JPH02254247A - Predicting device for pattern of occupied room - Google Patents

Abstract

PURPOSE:To enable automatic operation of an air conditioner by a method wherein through estimation of a present occupied condition of a room and living action, instructions on turning ON and OFF is outputted to an air conditioner. CONSTITUTION:According to digital information from an occupied room information sensor 4, an occupied room condition and living action estimating means 7 estimates which is a present occupied room condition based on a plurality of predetermined occupied room conditions and its transient conditions, and a change in occupied room conditions are estimated as a living action. In order to effect prediction, the features of conditions of using the room of an object house are picked up by a living pattern pick-up means 10, and a pick-up result is stored in a living pattern storage means 11. A means 16 for predicting coming-into-room and getting up predicts a coming-into- room time from a present occupied conditions and a picked up living pattern when the room is not occupied for a long period is estimated, and predicts a getting up time when sleeping is estimated. Since prediction is accompanied with probability, an effective time of a predicting result is also outputted. This constitution enables reduction of the number of control times of a user, by which an air conditioner is turned ON and OFF, to a minimum value.

Description

DETAILED DESCRIPTION OF THE INVENTION Field of Industrial Application The present invention is a room occupancy pattern prediction device used primarily for automatic operation control of domestic air conditioners (hereinafter referred to as air conditioners) and energy supply equipment.

Conventional Technology Conventionally, in order to operate an air conditioner when it is not possible to operate the device directly, such as when a person is absent or asleep, it has been done by setting a timer or by remote control using a telephone line. In addition, to stop the operation of the air conditioner,
Air conditioner users were operating the air conditioner using the control panel or remote control device.

Regarding prediction, there has been no device for predicting the putter 7 at home.

Problems to be Solved by the Invention However, with conventional air conditioners, even if there is a certain degree of lifestyle such as waking up, going to bed, going out, and returning home, the timer must be set every time, and setting the timer is also troublesome and has limited functionality. Despite being attached to air conditioners, it was rarely used.

Additionally, there is a problem with turning on and off the air conditioner, even if the room is in a poor thermal environment, the air conditioner will not operate unless operated by a person in the room.

The present invention was developed in view of the inconvenience of air conditioners for users as described above, and the first purpose is to minimize the number of operations required by the user to turn on and off the air conditioner. It is about reducing. The second purpose is to extract the characteristics of past room usage patterns, predict the time of entry when the room is not occupied, and the wake-up time when the user is busy sleeping, thereby creating information for predictive operation of the air conditioner. It is to put it out.

Means for Solving the Problems In order to achieve the above object, the technical solution of the present invention is to install a room occupancy information sensor in the 1st K1 room that detects the presence or absence of a person in the room or whether the lights are on or off. That's true. Secondly, this occupancy information sensor alone cannot determine whether you have left the room for a short time or whether you have gone out, and this is sufficient to operate the air conditioner, so the state detected by the occupancy information sensor is insufficient. The object of the present invention is to have a room occupancy state/living activity estimation means for estimating the room user's life as the room occupancy state from the elapsed time, past state, and current time. Thirdly, it is necessary to have means for predicting the wake-up time and room entry time when the user is absent or sleeping. Fourthly, in order to achieve the third objective, there is provided a lifestyle pattern extraction means for extracting characteristics of past room usage situations.

Based on the digital information from the occupancy information sensor, the occupancy status and
The living activity estimation means estimates which occupancy state the user is currently in based on a plurality of predetermined occupancy states and their transition conditions. In addition, changes in the state of being in the room are estimated as daily activities. In order to make a prediction, the lifestyle pattern extraction means extracts the characteristics of the usage status of the room in the target dwelling unit, and the lifestyle pattern storage means stores the extraction results. The room entry/wake-up prediction means predicts the entry time when the room is presumed to be absent for a long time from the current room occupancy state and the extracted lifestyle pattern, and predicts the wake-up time when the room is estimated to be asleep.

Furthermore, since the prediction accuracy is accompanied by certainty, the effective time of the prediction result is also output at the same time.

Embodiment Hereinafter, an embodiment of the present invention will be described with reference to the drawings. FIG. 1 is a block diagram showing the functional configuration of a room occupancy pattern prediction device for use in controlling an air conditioner in an embodiment of the present invention.

In FIG. 1, reference numeral 1 denotes an occupancy mode setting unit that accepts input from the user regarding the use of the room and sets parameters; a room mode setting unit 2 that accepts input of the purpose of the room such as a living room or a bedroom; and a family trip, etc. It consists of a long-term absence setting section that accepts long-term absence status. Reference numeral 4 denotes a presence information sensor that detects the status of people in the room, an optical sensor 5 that detects the on/off status of lights as digital information, and a sensor that changes to 1 when the person in the room is active and O when the person is absent or stationary. It consists of an infrared occupancy sensor set to Reference numeral 7 denotes a room occupancy state/living activity estimation means for identifying one of a plurality of room occupancy states with predetermined conditions based on the measurement results from the room occupancy mode setting means 1 and the room occupancy information sensor 4. , an occupancy state/life activity estimator 8 that identifies the occupancy state and the state according to an algorithm that determines the transition conditions thereof;
It consists of an event check section 9 that checks whether to perform lifestyle pattern extraction or prediction according to the occupancy state estimated within the daily activity estimation section 8.

