JP2020103508A - Environment control system and environment control method - Google Patents

Abstract

To provide an environment control system performing napping support having high napping effect.SOLUTION: An environment control system 10 includes: a first acquisition unit 110 acquiring input information including action index information showing the action of a user; a determination unit 120 determining control contents for making the user sleep and for controlling environment control equipment 400 including light output equipment, from the input information according to a control content determination rule; a control unit 130 for controlling the environment control equipment 400 on the basis of the determined control contents; a second acquisition unit 150 acquiring evaluation information showing user evaluation to the control contents; and an update unit 180 updating the control content determination rule according to machine learning using a value based on the evaluation information as a reward.SELECTED DRAWING: Figure 1

Description

The present invention relates to an environment control system and an environment control method for determining the control content of an environment control device for a nap.

Conventionally, a system that measures a user's sleep using a non-contact biological motion sensor or the like, and proposes sleep advice for promoting a good sleep habit of the user based on the sleep information and the like obtained by the measurement is disclosed. (For example, see Patent Document 1).

Japanese Patent Publication No. 2016-532481

However, the conventional system only proposes advice and does not actually control the environmental control device. That is, in the conventional system, the user needs to change the operation of the environmental control device by himself/herself according to the sleep advice, which is troublesome. In addition, it is desired that the environmental control system perform nap support that enhances the nap effect in short-time sleep (nap).

Therefore, the present invention provides an environment control system and an environment control method capable of executing nap support with a high nap effect.

An environment control system according to one aspect of the present invention, for obtaining a nap of the user from the input information, according to a first acquisition unit that acquires input information including behavior index information indicating the behavior of the user and a control content determination rule. A control unit that determines the control content of an environmental control device including a device that outputs light, a control unit that controls the environmental control device based on the determined control content, and the control content A second acquisition unit that acquires evaluation information indicating the user's evaluation with respect to, and an updating unit that updates the control content determination rule by machine learning using a value based on the evaluation information as a reward.

An environment control method according to an aspect of the present invention, in order to make a nap of the user from the input information according to a first acquisition step of acquiring input information including action index information indicating a user's action and a control content determination rule. The control content of the environmental control equipment including a device that outputs light, a control step of controlling the environmental control equipment based on the determined control content, and the control content A second acquisition step of acquiring evaluation information indicating the evaluation of the user by the user, and an updating step of updating the control content determination rule by machine learning using a value based on the evaluation information as a reward.

Note that these comprehensive or specific aspects may be realized by a recording medium such as a system, a method, an integrated circuit, a computer program, or a computer-readable CD-ROM, and the system, the integrated circuit, the computer program, and the recording. It may be realized by any combination of media.

According to the environment control system and the environment control method according to an aspect of the present invention, it is possible to execute nap support with a high nap effect.

FIG. 1 is a block diagram showing the configuration of the environment control system according to the first embodiment. FIG. 2A is an example of input information input to the environment control device according to the first embodiment. FIG. 2B is an example of output information output by the environment control device according to the first embodiment. FIG. 3 is a first diagram for explaining an example of control contents determined by the environment control device according to the first embodiment. FIG. 4 is a second diagram for explaining an example of control contents determined by the environment control device according to the first embodiment. FIG. 5 is a third diagram for explaining an example of control contents determined by the environment control device according to the first embodiment. FIG. 6 is an example of a reward input to the environment control device according to the first embodiment. FIG. 7 is a flowchart showing processing of the environment control device according to the first embodiment. FIG. 8 is a flowchart showing a process of calculating a reward in the environment control device according to the first embodiment. FIG. 9A is a flowchart showing a process of calculating an individual reward for the quality of a nap in the environment control device according to the embodiment. FIG. 9B is a flowchart showing a process of calculating an individual reward for the time until falling asleep in the environment control device according to the embodiment. FIG. 9C is a flowchart showing a process of calculating an individual reward for a recovery effect in the environment control device according to the embodiment. FIG. 9D is a flowchart showing a process of calculating an individual reward for comfort in the environment control device according to the embodiment. FIG. 10 is a block diagram showing the configuration of the environment control system according to the first modification of the first embodiment. FIG. 11 is a conceptual diagram showing an example of a neural network in the environment control device according to the second modification of the first embodiment. FIG. 12 is a diagram showing a specific example of a plurality of control contents in the second modification of the first embodiment. FIG. 13 is a block diagram showing the configuration of the environment control system according to the second embodiment.

