real-time environmental data with the environment caused by the device and scene operations in the recommendation results. If the two are not compatible, adjusting the corresponding devices and scenes in the recommendation results, or adjusting the recommendation order accordingly according to the set environmental data.
[0013] Beneficial effects: The method of the present invention further includes adjusting the device scene according to the real-time environment data, and adjusting the devices and scenes in the recommended results according to the real-time environment data and the set environment data. The method takes environmental factors into consideration, and the adjusted recommendation results are more accurate, more in line with the actual needs of users, and improves the accuracy of the recommendation.
[0014] Furthermore, the device data includes the acquisition date, the time period of the acquisition date, the host and the corresponding host ID, the device ID, the device name, the area ID and the number of times the device is used, and the scene data includes the acquisition date, the time period of the acquisition date, the host and the corresponding host ID, the scene ID, the scene name, the area ID and the number of times the scene is used.
[0015] Furthermore, the device recommendation results include the host ID, recommended device ID, recommended value corresponding to the recommended device and the corresponding recommended time period, and the scene recommendation results include the host ID, user area ID, recommended scene ID, recommended value corresponding to the recommended scene and the recommended time period.
[0016] Beneficial effects: The data displayed in the device recommendation results and the scene recommendation results can accurately determine the recommended device or scene, prioritize them according to the recommended value, and recommend the best scene or device.
[0017] Furthermore, the time period within the set date closest to the current time is each hour in the 7 days closest to the current time.
[0018] Furthermore, the time decay algorithm model is:
[0019]
[0020] G is the time attenuation coefficient, x1-x T is the number of device clicks at a certain time within 1-T days, X is the cumulative operation days vector of the device at that time, a is the cumulative days impact factor, and Y n is the recommendation index vector for devices and scenarios at point n.
[0021] A smart home control panel of the present invention comprises a processor and a display, wherein the processor is connected to the display, the processor is used to execute program instructions to implement the recommendation method of the smart home control panel as described in any one of the above items, and the display is used to display the recommendation results processed by the processor.
[0022] Beneficial effects: The smart home control panel of the present invention includes a processor and a display. The processor can execute program instructions to better implement the recommendation method of the smart home control panel. The display displays the recommendation results processed by the processor. The control panel considers weekdays and weekends for comprehensive recommendations. The recommendation results are more accurate and meet actual needs. The recommendations for different time periods are not exactly the same, ensuring the diversity and accuracy of the recommendations. BRIEF DESCRIPTION OF THE DRAWINGS
[0023] Figure 1 It is a schematic diagram of the control panel recommendation process flow in the method embodiment of the present invention. DETAILED DESCRIPTION
[0024] The present invention will be further described in detail below in conjunction with the accompanying drawings.
[0025] Control method embodiment of smart home control panel:
[0026] like Figure 1The control method of a smart home control panel shown in the figure first obtains the operation records of home appliances and scenes in each time period within the set date period closest to the current time according to the host, user, device and scene area, and forms device data and scene data respectively. According to the device data and scene data, a time decay algorithm is used to calculate each time period, and preliminary recommendation results of home appliances and / or scenes in each area in each time period are obtained respectively. The statistical results are sorted into a recommendation result table respectively, and the preliminary recommendation results are adjusted according to working days and rest days, and the devices and scenes are recommended respectively according to the priority order of the adjusted recommendation result table. The specific implementation method is as follows:
[0027] 1. According to the host, user, device and scene area, obtain the operation records of home devices and scenes in each time period within the set date period closest to the current time, form device data and scene data respectively, and use the time decay algorithm to calculate each time period based on the device data and scene data to obtain preliminary recommendation results for home devices and / or scenes in each area in each time period.