IO extracts parameters that characterize room use from the output parameters of the room occupancy status/life activity estimation means 7, such as the start time of absence, start time of entering the room, start time of sleep, start time of waking up, duration of absence, and longevity. The lifestyle pattern extraction means 11 is a lifestyle pattern storage means for storing the results of the lifestyle pattern extraction means 10. Life pattern memory means 1
1 has weekdays and holidays as long-term characteristics, divides the day into 10-minute intervals, and calculates the frequency of absence for each day characteristic in two dimensions of absence start time and absence duration, 144 x 144 two-dimensional. Absence of going out table 12 stored in an array; sleep table 13 storing sleep start times as frequencies in a one-dimensional array of 144; the next day is a weekday and the next day is a holiday; The wake-up table 14 stores the wake-up start time as a one-dimensional array of 144, and the time zone in which the user was in the room determined by the start time of being in the room and the start time of being absent, and the number of days until now. room rate,
It consists of a room occupancy table 15 plotted in different arrangements for weekdays and holidays.

Reference numeral 16 denotes a room entry/wake-up prediction unit that predicts when a predictive activation occurs as an output of the room occupancy status/life activity estimation unit 7 or when the previously predicted result is wrong, and a room entry prediction unit 17 that predicts the room entry time. and a wake-up prediction unit 18 that predicts the wake-up time. Reference numeral 19 denotes a prediction result determination means for determining the prediction result from the result predicted by the room entry/wake-up prediction means 16 and the occupancy state outputted by the room occupancy state/life activity estimation means 7; This is an air conditioner to which the prediction device is applied.

Next, the operation of the above embodiment will be explained with reference to FIGS. 2 and 3.

FIG. 2 is a state transition diagram showing the occupancy states and their transition relationships predetermined within the occupancy state/life activity estimation means of the above embodiment. First, based on FIG. Explain daily activities.

Eight states from a to h are set as the occupancy state.

In addition, the day is generally considered to start at 4 a.m., when the number of people in the room is extremely low. a is an active occupancy state in which there is a person in the room who is moving; b is a short-term moving absence state in which a person has been absent for more than 10 minutes; and C is an I from the start time of absence.
If minutes have passed and the lights are on, you are in a state of being away from home, d is a long-term absence state, and minutes have passed since the start of your absence, and the lights are off, and you are away from home. e is when you move from room to room to sleep. f is a sleeping absent state when the person is absent from the house, f is a sleeping absent state when there is no one entering the room even when the sleeping time has come from the absentee state, and g is a sleeping absent state when the person is absent from going out and going into the sleeping time. h is a state in which the user stays up all night and is not erased, and h is a state in which the user is in the room but asleep, which is a sleep occupancy state. The sleep time periods in the above eight room occupancy states are determined based on a certain threshold value from the sleep table 13 in FIG. The above-mentioned transition relationships among the eight occupancy states have 17 possible state transitions from i to y in FIG. 2, which are called daily activities. These transition conditions are determined by the output of the infrared occupancy sensor 6, the output of the optical sensor 5, the duration of absence, and the current time. The names of each transition condition are: i for leaving the room, j for short-term move entry, k for long-term move determination, 1 for long-term move entry, m for going out decision, n for return home entry, 0 for sleep exit, and p for wake-up entry. , q is a sleepover judgment,
r is staying out and coming home, S is determining whether to stay up all night, t is staying overnight, U is sleeping, lights off, ■ is sleeping all night, lights are off, W is going out, lights off or sleeping, and y is waking up. Based on these daily activities, we predict entering the room when it is determined to be out for a long period of time (m) and when the lights are turned off when going out (W).
Predicts when you will wake up.

Next, the prediction method will be explained with reference to FIG.

FIG. 3 is a schematic diagram of the prediction method. As an example, assume that the decision to go out is made at 9:56 a.m. on a weekday. At this time, the absence start time is 9:26, which is before the addition.

A method of predicting room entry at this time will be explained. In FIG. 3, 21 is the absence start time 9 of the weekday outing/absence table 12.
The frequency for each absence duration in the 9:00 (9) minute interval, the 9 hour interval from 9:00 IO, and the 9:00 minute interval from 9:00 to 9:40 is summed up by absence duration.
It is the sum of interval classification degrees. In this household, absences of less than 10 minutes are the most common, followed by periods of 10 to 20 minutes, while absences exceeding that period are rare, and long-term absences are appearing. Party B is
This is the threshold value for determining the effective time of prediction. 23 is the dust prediction result for the first time period in which the frequency of absence equal to or higher than the threshold value 2 appears. A is the next time period in which the occurrence frequency of absences equal to or higher than the threshold value 2 will appear, which is the prediction result.