Hereinafter, an environment control system and an environment control method according to embodiments will be described with reference to the drawings. It should be noted that each of the embodiments described below shows a comprehensive or specific example. Numerical values, shapes, materials, constituent elements, arrangement positions of constituent elements, connection forms, steps, order of steps, and the like shown in the following embodiments are examples, and are not intended to limit the present invention. Further, among the constituent elements in the following embodiments, constituent elements not described in independent claims are described as arbitrary constituent elements.

It should be noted that each drawing is a schematic diagram and is not necessarily strictly illustrated. Further, in each drawing, the substantially same configurations are denoted by the same reference numerals, and overlapping description may be omitted or simplified.

Further, in the present specification, numerical values and numerical ranges are expressions that represent not only a strict meaning but also substantially equivalent ranges, for example, including a difference of about several percent.

(Embodiment 1)
Hereinafter, the environment control system 10 according to the present embodiment will be described with reference to FIGS. 1 to 9D. The environment control system 10 is a nap support system that performs environment control when a user takes a nap. Note that the nap is short-time sleep, for example, sleep performed by interrupting the work in the middle of the work. The work may be, for example, work, work, study, or intellectual work that uses the head such as reading, or work that uses limbs such as driving a car, operating a machine, or creating an article. Good. The nap may be sleep performed during work breaks, and is, for example, a nap. Further, the nap includes sleep taken by security guards such as the night shift during the work. A nap is sleep within 1 to 2 hours, for example, sleep of about 30 minutes. Moreover, a nap is a sleep that is lighter than a long sleep (hereinafter also referred to as a main sleep) taken by a person who is active during the daytime at night.

In the following description, a case where a user who uses a personal computer in an office for business takes a nap in a nap room installed in the office will be described as an example.

[1. Environmental control system configuration]
First, the overall configuration of the environment control system 10 will be described. FIG. 1 is a block diagram showing the configuration of the environment control system 10 according to the present embodiment. The environment control system 10 according to the present embodiment includes an environment control device 100, a sensor 200, an input device 300, and an environment control device 400.

The environment control device 100 acquires input information including action index information indicating a user's action, determines the control content of the environment control device 400 in the nap room for the user to take a nap in accordance with the acquired input information, The control content is output to the environment control device 400 and executed. Details of this environment control device 100 will be described later. The control content includes at least a control condition regarding light. Further, the control content may further include a control condition regarding at least one of video, sound, scent, vibration, temperature and humidity, and air flow.

The sensor 200 detects a user's action. Further, the sensor 200 sets information based on the detected user's action as action index information, and outputs input information including the action index information to the first acquisition unit 110 of the environment control device 100. The sensor 200 outputs the input information including the action index information to the first acquisition unit 110. The sensor 200 may be an imaging device that captures an image of the user, an illuminance sensor that measures the illuminance of the environment in which the user is active, or a pressure sensor provided in a chair on which the user sits. It may be a temperature sensor. The image pickup device is an image sensor having sensitivity in the visible light band, but may be an infrared image sensor or a thermal image sensor. The image captured by the image capturing device is, for example, a moving image, but may be a still image. Further, the sensor 200 may be another sensor capable of detecting the action of the user or a combination thereof. The sensor 200 is an example of a detection device. In the following, it is assumed that the sensor 200 is an imaging device and the user's type speed is measured from an image of the user.

The sensor 200 is provided so that a user can be sensed in each of an office and a nap room. The sensor 200 may be arranged in each of an office and a nap room, for example. The sensor 200 outputs the sensing result of sensing the user who is working before taking a nap to the first acquisition unit 110 as action index information.

Further, the sensor 200 may output the sensing result obtained by sensing the user who is taking a nap and the user who is working after taking a nap to the second acquisition unit 150 as evaluation information. That is, the sensor 200 may output the evaluation information in addition to the input information. The evaluation information is information indicating the evaluation in the nap.

The input device 300 acquires the evaluation information from the user and outputs the acquired evaluation information to the second acquisition unit 150. The input device 300 may output, for example, the information acquired from the user after the nap as evaluation information to the second acquisition unit 150. The input device 300 is, for example, a smartphone or a tablet computer, and receives input of evaluation information for a nap via a touch display. The input device 300 may be a wearable device that can be worn by the user (for example, a smart watch). Further, the input device 300 may be, for example, a microphone. In this case, the input device 300 receives a voice input from the user. Further, the input device 300 may be, for example, a mechanical push button, a keyboard, a mouse, or the like.

