Method and device for judging whether there is person in room or not

The decision tree model adjusts the duration of unjusted judgment, which solves the problem of misjudgment of fixed time period judgments in the prior art, improves the accuracy and user experience of room status judgments, and reduces energy waste.

CN119026024AActive Publication Date: 2024-11-26ULTIMATE IOT (HENAN) TECHNOLOGY LTD +1
View PDF 6 Cites 0 Cited by

Patent Information

Application Number
CN202310608138.9
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2023-05-26
Publication Date
2024-11-26
Estimated Expiration
2043-05-26

AI Technical Summary

Technical Problem

The prior art determines the room status by setting a fixed time period for person-free jump, which can easily cause misjudgment, affect the user experience, and may lead to reduced equipment service life and waste of energy.

Method used

Use the decision tree model to train based on historical data to adjust the duration of unjusted people. By collecting the duration of unjusted people and the cumulative number of people in each time period, adjusting the duration of unjusted people, improving the accuracy of judgment, and avoiding unjusted people through restrictions. The duration is too long or too short.

Benefits of technology

It improves the accuracy and user experience of unauthorized judgments, ensures that room status judgments are more in line with practical applications, and reduces the problems of energy waste and frequent equipment conversion.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN119026024A_ABST
    Figure CN119026024A_ABST
Patent Text Reader

Abstract

The invention relates to the technical field of smart home, in particular to a method and device for judging whether a person is in a room or not. According to the scheme, the decision tree model is trained according to the set decision condition by collecting historical data related to judgment of whether there is a person or not, and the decision tree model is utilized to obtain an adjustment mode of the duration of judgment of no person in different rooms in different time periods. And the output result of the decision tree model is corrected by using a set unmanned judgment duration adjustment logic to obtain the unmanned judgment duration of each room in each time period, and finally, according to the unmanned judgment duration of each room in each time period, the manned / unmanned state of each room in each time period is judged. According to the scheme, the adjustment of the unmanned judgment duration can be realized, and the accuracy of the unmanned judgment duration can be improved, so that the judgment on the state that people are in the room or not in different scenes is more accurate, the method is more suitable for the practical application, and the user experience can be improved.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the field of smart home technology, and in particular to a method and device for determining whether a room is occupied or not. Background Art

[0002] The existing occupied or unoccupied algorithm can judge whether a room is occupied by a user manually or reported by a security device. Therefore, the judgment of occupied is often accurate, and the room status will be changed to occupied. However, for the judgment of unoccupied, the main judgment logic of the existing technology is to change the room status to unoccupied after no manual operation record or security device report within a period of time (such as 30 minutes or 15 minutes).

[0003] However, after analyzing the data related to occupancy and unoccupancy, and combining it with the actual situation of the office area of ​​the office building, the existing technology of setting a fixed occupancy and unoccupancy jump time period to determine the occupancy and unoccupancy status is prone to misjudgment. For example, in certain time periods, there are rooms that become unoccupied and then become occupied in a short period of time. In particular, rooms that are equipped with occupancy-related equipment and scene linkage will cause frequent changes in the room status, which will not only affect the service life of the equipment, but also affect the user experience; and when working in the office area at night or having a meeting in the conference room, there may be a long period of small movements without triggering the human infrared sensor, resulting in the misjudgment of occupancy as unoccupied and the lights being turned off. The equipment needs to be manually adjusted to the occupancy state and then the lights are turned on, which will also affect the user experience. In both of the above situations, it is easy to misjudge occupancy as unoccupied, so it is necessary to set a longer unoccupied duration before turning the room into an unoccupied room.

[0004] Similarly, there are also cases where a room remains unoccupied for a long time after it becomes unoccupied. For example, after the last person leaves the office area at night, the occupied and unoccupied state will still be judged according to the fixed occupied and unoccupied transition time period set, so that it takes a long time to judge that the room is unoccupied, which wastes energy. Summary of the invention

[0005] The purpose of the present invention is to provide a method for determining whether a room is occupied or unoccupied, so as to solve the problem that the prior art method of determining the occupied or unoccupied status by setting a fixed occupied or unoccupied jump time is prone to misjudgment, thereby affecting the user experience; and also provides a device for determining whether a room is occupied or unoccupied, which is used to implement the process of the above method.