[0028] Specifically, the time period within the set time closest to the current time can be set according to the actual situation, which can be every hour within 3 days or every hour within 9 days. The set time period obtained in this embodiment is the data of each day in the past 7 days closest to the current time, and is counted by hour and region. For example, the recommendation on January 8th needs to obtain the operation record data of household devices and scenes in each area every day and every hour from January 1st to January 7th, and the recommendation on January 9th needs the operation record data of household devices and scenes in each area every day and every hour from January 2nd to January 8th. The obtained household device operation records and operation time and other information form the device data table, and the obtained scene operation records, operation time and other information form the scene data table.
[0029] The sample data of the device data sheet is shown in Table 1:
[0030] Table 1
[0031]
[0032] In Table 1, create_data is the statistical date, time is the specific time period of the statistical date, sn is the host name, reg_sn_id is the host ID number, dev_id is the device ID, dev_name is the device name, area_id is the area ID, and dev_use_num is the number of times the device is used. The meaning shown in Table 1 is: the user with host ID 66 operated the large living room downlight with device ID 65 located in area 1 once from 6 to 7 o'clock on December 25, 2021.
[0033] The scene data table includes the statistical date (create_data), the specific time period of the statistical date (Time), the host name (sn), the host ID number (reg_sn_id), the scene ID (dev_id), the scene name (dev_name), the area ID (area_id), and the scene usage times (dev_use_num). The formats of the host and the host ID are different, but they represent the same host, such as: the host is 06263c6400924c34b37fdac9351afc2b, and the host ID is 66. The sample data of the scene data table is shown in Table 2:
[0034] Table 2
[0035]
[0036] The meaning shown in the table is: the user with host ID 453 executed the daytime home scene with scene ID 2 located in area 1 once from 9:00 to 10:00 on November 23, 2021.
[0037] This embodiment adopts an improved time decay algorithm, and the time decay algorithm model is:
[0038]
[0039] Where G is the time attenuation coefficient, x1, x2, x3…x T Y is the device click count vector from day 1 to day T at n points, X is the cumulative operation days vector of the device at point n, and a is the cumulative days impact factor, which is the weight coefficient for regulating the proportion of X. The time attenuation coefficient G is 3, and the cumulative days impact factor a is 5. n is the recommendation index vector for devices and scenarios at point n.
[0040] The time decay algorithm of this embodiment takes into account three main influencing factors: the distance from the historical date to the recommendation prediction day (current time or next day), the click frequency of the device, and the number of days clicked. The closer the operation is to the prediction day, the greater the impact. The more days the device is operated, the easier it is to be recommended. Devices with more recent click frequencies are recommended first than those with fewer clicks. For example, in the recommendation results on January 8, devices operated on January 7 are more likely to be recommended than those operated on January 1. Devices operated on 4 days from January 1 to January 7 are more likely to be recommended than those operated on only 1 day. Devices operated 5 times on January 7 are more likely to be recommended than devices operated once on January 7.
[0041] Integrate and count the obtained home device data table and device recommendation results to obtain the corresponding statistical results, and prioritize the statistical results to obtain a preliminary recommendation result table, that is, the preliminary recommendation result table includes: host ID (reg_sn_id), area (area_id), recommended device ID table, and the recommended value corresponding to the recommended device and the corresponding recommended time period. Integrate and count the scene data table and the recommendation results of the scene to obtain the statistical results, and prioritize the statistical results to obtain a preliminary recommendation result table of the scene. The preliminary recommendation result table of the scene includes: host ID (reg_sn_id), area (area_id), recommended scene ID table, and the recommended value corresponding to the recommended scene and the corresponding recommended time period. Among them, the elements in the recommendation result table can be adjusted according to actual conditions. Taking the recommended device as an example, the recommendation result table 3 is shown:
[0042] Table 3
[0043]
[0044] Table 3 shows that for a user with host ID 453, devices with device IDs 10, 38, and 39 are recommended at 20:00-21:00 in area 1.
[0045] 2. Adjust the preliminary recommendation results according to weekdays and weekends, delete the weekend type devices or scenes in the weekday recommendation results, delete the weekday type devices or scenes in the weekend recommendation results, and recommend corresponding devices and / or scenes based on the adjusted recommendation results for each time period in the query information.