The input device 300 may further output information acquired from the user before the nap as input information to the first acquisition unit 110. That is, the input device 300 may output the input information in addition to the evaluation information. In addition, the input device 300 may acquire an operation for starting control of the nap environment from the user.

The numbers of the sensors 200 and the input devices 300 included in the environment control system 10 are not particularly limited.

The environment control device 400 is a device for supporting a nap. For example, the environment control device 400 controls the environment inside the nap room. The environment control device 400 includes a device that outputs light (for example, a lighting device), that is, a device that stimulates a user with light. The environment control device 400 may further include a device other than the lighting device and capable of controlling the environment for the nap. In addition to light, the environment control device 400 may further include a device that stimulates the user with images, sounds, scents, vibrations, temperature and humidity, airflow, and tactile sensations. For example, the environment control device 400 may further include an air conditioner, an air purifier, a ventilation fan, a fan, or floor heating. Further, for example, the environment control device 400 may include an opening/closing device for a light-shielding facility (for example, a blind and a curtain) that covers the window so that the window can be opened and closed. Further, for example, the environment control device 400 may include a video device, an audio device, or an audiovisual device. Further, for example, the environment control device 400 may include an aroma machine. Further, for example, the environment control device 400 may include a massager. In the present embodiment, a case will be described where environment control device 400 is a lighting device.

[1-1. Configuration of environmental control device]
Next, a detailed configuration of the environment control device 100 will be described. As shown in FIG. 1, the environment control device 100 includes a first acquisition unit 110, a determination unit 120, a control unit 130, a display unit 140, a second acquisition unit 150, a reward calculation unit 160, and a reward condition. A setting unit 170, an updating unit 180, and a storage unit 190 are provided. Further, in the environment control device 100, the reward calculation unit 160, the reward condition setting unit 170, the update unit 180, and the storage unit 190 constitute the learning unit 100b, and the first acquisition unit 110, the determination unit 120, and The display unit 140, the second acquisition unit 150, and the learning unit 100b constitute a learning device 100a (agent). The learning unit 100b learns the control content in association with the input information and the evaluation information.

The environment control device 100 includes, for example, a processor and a memory. For example, when the processor executes the software program stored in the memory, the processor causes the first acquisition unit 110, the determination unit 120, the control unit 130, the display unit 140, the second acquisition unit 150, It functions as the reward calculation unit 160, the reward condition setting unit 170, and the update unit 180. Further, the environment control device 100 corresponds to the first acquisition unit 110, the determination unit 120, the control unit 130, the display unit 140, the second acquisition unit 150, the reward calculation unit 160, the reward condition setting unit 170, and the update unit 180. May be implemented as one or more electronic circuits dedicated to

The 1st acquisition part 110 acquires input information containing action index information which shows a user's action. The first acquisition unit 110 acquires input information from the sensor 200. The first acquisition unit 110 may acquire the input information from the input device 300.

According to the control content determination rule, the determination unit 120 determines the control content for napping the user from the input information and the control content for controlling the environment control device 400 including the device that outputs light. The control content determination rule is stored in the storage unit 190.

The control content determination rule is represented by, for example, an action value function that determines the value of the control content. For example, the determining unit 120 determines that the control content that maximizes the reward from the input information (for example, the control content that obtains the maximum reward for the sleep-inducing effect) is the control content for the input information by using the action value function. decide. The control content that gives the highest reward may differ for each user. Therefore, the determination unit 120 may calculate the reward of the user for the control content according to the control content determination rule corresponding to the user. The action value function is an example of a value function.

When the environment control device 400 has another device (for example, a device other than the lighting device), the determining unit 120 determines the control content of the environment control device 400 including the other device.

The control unit 130 controls the environment control device 400 based on the control content determined by the determination unit 120. Specifically, the control unit 130 outputs a control signal corresponding to the control content to the environment control device 400.

The display unit 140 is a device that displays various types of information determined by the determination unit 120. The display unit 140 outputs, for example, at least one of the input information, the control content, and the information based on the evaluation result as an image. The image includes a photograph, a moving image, an illustration, characters, and the like. The display unit 140 is a liquid crystal display or the like. The display unit 140 is an example of an output unit.