[0006] In order to solve the above technical problems, the present invention provides a method for determining whether a room is occupied or not, comprising the following steps:

[0007] 1) Collect historical data related to occupancy and unoccupancy, and use the collected data to train a decision tree model according to the set decision conditions; data related to occupancy and unoccupancy include the duration of occupancy and cumulative number of occupancy in each room in each collected time period; duration of occupancy refers to the time when the current room status changes from occupied to unoccupied; cumulative number of occupancy refers to the number of times occupancy is collected in each time period; decision conditions refer to the number of changes in the occupancy and unoccupancy status in each time period;

[0008] 2) Collect data related to judging whether there are people in different rooms at different time periods, input the collected data into the decision tree model, and obtain the adjustment method of the length of time when no one is judged; the adjustment method of the length of time when no one is judged includes increasing, keeping unchanged and decreasing;

[0009] 3) The output result of the decision tree model is corrected by using the set unmanned judgment time adjustment logic to obtain the unmanned judgment time of each room in each time period; and according to the unmanned judgment time of each room in each time period, the occupied or unmanned status of each room in each time period is judged.

[0010] Beneficial effect: Compared with the method of setting a fixed position to judge whether there is someone or not in the prior art, on the one hand, the present invention uses the collected data related to judging whether there is someone or not to train a decision tree model according to the set decision conditions, and uses the output result of the decision tree model to determine the adjustment method of the unmanned judgment time, which can realize the adjustment of the unmanned judgment time, thereby improving the accuracy of the unmanned judgment time; on the other hand, the present invention can further improve the accuracy of the unmanned judgment time by correcting the output result of the decision tree model by using the unmanned judgment time adjustment logic, thereby making the judgment on the status of someone or not in the room in different scenarios more accurate and more in line with the actual application, which is conducive to improving the user experience.

[0011] Furthermore, the logic for adjusting the unmanned judgment time includes: if the current adjustment method of the unmanned judgment time is unchanged, then remain unchanged; if the current adjustment method of the unmanned judgment time is increase, and the adjustment method of the unmanned judgment time in the same time period of the previous day was not increase or the adjustment method of the unmanned judgment time in the same time period of the previous several days was not increase, then increase, otherwise remain unchanged; if the current adjustment method of the unmanned judgment time is decrease, and the adjustment method of the unmanned judgment time in the same time period of the previous day was not decrease or the adjustment method of the unmanned judgment time in the same time period of the previous several days was not decrease, then decrease, otherwise remain unchanged.

[0012] Beneficial effect: The present invention can avoid continuous adjustment causing the unmanned judgment time to be too long or too short by adding restrictions on the adjustment method of not making the same unmanned judgment time for adjacent time periods, which is beneficial to improving the accuracy of the adjustment of the unmanned judgment time.

[0013] Furthermore, the logic for adjusting the unmanned judgment duration is: if the output result of the decision model for the current time period is an increase, the output result of the decision model for the previous time period is adjusted to an increase; if the output result of the decision model for the current time period is a decrease, and the output result of the decision model for the adjacent previous time period is unchanged, the output result of the decision model for the previous time period is adjusted to a decrease.

[0014] Beneficial effect: Since the decision tree model considers that the time period in which the output result is located is the current processing time period, in fact, the output result of the current time period corresponds to the judgment result of the unmanned judgment time length of the previous time period. The present invention increases the accuracy of adjusting the unmanned judgment time length of each time period by adding this restriction condition, thereby improving the rationality of the algorithm's adjustment of the unmanned judgment time length when outputting results across time periods.

[0015] Further, if the output result of the decision tree model is an increase, and if the original unmanned judgment time of the current time period is the first unmanned judgment time, it is adjusted to the second unmanned judgment time; if it is the second unmanned judgment time, it is adjusted to the third unmanned judgment time; if it is the third unmanned judgment time, it remains unchanged; if the output result of the decision tree model is unchanged, the unmanned judgment time of the current time period remains unchanged; if the output result of the decision tree model is a decrease, the original unmanned judgment time of the current time period remains unchanged if it is the first unmanned judgment time; if it is the second unmanned judgment time, it is adjusted to the first unmanned judgment time; if it is the third unmanned judgment time, it is adjusted to the second unmanned judgment time; the first unmanned judgment time, the second unmanned judgment time, and the third unmanned judgment time increase sequentially.