[0046] Based on statistical data and algorithm models, we get the hourly recommendation results for each user. We adjust the initially recommended devices and scenarios based on working days and rest days (working days and rest days are calculated based on legal dates. Working days usually refer to Monday to Friday, and rest days are usually Saturday and Sunday, and also include legal holidays). The adjustment logic is as follows:
[0047] When there are working days and holidays in the past 7 days, the recommended equipment and scenes have been operated more than 2 times at this time on working days in the past 7 days, but not on holidays, then the equipment is a working day type equipment at this time, and the equipment and scenes of the holiday type are deleted from the recommendation results. For example, a family will perform the leaving home scene at around 8 am on working days, but will not perform the leaving home scene at 8 am on holidays. Then the recommendation will only recommend the leaving home scene at 8 am on working days, and will not recommend the leaving home scene at 8 am on holidays.
[0048] When there are working days and holidays in the past 7 days, the recommended equipment and scenes have been operated more than 2 times during the holidays in the past 7 days, but not during the working time, then the equipment is a holiday type equipment at that time, and the equipment and scenes of the working day type are deleted from the recommendation results. For example, if an entertainment device is turned on from 3 to 5 pm on a holiday, but not on a working day, then the entertainment device is a holiday type entertainment device.
[0049] If there are working days and holidays in the past 7 days, and the recommended devices and scenes have been operated at the same time on working days and holidays in the past 7 days, then the devices and scenes are considered to be general type devices. For example, if a light device is turned on every night, then the light device is a general type device.
[0050] If the current day is a weekday, weekday-type and general-type devices and scenarios are recommended. If the current day is a holiday, holiday-type and general-type devices and scenarios are recommended.
[0051] According to the adjusted recommendation results, the final recommendation results are prioritized to obtain a recommendation result table, and the corresponding devices and / or scenarios are recommended according to the adjusted recommendation results for each time period in the query information.
[0052] The recommended devices and scenes for each time period are different, and the number of recommendations may also be different. Each time period is judged based on the operation records of home devices and scenes in each time period within the set date period (7 days) closest to the current time. If the user has no device operation record in a certain time period, the recommended number is 0; if the number of devices operated in a certain time period is small (less than or equal to 3), the recommended number is less than or equal to 3. If the number of devices operated in a certain time period is large (greater than 3), the first 3 recommendations are taken from the recommended value in descending order, and the recommended number is equal to 3.
[0053] Considering that the actual environment may affect the user's operation, the method of this embodiment also includes a step of adjusting the recommendation result obtained in step 2) according to the real-time environment data.
[0054] Specifically, when the smart control panel is awakened by the user to display the recommended devices and scenes, the current environmental conditions are taken into consideration at the same time. If the current room has environmental sensing devices such as an environmental box (that is, real-time monitoring of temperature, humidity, brightness, CO2 concentration, formaldehyde, and PM2.5 environmental data), the smart control panel will combine the sensor data and the environment caused by the operation of the equipment and scenes in the recommended results. If the two are not compatible, the corresponding devices and / or scenes in the recommended results will be adjusted, or the recommended order will be adjusted accordingly. For example, during the daytime on holidays, the recommended result is a daytime scene, and lighting equipment will not be turned on, and ventilation equipment will be turned on. However, the weather during the daytime on the current holiday is rainy, and turning on the ventilation equipment is not compatible with the current environment. It is necessary to adjust the ventilation equipment to turn off and turn on the lighting equipment when the brightness is low.
[0055] After the cloud sends the adjusted recommendation results to the host, the host displays them on the intelligent control panel by hour, region, and user.
[0056] Smart home control panel implementation example:
[0057] A smart home control panel in this embodiment includes a memory, a processor and an internal bus, and the processor and the memory communicate and exchange data with each other through the internal bus. The memory includes at least one memory capable of storing the time decay algorithm and the data and recommendation results of the device and the scene. The processor executes various functional applications and data processing by running the software programs and modules stored in the memory, thereby realizing the recommendation method of the smart home control panel introduced in the method embodiment of the present invention.