In addition, the environment control device 100 may have a device that outputs a sound as an output unit instead of the display unit 140 or together with the display unit 140. Further, the environment control device 100 may include, as the display unit 140, a device such as a projector that displays information on an object (for example, a screen).

The second acquisition unit 150 acquires, from the input device 300, evaluation information indicating the evaluation of the user's nap for the control content. The evaluation information includes information indicating an evaluation of a result of controlling the environment control device 400 based on the control content determined by the determination unit 120, and includes, for example, information input by the user via the input device 300 after the environment control. .. For example, the second acquisition unit 150 acquires, from the input device 300, a value indicating at least one of the quality of the nap, the effect of introducing sleep, the recovery effect, and the comfort feeling input by the user. In addition, for example, the second acquisition unit 150 may acquire the evaluation information by receiving a voice signal from the input device 300 and detecting a speech of a predetermined keyword by voice recognition. Explaining comfort as an example, the predetermined keyword is a predetermined keyword indicating the comfort of the user with respect to a nap. For example, the predetermined keyword is “comfortable” or “uncomfortable”.

The reward calculation unit 160 calculates a reward based on the evaluation information acquired by the second acquisition unit 150. The evaluation information includes one or more item information, and the reward calculating unit 160 calculates the individual reward for each of the one or more item information and calculates the calculated one or more individual rewards to calculate the reward for the control content. To do. Details of the reward will be described later with reference to FIG. Note that the calculation is, for example, weighted addition, but is not limited to this. The process of calculating the reward will be described later with reference to FIGS. 7 to 9D.

The reward condition setting unit 170 sets conditions for calculating the reward in the reward calculating unit 160. The reward condition setting unit 170 sets weights for weighting addition of one or more individual rewards, for example. The condition may be stored in the storage unit 190 in advance, or may be set by the user via the input device 300. The reward condition setting unit 170 may change the condition (for example, weight) according to at least one of the weather, season, and time zone when the user takes a nap. When the weather, season, and time zone are included in the input information acquired by the first acquisition unit 110, for example, the reward condition setting unit 170 acquires information about the weather, season, and time zone from the first acquisition unit 110. You may. In addition, the reward condition setting unit 170 may acquire the information regarding the season and the time zone from a general-purpose timer IC (timer circuit) that measures the current time, which is included in the environment control device 100, or a real-time clock IC. Note that the reward condition setting unit 170 may not be provided when a fixed value is used as the weight and when the weighting is not performed.

The update unit 180 updates the control content determination rule by machine learning using a value based on the evaluation information acquired by the second acquisition unit 150 as a reward. The update unit 180 performs control learning adapted to the user (that is, the reward for the user is the most obtained by the reinforcement learning based on the reward for the control content determined by the determination unit 120 and calculated by the reward calculation unit 160). Learn the decision of control content). In the present embodiment, the updating unit 180 updates the control content determination rule by updating the action value function.

Q learning and TD learning are known as typical methods of reinforcement learning. Hereinafter, Q learning will be described as an example. The Q-learning is a method of learning the value Q(s, a) of selecting the control content a under the physical condition s of the user indicated by the input information, and the value Q(s, a The highest control content a of a) is selected as the optimum control content. The learning device 100a selects various control contents a under a certain physical state s, and a reward is given to the control contents a at that time. Thereby, the learning device 100a learns better selection of control content, that is, the correct value Q(s,a). Such an updating formula of the value Q(s, a) can be expressed by Formula 1, for example.

Here, s _t represents the state of the physical condition s at time t, a _t represents the control content at time t. The physical state changes to s _t+1 depending on the control content a _t . r _t+1 represents the reward obtained by the change in the physical condition. Further, the term with max is the Q value obtained by multiplying γ under the physical condition s _t+1 when the control content a _t+1 having the highest Q value known at that time is selected. Here, γ is a parameter of 0<γ≦1 and is called a discount rate. Further, α is a learning coefficient, and is set in a range of 0<α≦1.

Referring to FIG. 1 again, storage unit 190 stores the control content determined by determination unit 120. The storage unit 190 stores only the control content of the input information and the control content. The storage unit 190 stores, for example, the control content determined by the determination unit 120 without being associated with the input information corresponding to the control content. The storage unit 190 further stores the above-described control content determination rule and the like. The control content determination rule is an example of a learning result which is a learned result, for example.