[0016] Beneficial effect: The present invention increases or decreases the unoccupied time in the corresponding time period according to the output result of the decision tree model, or keeps the unoccupied time in the corresponding time period unchanged, which can make the judgment result of whether there is someone in the room more humane, thereby improving the user experience.

[0017] Furthermore, the unmanned judgment duration of each room in each time period in step 3) is regularly stored in the cloud at set time intervals. When the cloud cannot receive data within the corresponding time interval, the last unmanned judgment duration is kept.

[0018] Beneficial effect: By regularly storing the occupancy judgment duration of each room in each time period in the cloud at set time intervals, the occupancy judgment duration can be accurately obtained based on the records of changes in the occupancy status, which is convenient for data statistics and maintenance.

[0019] Furthermore, the decision tree model is repeatedly trained using the collected data related to judging whether there are people or not, so as to obtain an optimized decision tree model.

[0020] Beneficial effect: By repeatedly training the decision tree model according to a set time period using continuously updated data related to judging whether there is someone or not, the decision tree model is made closer to reality, which is beneficial to improving the accuracy of the algorithm in adjusting the length of time for judging whether there is no one.

[0021] The present invention also provides a device for determining whether a room is occupied or not, the device comprising a processor and a memory, wherein the memory is used to store executable instructions of the processor; the processor is configured to execute any of the above-mentioned methods for determining whether a room is occupied or not by executing the executable instructions. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] Figure 1 It is a schematic diagram of a method for determining whether a room is occupied or not according to the present invention. DETAILED DESCRIPTION

[0023] In order to make the purpose, technical solutions and advantages of the present invention more clear, the technical principle and practical application of the present invention are further described in detail below in conjunction with the accompanying drawings and embodiments.

[0024] Example of method for determining whether a room is occupied or not:

[0025] In view of the above problems, the present invention mainly uses a decision tree model to calculate the time required for the room occupancy and unoccupancy state conversion corresponding to each time period in a day based on the historical occupancy and unoccupancy state change record data of each room.

[0026] The room occupancy status determination method of this embodiment is generally divided into three stages: model training stage, cloud computing stage and host application stage. Figure 1 The implementation process of each stage is as follows:

[0027] 1. Model training stage

[0028] The purpose of the model training phase is to obtain a decision tree model, which includes three steps: collecting data, labeling data, and training the decision tree model. The entire phase can be run once to obtain a model with good results. This phase can be re-executed from time to time to obtain a more suitable model based on the algorithm operation status.

[0029] A decision tree is a tree structure in which each node represents a test on an attribute, each branch represents a test output, and each leaf node represents a classification category. By building a decision tree with training data, unknown data can be classified.

[0030] By collecting and processing the relevant data on manned and unmanned driving, a decision tree model is trained according to the decision conditions to obtain the adjustment method of the unmanned driving judgment time. The decision conditions refer to the number of times the manned and unmanned driving status changes in each time period, and the adjustment methods of the unmanned driving judgment time include increase, unchanged and decrease.

[0031] The algorithm mainly uses the room occupancy change data, including specific time and occupancy value, and analyzes the data characteristics according to the room occupancy change data, annotates and processes the data by the duration of occupancy, and obtains and judges the occupancy-related data, namely: the number of times the occupancy duration exceeds 30 minutes, the number of times the occupancy duration is less than 15 minutes, the number of times the occupancy duration is between 15-30 minutes, and the cumulative number of occupancy. Among them, the duration of occupancy refers to the duration from the room becoming occupancy to the next time the room becomes occupancy, in minutes. For example, in the second data, the room becomes occupancy at 7:41 and becomes occupancy at 8:58, so the duration of occupancy is 77 minutes; the cumulative number of occupancy refers to the number of times there are people in each time period during the collection and judgment of occupancy-related data.

[0032] Then, the data related to judging whether there is someone or not is input into the decision tree model for training, and the output results are -1, 0, and 1. According to the output results, the corresponding unmanned judgment time is adjusted to 15 minutes, 30 minutes, and 60 minutes. Among them, -1 means shortening the unmanned judgment time, that is, changing the room status to unmanned after 15 minutes of no operation; 0 means keeping the current unmanned judgment time unchanged, that is, changing the room status to unmanned after 30 minutes of no operation; 1 means increasing the unmanned judgment time, that is, changing the room status to unmanned after 60 minutes of no operation.