[0058] The processor may be a microprocessor MCU, a programmable logic device FPGA or other processing device.
[0059] The memory may be various memories that use electrical energy to store information, such as RAM, ROM, etc.; it may also be various memories that use magnetic energy to store information, such as hard disks, floppy disks, magnetic tapes, magnetic core memories, bubble memories, USB flash drives, etc.; it may also be various memories that use optical methods to store information, such as CDs, DVDs, etc.; of course, it may also be other types of memories, such as quantum memories, graphene memories, etc.
[0060] Specific implementation methods are given above, but the present invention is not limited to the described implementation methods. The basic idea of the present invention lies in the above basic scheme. For ordinary technicians in this field, it does not take creative work to design various deformed models, formulas, and parameters according to the teachings of the present invention. Changes, modifications, substitutions, and variations of the implementation methods without departing from the principles and spirit of the present invention still fall within the scope of protection of the present invention.
Claims
1. A method for recommending a smart home control panel, characterized in that: The following steps are involved: 1) According to the host, user, device and scene area, obtain the operation records of home devices and scenes in each time period within the set date period closest to the current time, form device data and scene data respectively, and use the time decay algorithm to calculate each time period based on the device data and scene data to obtain the preliminary recommendation results of home devices and / or scenes in each area in each time period; 2) Adjust the preliminary recommendation results according to weekdays and weekends, delete the weekend type devices or scenes in the weekday recommendation results, and delete the weekday type devices or scenes in the weekend recommendation results to obtain adjusted recommendation results.
2. The method for recommending a smart home control panel according to claim 1, characterized in that: The devices and scenes of the weekday type are those for which the number of operations in the first set time period on weekdays is greater than the set number of operations, and there is no operation in this time period on weekends; the devices and scenes of the holiday type are those for which the number of operations in the second set time period on weekends is greater than the set number of operations, and there is no operation in this time period on weekdays.
3. The method for recommending a smart home control panel according to claim 1 or 2, characterized in that: The method also includes the step of adjusting the recommendation result obtained in step 2) according to the real-time environmental data.
4. The method for recommending a smart home control panel according to claim 3, characterized in that: The steps of adjusting according to the real-time environment data are as follows: comparing the obtained real-time environment data with the environment caused by the device and scene operation in the recommendation results. If the two are not compatible, adjusting the corresponding devices and scenes in the recommendation results, or adjusting the recommendation order accordingly according to the set environment data.
5. The method for recommending a smart home control panel according to claim 1, characterized in that: The device data includes the acquisition date, the time period of the acquisition date, the host and the corresponding host ID, the device ID, the device name, the area ID and the number of times the device is used; the scene data includes the acquisition date, the time period of the acquisition date, the host and the corresponding host ID, the scene ID, the scene name, the area ID and the number of times the scene is used.
6. The method for recommending a smart home control panel according to claim 3, characterized in that: The device recommendation results include the host ID, recommended device ID, recommended value corresponding to the recommended device, and the corresponding recommended time period. The scene recommendation results include the host ID, user area ID, recommended scene ID, recommended value corresponding to the recommended scene, and recommended time period.
7. The method for recommending a smart home control panel according to claim 1, characterized in that: The time period within the set date closest to the current time is every hour in the 7 days closest to the current time.
8. The method for recommending a smart home control panel according to claim 1, characterized in that: The time decay algorithm model is: G is the time attenuation coefficient, x1-x T is the number of device clicks at a certain time within 1-T days, X is the cumulative operation days vector of the device at that time, a is the cumulative days impact factor, and Y n is the recommendation index vector for devices and scenarios at point n.
9. A smart home control panel, comprising a processor and a display, wherein the processor is connected to the display, characterized in that: The processor is used to execute program instructions to implement the recommendation method of the smart home control panel as described in any one of claims 1-8, and the display is used to display the recommendation results processed by the processor.