[1-2. Explanation of input information]
Next, the input information used by the determination unit 120 and the control content determined by the determination unit 120 will be described.

First, input information output from at least one of the sensor 200 and the input device 300 will be described with reference to FIG. 2A. FIG. 2A is an example of input information input to the environment control device 100 according to the present embodiment. Specifically, this is an example of input information acquired by the first acquisition unit 110.

As shown in FIG. 2A, the input information includes “behavior index information” indicating the user's behavior, “subjective index information” indicating the result of self-evaluation in the physical and mental state of the user, and “schedule information indicating the user's schedule”. , "Use history information" indicating the history of the user's past naps, "environment information" indicating the current date and time and information on the space where the user is, and "user information" indicating the user's attributes.

The "behavior index information" includes the type speed at which the user operates the keyboard, the sleep time of the previous or previous main sleep, the quality of the main sleep, the wake-up and bedtime, the number of mouse clicks, the frequency of leaving the seat (toilet and hot water supply room). Etc.), the number of conversations and the state of conversation (for example, whether or not they are discussing, the speed of conversation such as talking slowly, and the number of remarks), rest time, drug intake history (whether or not there is sleep action (whether or not there is a sleep effect). For example, whether or not it is a sleeping pill, and whether or not it has a wakefulness effect (for example, whether or not caffeine is included), and the eating and drinking history including the previous day and the day (what was eaten, amount eaten, drinking, And including food and drink times). The “action index information” is acquired from at least one of the sensor 200 and the input device 300, for example.

By including at least “action index information” in the input information, the determination unit 120 can determine the control content according to the user's fatigue level and the like. For example, when the determination unit 120 determines from the “behavior index information” that the user is extremely tired and sleepy, the nap introduction period can be set short and the nap period can be set long. In other words, the action value function is learned so that the control content having a short nap introduction period and a long nap period has a high value, for example, when the user is very tired and the sleepiness is high.

The “subjective index information” includes the physical and mental condition and degree of the user. The “subjective index information” includes drowsiness, motivation, physical condition, degree of physical fatigue, degree of stress, concentration, performance, degree of tension or relaxation, degree of irritability, degree of anger or sadness, and the like. The “subjective index information” is acquired from the input device 300, for example.

By further including "subjective index information" in the input information, the determination unit 120 can determine the control content according to the state of mind and body (for example, the degree of fatigue) felt by the user himself.

The “schedule information” includes the usage date, the day before the usage day, and the user's activity schedule (meeting, etc.) on the day after the usage day, the work status (taking in, leaving, meeting, etc.) and the like. The “schedule information” is acquired from the input device 300. For example, the input device 300 acquires “schedule information” from a server that manages a schedule.

When the input information further includes the “schedule information”, the determining unit 120 determines the control content in which both the nap introduction period and the nap period are shortened when the user cannot take a long nap from the schedule content of the usage day. can do.

The "usage history information" includes a usage history of the user's past usage time, time, day of the week, timing (for example, timing at which a nap is easily taken), and data such as input information, output information, and reward. Etc. are included. The “usage history information” is acquired from the input device 300, for example. For example, the input device 300 may acquire the “usage history information” from the server device. The “usage history information” may be stored in the storage unit included in the environment control device 100.

Since the input information further includes “usage history information”, the determination unit 120 can determine the control content by referring to the past control content or the like for which the reward is high.

The “environmental information” includes weather (amount of solar radiation), amount of received light (light environmental information), season, time zone, environmental temperature, humidity, odor, sound environmental information, CO ₂ concentration, and the like. The environmental information is information on the environment in the partitioned space. Specifically, the “environmental information” may include, for example, information about the temperature, humidity, airflow, etc. in the office where the user works and in the nap room. The “environmental information” is acquired from at least one of the sensor 200 and the input device 300, for example.

By including the “environmental information” in the input information, the determining unit 120 can determine the control content so that the air conditioner is set to the heating setting in winter, for example.

The "user information" includes sex, age, race, place of origin, occupation, constitution (for example, photosensitivity), and medical history (for example, insomnia). The “user information” is obtained from the input device 300, for example. The input device 300 may acquire the “user information” by receiving an operation from the user, for example. Moreover, the “user information” may be stored in a storage unit included in the environment control device 100.