[0033] For example, the specific time and status of occupancy and occupancy of a negotiation room at different times on August 3, 2022 are shown in Table 1.

[0034] Table 1

[0035] name No one time Is there anyone in the negotiation room? 1 2022 / 8 / 3 7:15 Is there anyone in the negotiation room? 0 2022 / 8 / 3 7:46 Is there anyone in the negotiation room? 1 2022 / 8 / 3 8:02 Is there anyone in the negotiation room? 0 2022 / 8 / 3 12:35 Is there anyone in the negotiation room? 1 2022 / 8 / 3 13:46 Is there anyone in the negotiation room? 0 2022 / 8 / 3 18:41 Is there anyone in the negotiation room? 1 2022 / 8 / 3 19:20 Is there anyone in the negotiation room? 0 2022 / 8 / 3 22:12

[0036] According to the characteristics of the data, the data is labeled and processed by the duration of unattended operation. The rules are as follows:

[0037] 1. If the room status is changed to occupied within 15 minutes after the room status is changed to unoccupied, it means that the timer duration should be increased at this time point.

[0038] 2. If the room status is not changed within 15 minutes after it is changed to No One, it means that the timer duration remains unchanged at this time point.

[0039] 3. If the room status is not changed within 30 minutes after it is changed to unoccupied, it means that the timer duration should be reduced at this time point.

[0040] The historical occupancy and unoccupancy data of different rooms in different time periods are counted and processed. Here, 7 days is taken as a cycle. The occupancy and unoccupancy data of a negotiation room from July 25th to July 31st is taken as an example. The data after annotation and processing are shown in Table 2.

[0041] Table 2

[0042] time Number of times no one was there for more than 30 minutes Number of times no one was there for more than 15 minutes Number of times no one was there for less than 15 minutes Total number of days with people 0 0 0 0 0 1 0 0 0 0 2 0 0 0 0 3 0 0 0 0 4 0 0 0 0 5 0 0 0 0 6 0 0 0 0 7 5 0 0 5 8 0 0 1 3 9 0 0 1 6 10 0 0 0 6 11 0 0 0 6 12 3 0 3 6 13 1 1 0 4 14 1 0 1 6 15 1 1 0 5 16 1 0 0 4 17 1 0 0 3 18 1 0 0 4 19 0 2 0 4 20 1 0 3 5 21 3 0 0 4 22 1 0 1 1 23 0 0 0 0

[0043] Among them, time refers to the moment, such as 1 refers to from 1:00 to 2:00; unoccupied time refers to the duration of unoccupied time, that is, the duration of unoccupied time after the room status changes from occupied to unoccupied at the current time, and is counted according to the intervals of unoccupied time of less than 15 minutes, 15 minutes to 30 minutes, and more than 30 minutes; the cumulative number of occupied days refers to the number of days in the past 7 days when the occupant time at this moment exceeded 0 minutes.

[0044] Take the eighth data in Table 2 as an example to illustrate the physical meaning of each data in each row. The data is [7, 5, 0, 0, 5], which means that in the time period of 7:00-8:00, there were 5 times in the past 7 days when the room status changed from occupied to unoccupied in this time period and the unoccupied duration exceeded 30 minutes. The number of times below 15 minutes and between 15 minutes and 30 minutes was 0. In the past 7 days, there were 5 days when the room status was occupied in this time period.

[0045] 2. Cloud computing stage

[0046] The purpose of the cloud computing stage is to obtain the unmanned judgment duration of each room in each period of each host based on the data and decision tree model and store it in the cloud. It includes three steps: collecting and processing the data related to occupancy, inputting the collected data into the decision tree model to obtain the unmanned judgment duration of each room at each time, and storing the results in the cloud. The entire stage needs to be run once a day to obtain the unmanned judgment duration of each room in each period of the day.

[0047] The cloud computing logic is mainly:

[0048] (1) Based on the changes in manned and unmanned conditions and the decision tree, it is determined whether the unmanned time per hour increases, remains unchanged, or decreases.