δ is the start interval of the absence, plus or minus one interval including the start time of the absence, and 9 o'clock is a single term in meaning. is the outing determination time and the start time of entering the room prediction. r is the valid time of the prediction result n, public is the predicted room entry time of the prediction result sequence, and 4 is the valid time of the prediction result sequence.

As a prediction method, a threshold value n is set in advance at a certain ratio from the total frequency of 21 in three interval classifications.

In order to predict entering the room at the outing judgment time, the total classification level of 3 sections is 2.
1 is calculated from the weekday outing/absence table.

Therefore, the first time period in which the frequency is equal to or higher than the threshold value 2 is output as the predicted room entry time. Since the prediction result is the current entry start time, only the effective time is output. In this case, since the duration of the absence is 60 minutes, "someone entered the room within 30 minutes". Therefore, it is sufficient if someone enters the room within this time, but if no one enters the room, the prediction is performed again at the effective time τ. As a result, since there is still a time period in which the threshold value 2 is exceeded, a prediction result sequence is output. That is, the entry start time and its valid time are 4. If no one enters the room at this time, 1. The time with the highest occupancy rate is output as the predicted room entry time. If it is determined that the person will not be staying overnight during or after this time, the prediction will be stopped.

Effects of the Invention As described above, according to the present invention, since the current occupancy status and daily activities are estimated, it is possible to issue on/off instructions to the air conditioner, and automatic operation of the air conditioner becomes possible. In addition, since it outputs predictions of entering the room and waking up when you are absent or asleep, you only need to set the timer to the bare minimum when you want to use it for a special purpose.

[Brief explanation of drawings]

FIG. 1 is a block diagram showing the configuration of a room occupancy pattern prediction device for use in controlling an air conditioner according to an embodiment of the present invention, and FIG. FIG. 3 is a state transition diagram showing predetermined room presence states and their transition relationships. FIG. 3 is a schematic diagram for explaining a prediction method for predicting room entry using an out-of-office table when a person is out of the house. DESCRIPTION OF SYMBOLS 1... Room occupancy mode setting means, 2... Room mode setting section, Exit... Long-term absence setting section, 4... Room occupancy information sensor,
5... Optical sensor, 6... Infrared room occupancy sensor, 7...
Room presence/living activity estimation means, 8... Living status/living activity estimation section, 9... Event check section, 10... Life pattern extraction means, 11... Life pattern storage means, 12... Chiful who does not go out, 13...Sleep table, 14...Wake up table, 15... Room presence table, 16... Room entry/wake up prediction means, 17... Room entry prediction unit, 18... Wake up prediction Section 19: Prediction result determination means; Addition: Air conditioner. Name of agent: Patent attorney Shigetaka Awano and one other person

Claims (8)

[Claims] (1) A room occupancy mode setting means that accepts registration from a room user in the case of being absent from the room for a long period of time, a room occupancy information sensor that detects the presence or absence of a person in the room, the room occupancy information sensor; occupancy state and living activity estimation means for identifying an indoor state for which a plurality of transition conditions are predetermined based on settings from the room occupancy mode setting means; and occupancy state and living activity estimation means from the room occupancy state and living activity estimation means. a lifestyle pattern extraction means for extracting a lifestyle pattern related to entering and exiting the target dwelling unit based on the transition status of the living unit; a lifestyle pattern storage unit for storing the results extracted by the lifestyle pattern extraction unit; room entry/wake-up prediction means for predicting room entry time or wake-up time based on information stored in the pattern storage means; A room occupancy pattern prediction device having a prediction result determination means for determining whether a prediction result deviates from a current room occupancy state that is an output. (2) The room occupancy pattern prediction device according to claim 1, further comprising, as the room occupancy information sensor, a sensor for detecting presence or absence of a room occupancy, and an optical sensor for detecting turning on/off of lighting. (3) The lifestyle pattern extraction means calculates daily characteristics such as the start time of going out, the duration of the absence, the bedtime, the wake-up time, the start time of being in the room, and the occupancy rate up to the start time of the absence, in units of one day. 2. The room occupancy pattern prediction device according to claim 1, wherein the device determines and outputs the determined result and stores it in a lifestyle pattern storage means. (4) The occupancy pattern prediction device according to claim 3, wherein the day characteristics are weekdays and holidays. (5) The occupancy pattern prediction device according to claim 3, wherein the day characteristic is a day of the week. (6) The occupancy pattern prediction device according to claim 3, wherein the day characteristic is that the next day is a weekday and the next day is a holiday. (7) The room occupancy pattern prediction device according to claim 1, wherein the output of the room entry/wake-up prediction means is a predicted room entry time or a predicted wake-up time, and a valid time of the room entry or wake-up prediction. (8) The room occupancy pattern prediction device according to claim 1, wherein the room occupancy mode setting means includes means for setting long-term absence and setting the purpose of the room.