By including “user information” in the input information, the determination unit 120 can determine the control content more suitable for the user.

Next, the determination unit 120 that determines the control content will be described with reference to FIG. 2B. FIG. 2B is an example of output information output by the environment control device 100 according to the present embodiment.

As shown in FIG. 2B, the control content determined by the determination unit 120 includes the total environmental control time indicating the time for environmental control, the time of each phase of the nap (relaxation period, nap introduction period, nap period, etc.), light Fluctuation parameters (for example, cycle, MAX and MIN of brightness, light color, etc.), and at least one of the light control of the passage to the nap room and the rest room (light control according to refresh or nap). ..

In addition, the determination unit 120 may present the control content determined based on the input information to the user via the display unit 140. When the input information includes information other than the user's action index information (for example, schedule information), the determination unit 120 is suitable for the user based on the other information and the action index information. One or more control contents may be presented. For example, the determination unit 120 may present, as a recommendation, a control content for shortening the fluctuation period and prolonging the awakening when the user is a drowsiness person or when the user is in a drowsiness state. The determination unit 120 may present, for example, one or more control contents whose value is higher than a predetermined value. The presentation mode of the control content presented to the user by the determination unit 120 is not particularly limited. For example, it may be a character, a graph (see FIGS. 3 and 4), or a voice.

Further, when the first acquisition unit 110 has not acquired the input information, the determination unit 120 may cause the display unit 140 to display that effect when the operation to start the environmental control is acquired. In this case, the determination unit 120 may display the control content stored in the storage unit 190 on the display unit 140. At this time, the determination unit 120 can cause the user to more appropriately select the control content by displaying the plurality of control content.

Each phase of the nap is a period during which the nap environment control is being executed, and includes three periods of a relaxation period, a nap introduction period, and a nap period. The relaxation period is a period in which the user relaxes before going to sleep, or a period in which environment control suitable for relaxation is performed. The nap introduction period is a period in which the user is going to fall asleep or a period in which environmental control suitable for putting the user to sleep is performed. During the nap introduction period, the user is not sleeping yet. The nap period is a period in which the user is sleeping or a period in which environmental control suitable for sleeping is performed.

The determination unit 120 determines the control content in at least one of the above three periods. The determination unit 120 may determine the control content including, for example, the first control content in the nap introduction period and the second control content in the nap period after the nap introduction period. Further, the determining unit 120 may determine the control content in each of the above three periods. The determining unit 120 may determine, for example, at least one of the three periods to have a control content different from that of the other periods. In other words, the determining unit 120 may cause the environment control to be performed in at least one of the three periods, which is different from the other periods. For example, when the user is asleep and cannot sleep due to anxiety when the light is not lit, the determination unit 120 keeps the light lit so that the user falls asleep during the nap introduction period, and the user falls asleep. After setting, the control content including turning off the light may be determined so that the light does not disturb the sleep as a visual stimulus.

Here, the control content will be described in more detail with reference to FIGS. FIG. 3 is a first diagram for explaining an example of control contents determined by the environment control device 100 according to the present embodiment. FIG. 4 is a second diagram for explaining an example of control contents determined by the environment control device 100 according to the present embodiment. 3 and 4, the control content of the nap introduction period is mainly described among the relaxation period, the nap introduction period, and the nap period. 3 and 4, the control content of the brightness will be described.

As shown in FIG. 3, the control contents are the length of the entire nap introduction period, the maximum and minimum values of brightness, the rising period for increasing the brightness, the maintaining period for the maximum brightness, and the falling period for decreasing the brightness. , A minimum brightness maintenance period, a cycle, an ascending curve showing how to change the brightness (for example, slope), and a descending curve that shows how to change when the brightness is decreased (for example, slope) Including a curve. The maximum value and the minimum value mean the maximum value and the minimum value of the brightness when the brightness is changed periodically.

Further, at least one of the maximum value, the minimum value, various periods, and various curves may change over time. FIG. 4 shows an example in which the maximum value changes with time.

As shown in FIG. 4, the control content parameter further includes a change start time at which the change of the maximum value starts, a change end period at which the change of the maximum value ends, a change curve indicating how the maximum value changes, and It includes the change target value that indicates the maximum value after the change. When changing from the maximum value to the change target value, the determining unit 120 may determine the control content such that the change value does not suddenly change to the change target value, but gradually changes from the maximum value to the change target value. For example, when changing from the first maximum value to the second maximum value that is darker than the first maximum value, the determination unit 120 sets the brightness of the third maximum value between the first maximum value and the second maximum value to the maximum value. The control content is determined so as to include a cycle such as

The deciding unit 120 decides the above-mentioned 14 parameters so that the value of the control content with respect to the input information becomes highest, for example, by using the action value function.