[0049] The historical occupancy data is counted and processed by room and time period. Taking 7 days as a cycle, the number of times when no one was present for less than 15 minutes, 15 minutes to 30 minutes, and more than 30 minutes is counted, and the number of days when the occupancy time in this time period was more than 0 minutes in the past 7 days is counted. Generally speaking, the time should be increased for the case where no one is present for less than 15 minutes, and the time should be reduced for the case where no one is present for more than 30 minutes.

[0050] For example, if there is a high probability that someone is in a meeting room after 9 o'clock, but the movements of the people in the meeting room are relatively small and not easily detected by infrared sensing. Assuming that the original unmanned judgment time is 30 minutes, it is judged as unmanned at 9:30, and the people in the meeting are sensed by infrared at 9:35. If there is no one in the room for 5 minutes (within the interval of no one duration of less than 15 minutes), it becomes occupied. This is a situation where there are frequent activities in a short period of time, and the unmanned judgment time of this time period should be increased.

[0051] For example, if statistics show that there are many times when no one is around for more than 30 minutes after 18:00, the length of time for judging no one in this time period should be reduced.

[0052] The specific unmanned judgment time is determined based on the algorithm result and the current unmanned judgment time. For example, the unmanned judgment time is set to three levels: 15, 30, and 60. The unmanned judgment time of Room A at 8 o'clock is 15 minutes, and the unmanned judgment time at 9 o'clock is 30 minutes. The result of the algorithm judging at 8 o'clock and 9 o'clock on the same day is to increase the time, from 15 minutes to 30 minutes at 8 o'clock, and from 30 minutes to 60 minutes at 9 o'clock. As other implementation methods, the set time and / or the set level can be adjusted according to the unmanned judgment time of the current room and the decision tree optimization results to meet the personalized needs in different scenarios.

[0053] (2) Correct the current results based on the algorithm results of the previous week.

[0054] Because the calculation results are based on the data of the past week, it may happen that the time should be increased or decreased after the time has been increased or decreased, so it is necessary to add restrictions: do not increase or decrease the time without judgment continuously. If the algorithm judgment result is unchanged, then the result remains unchanged; if the judgment result is increased, and the previous day was not increased or the judgment results were increased for 7 consecutive days, then increase, otherwise remain unchanged; if the judgment result is decreased, and the previous day was not decreased or the judgment results were decreased for 7 consecutive days, then decrease, otherwise remain unchanged.

[0055] If the duration is increased on January 1, it will be increased on January 1; on the second day, the algorithm still determines that it is increased on January 2, because January 1 is an increase. Based on the principle of discontinuous increase or decrease, there will be no increase on January 2 and it will remain unchanged; if the algorithm determines that it is unchanged on January 2, it will remain unchanged on January 2; if the algorithm determines that it is reduced on January 2, no one determines that the duration is reduced on January 2. If January 1 is unchanged, it will not affect the judgment on January 2. If January 1 is reduced, January 2 is also reduced. Based on the principle of discontinuous increase or decrease, the result of January 2 is modified to be unchanged. If January 2 is an increase, the duration will be increased; if January 2 is unchanged, the duration will remain unchanged.

[0056] For example, if we use the occupied and unoccupied data from July 1 to July 7 to judge the unoccupied judgment time on July 8, if the decision tree model outputs 1, we need to increase the unoccupied judgment time, that is, 60 minutes of no operation to change the room status to unoccupied. Then, we use the occupied and unoccupied data from July 2 to July 8 to judge the unoccupied judgment time on July 9. If the decision tree model outputs 1, since the occupied and unoccupied data on July 8 used are the data with increased unoccupied judgment time, if the unoccupied judgment time is continuously increased in adjacent time periods, it may affect the accuracy of the results output by the decision tree model. Therefore, we add the restriction condition of discontinuously increasing or decreasing the unoccupied judgment time to ensure the accuracy of the results output by the decision tree model.

[0057] (3) Modify the results of the previous hour based on the decision tree results of the current hour.

[0058] If the output result of the decision model for the current time period is an increase, the output result of the decision model for the previous time period is adjusted to an increase; if the output result of the decision model for the current time period is a decrease, and the output result of the decision model for the adjacent previous time period is unchanged, the output result of the decision model for the previous time period is adjusted to a decrease.

[0059] For the case where the result obtained by the algorithm using the data of the previous time period occurs in the next time period, this embodiment increases the current hour to increase, and the previous hour should also increase; if the current hour decreases, the previous hour remains unchanged, and the previous hour is changed to a decreasing restriction condition to improve the rationality of the algorithm's adjustment of the unattended judgment time when outputting results across time periods.

[0060] Note that in step (2), the time unit for correcting the current result based on the algorithm result of the previous week is day, that is, the time unit for discontinuous increase or decrease judgment is day; in step (3), the time unit for correcting the result of the previous hour based on the decision tree result of the current hour is hour, that is, the time unit for continuous increase or decrease of the time period is hour.

[0061] The output result of the decision tree model refers to the judgment result of increase, unchanged, or decrease obtained after inputting the historical manned and unmanned data into the model. The judgment result of the above algorithm refers to the judgment result of increase, unchanged, or decrease obtained by combining the decision tree model and logical judgment.

[0062] The following examples illustrate the applicable scenarios of the above restrictions. For the current hour, the previous hour should also be increased. For example, suppose there is always someone in a room during the 7:00-9:00 time period, but the person's movements in the room are relatively small, and there may be a situation where the infrared sensor cannot sense it. If 7:50 is the last manual operation or infrared sensing record during the 7:00-8:00 time period, and the unmanned judgment time is 30 minutes, the room will become unmanned at 8:20. However, in fact, there is someone in the room and there is a manual operation record at 8:25, and the room changes from unmanned to manned, that is, the room becomes manned again within 15 minutes after it becomes unmanned during the 8:00-9:00 time period. Therefore, the decision tree model may determine that the unmanned judgment time needs to be increased at 8 o'clock. However, in fact, this judgment result is based on the last manual operation or infrared sensing record at 7 o'clock, so the unmanned judgment time at 7 o'clock should be increased.

[0063] For the current hour, the previous hour remains unchanged, and the previous hour is changed to a decrease. For example, when the judgment starts at 7:50 and the unattended judgment time is 30 minutes, then 8:20 is judged as unattended. If there is no action for a long time afterwards, the unattended judgment time at 8 o'clock will be changed from 30 minutes to 15 minutes. In fact, this result was obtained when the judgment started at 7 o'clock, so the unattended judgment time at 7 o'clock should be changed from 30 minutes to 1 minute. The premise is that the decision tree result at 7 o'clock remains unchanged.

[0064] 3. Host application stage

[0065] The purpose of the host application stage is to set the occupancy judgment duration of the occupancy algorithm in each room and each period according to the cloud data. It is necessary to receive cloud data once a day for setting. In particular, when no data is received, the occupancy duration will remain the same as the last setting.

[0066] Embodiment of the device for determining whether a room is occupied or not:

[0067] The room occupancy status determination device of this embodiment includes a processor and a memory, and the processor executes a computer program stored in the memory so that the present invention implements the method of the above method embodiment. That is to say, the method in the above method embodiment should be understood as a process based on the above room occupancy status determination method that can be implemented by computer program instructions. These computer programs can be provided to issue instructions to the processor, so that the processor executes these instructions to generate functions specified by the above method flow.

[0068] The processor referred to in this embodiment refers to a processing device such as a microprocessor MCU or a programmable logic device FPGA; the memory referred to in this embodiment includes a physical device for storing information, which usually digitizes the information and then stores it in a medium using electrical, magnetic or optical methods. For example: various memories that use electrical energy to store information, such as RAM, ROM, etc.; various memories that use magnetic energy to store information, such as hard disks, floppy disks, tapes, magnetic core memories, bubble memories, and U disks; various memories that use optical methods to store information, such as CDs or DVDs. Of course, there are other types of memories, such as quantum memories, graphene memories, etc.

[0069] The device is composed of the above-mentioned memory, processor and computer program, and is implemented by the processor executing corresponding program instructions in the computer. The processor can be equipped with various operating systems, such as Windows operating system, Linux system, Android, iOS system, etc.

[0070] As another embodiment, the device may further include a display, and the display is used to display the classification results for reference by staff.