Although FIGS. 3 and 4 show, as an example, control contents for periodically changing the brightness during the nap introduction period, the present invention is not limited to the periodical change. The control content may be, for example, to keep the brightness constant during the nap introduction period, or may be to decrease the brightness over time. Further, in the above description, the control content has been described as an example having 14 parameters, but the present invention is not limited to this. The parameter of the control content may be appropriately determined according to the user.

FIG. 5 is a third diagram for explaining an example of control contents determined by the environment control device 100 according to the present embodiment. FIG. 5 shows xy chromaticity coordinates used when the control content in the nap introduction period shown in FIGS. 3 and 4 is color control.

For example, the determining unit 120 sets one of the two points shown in FIG. 5 as the maximum value of chromaticity and the other point b as the minimum value of chromaticity, and determines the various periods and curves shown in FIG. Determined as control content. Note that color coordinates other than the xy chromaticity coordinates may be used.

Next, the reward for updating the action value function will be described with reference to FIG. FIG. 6 is an example of a reward input to the environment control device 100 according to the present embodiment.

As shown in FIG. 6, the reward is a value based on the evaluation information of each of the four items of “quality of nap”, “introduction effect of sleep fall”, “restoration effect of nap” and “comfort of nap environment” ( Hereinafter also referred to as individual reward). In addition, the reward may include at least one of the four individual rewards. Further, the individual reward is determined to have a higher value as the evaluation information shows a higher evaluation result, for example.

The individual reward in the “quality of nap” is calculated, for example, based on the number of awakenings and sleep efficiency of the user during the nap period. The number of awakenings during sleep and the sleep efficiency are examples of the second evaluation information for the “quality of nap”, and can be acquired by, for example, the sensor 200 measuring the body movement of the user during the nap period. For example, when the number of body movements during the nap period is small, the reward calculation unit 160 determines that the individual reward in the “quality of the nap” is high. The individual reward in the “quality of nap” indicates an evaluation value for the control content during the nap period.

The individual reward in the "sleep onset introduction effect" is calculated based on the sleep onset latency and the like. The time until falling asleep is an example of the first evaluation information for the “sleep falling induction effect”, and can be acquired by, for example, the sensor 200 measuring the body movement of the user. For example, when the number of body movements or the magnitude of body movements is equal to or less than a predetermined value, it may be determined that the user has started to take a nap (that is, the sleep onset period has started). The individual reward in the "sleep onset introduction effect" indicates an evaluation value for the control content during the nap introduction period.

The individual reward in the "restoration effect by nap" is calculated based on the physical condition of the user after waking up (for example, a behavior index and a subjective index). The physical condition of the user after waking up is an example of the third evaluation information for the “restoration effect by nap”, and can be acquired by, for example, the sensor 200 measuring the behavior of the user after waking up. For example, the evaluation information may include one of the action index information shown in FIG. 2A. The evaluation information may include, for example, type speed and the like. The third evaluation information may be a difference in the behavior index information of the user before and after a nap (for example, a difference in type speed). The “restorative effect by nap” indicates an evaluation value for the control content in the relaxation period, the sleep onset introduction period, and the nap period.

The individual reward in the "comfort of the sleep environment" is calculated based on the comfort of the sleep environment. Comfort is an example of the first evaluation information for “comfort in the sleep environment”, and can be acquired, for example, by the user operating the input device 300. “Comfort in the sleep environment” indicates an evaluation value for the control content during the nap introduction period.

The evaluation information includes first evaluation information based on the user's body movement during the nap introduction period, second evaluation information based on the user's body movement during the nap period after the nap introduction period, and before and after the nap introduction period. Third evaluation information based on the difference in the behavior index information of the user after the period.

The evaluation information on each of the above four items may be acquired from either the sensor 200 or the input device 300.

[2. Operation of environmental control device]
Next, the operation of the environment control device 100 configured as above will be described with reference to FIGS. 7 to 9D. FIG. 7 is a flowchart showing processing of the environment control device 100 according to this embodiment. This process is periodically executed, for example, at a predetermined time interval. Further, for example, this processing may be executed according to a predetermined time schedule, or may be executed based on an instruction from the user.