Claims

1. A method for determining whether a room is occupied or not, characterized in that: The steps include: 1) Collect historical data related to judging whether a room is occupied or not, and use the collected data to train a decision tree model according to the set decision conditions; the data related to judging whether a room is occupied or not includes the duration of each room being occupied and the cumulative number of times that a room is occupied in each time period collected; the duration of each room being occupied refers to the time that the room is occupied after the current room status changes from occupied to unoccupied; the cumulative number of times that a room is occupied refers to the number of times that a room is occupied in each time period; the decision condition refers to the number of times that the occupied or unoccupied state changes in each time period; 2) Collecting data related to judging whether there are people in different rooms in different time periods, inputting the collected data into the decision tree model, and obtaining an adjustment method for judging whether there are people in the room; the adjustment method for judging whether there are people in the room includes increasing, keeping unchanged and decreasing; 3) The output result of the decision tree model is corrected by using the set unmanned judgment time adjustment logic to obtain the unmanned judgment time of each room in each time period; and according to the unmanned judgment time of each room in each time period, the occupied or unmanned status of each room in each time period is judged.

2. The method for determining whether a room is occupied or not according to claim 1, characterized in that: The logic for adjusting the unmanned judgment time period includes: if the current adjustment method of the unmanned judgment time period is unchanged, then remain unchanged; if the current adjustment method of the unmanned judgment time period is increase, and the adjustment method of the unmanned judgment time period in the same time period of the previous day was not increase or the adjustment methods of the unmanned judgment time period in the same time period of the previous several days were not increase, then increase, otherwise remain unchanged; if the current adjustment method of the unmanned judgment time period is decrease, and the adjustment method of the unmanned judgment time period in the same time period of the previous day was not decrease or the adjustment methods of the unmanned judgment time period in the same time period of the previous several days were not decrease, then decrease, otherwise remain unchanged.

3. The method for determining whether a room is occupied or not according to claim 1 or 2, characterized in that: The logic for adjusting the unattended judgment time is as follows: if the output result of the decision model for the current time period is an increase, the output result of the decision model for the previous time period is adjusted to an increase; if the output result of the decision model for the current time period is a decrease, and the output result of the decision model for the adjacent previous time period is unchanged, the output result of the decision model for the previous time period is adjusted to a decrease.

4. The method for determining whether a room is occupied or not according to claim 3, characterized in that: If the output result of the decision tree model is an increase, and the original unmanned judgment time of the current time period is the first unmanned judgment time, it is adjusted to the second unmanned judgment time; if it is the second unmanned judgment time, it is adjusted to the third unmanned judgment time; if it is the third unmanned judgment time, it remains unchanged; if the output result of the decision tree model is unchanged, the unmanned judgment time of the current time period remains unchanged; if the output result of the decision tree model is a decrease, the original unmanned judgment time of the current time period remains unchanged if it is the first unmanned judgment time; if it is the second unmanned judgment time, it is adjusted to the first unmanned judgment time; if it is the third unmanned judgment time, it is adjusted to the second unmanned judgment time; the first unmanned judgment time, the second unmanned judgment time, and the third unmanned judgment time increase sequentially.

5. The method for determining whether a room is occupied or not according to claim 4, characterized in that: The unmanned judgment duration of each room in each time period in step 3) is regularly stored in the cloud at set time intervals. When the cloud cannot receive data within the corresponding time interval, the last man-or-no man judgment duration is kept.

6. The method for determining whether a room is occupied or not according to claim 5, characterized in that: The decision tree model is repeatedly trained using the collected data related to determining whether there is someone or not, so as to obtain an optimized decision tree model.

7. A device for determining whether a room is occupied or not, characterized in that: The device comprises a processor and a memory, wherein the memory is used to store executable instructions of the processor; the processor is configured to execute the method for determining whether a room is occupied or not as described in any one of claims 1 to 6 by executing the executable instructions.

Citation Information

Patent Citations

  • Air conditioner running control method and device

    CN106052051A

  • Air conditioner running control method and device

    CN106225166A

  • Manned-unmanned detection method and device, electronic equipment and system

    CN110209059A

  • Room state prediction method and device, electronic equipment and storage medium

    CN110717376A

  • Indoor personnel state acquisition method and device, storage medium and electronic equipment

    CN111198508A