First, the 1st acquisition part 110 acquires input information containing action index information which shows a user's action (S110). Step S110 is an example of the first acquisition step. The determination unit 120 determines the control content based on the action value function (S120). The determination unit 120 determines the control content for the nap from the input information, which is the control content for controlling the environment control device 400 including the device that outputs light. The determination unit 120 determines the control content that maximizes the reward for the input information, based on the action value function. Step S120 is an example of a determination step. The control unit 130 controls the environment control device 400 based on the determined control content (S130). Thereby, the environmental control of the nap room where the user takes a nap is executed based on the determined control content. Step S130 is an example of a control step.

Then, the 2nd acquisition part 150 acquires the evaluation information which shows the evaluation in the user's nap with respect to the control content determined by the determination step (S140). The second acquisition unit 150 acquires, for example, the evaluation information of the four items shown in FIG. 5, for example, from either the sensor 200 or the input device 300. For example, when the input device 300 has a display, the second acquisition unit 150 acquires a value indicating the user's evaluation of the nap via the graphical user interface (GUI). The GUI may include a slider for inputting a value indicating a nap evaluation, a text box for directly inputting a numeric value, a numeric value increase/decrease button, or a combination thereof. May be included. Step S140 is an example of the second acquisition step.

Subsequently, the reward calculating unit 160 calculates the reward for the control content determined in step S120 based on the evaluation information (S150).

Then, the updating unit 180 updates the action value function by machine learning using the reward (S160). Step S160 is an example of an updating step.

The evaluation information may not be acquired every time the control content is determined. That is, step S140 may be skipped. In this case, the updating unit 180 may update the action value function using a predetermined value (for example, 0) as a reward.

The process of step S140 may be acquired during the control of the environmental control device 400, that is, during the environmental control with the control content determined in step S120. The update unit 180 may update the action value function during the environmental control based on the evaluation information acquired by the second acquisition unit 150 during the environmental control with the control content determined in step S120. In this case, the process may return to step S120 again, and the determination unit 120 may redetermine the control content for the input information during the environmental control according to the action value function updated during the environmental control.

The evaluation information acquired during the environmental control includes, for example, information indicating the evaluation of the “sleep-introduction effect”, and includes, for example, information about the body movement of the user detected by the sensor 200 during the sleep-introduction period.

Here, the process of calculating the reward from the evaluation information in the reward calculating unit 160 will be described with reference to FIGS. 8 to 9D. FIG. 8 is a flowchart showing a process of calculating a reward in environment control device 100 according to the present embodiment. Specifically, FIG. 8 is a diagram for explaining step S150 shown in FIG. 7 in detail.

As shown in FIG. 8, the reward calculation unit 160 determines the individual reward Fa for the quality of the nap (S210), the individual reward Fb for the sleep onset induction effect (S220), and the individual reward Fc for the recovery effect by the nap. Then, the individual reward Fd for the comfort of the sleep environment is determined (S230) (S240). Then, the reward calculating unit 160 determines the reward F for the control content based on the individual rewards Fa to Fd determined in steps S210 to S240 (S250). At this time, the reward calculation unit 160 may calculate the reward F by weighting and adding the individual rewards Fa to Fd. For example, the reward calculation unit 160 may calculate the reward F based on Expression 2.

F=w1×Fa+w2×Fb+w3×Fc+w4×Fd (Equation 2)

w1 to w4 are weights of items set by the reward condition setting unit 170, and are examples of reward conditions. That is, the reward condition setting unit 170 may set the weights w1 to w4 when the reward calculation unit 160 performs weighted addition.

Next, the processing of steps S210 to S240 will be described with reference to FIGS. 9A to 9D. 9A to 9D show a process of determining the individual reward of each item based on the evaluation information.

FIG. 9A is a flowchart showing a process of calculating the individual reward Fa for the quality of the nap in the environment control device 100 according to the present embodiment. FIG. 9A shows details of the process in step S210 shown in FIG.

As shown in FIG. 9A, the reward calculation unit 160 determines whether or not the quality of the nap is high based on the evaluation information (for example, the second evaluation information) (S211). The reward calculation unit 160 determines that the quality of the nap is high when, for example, the number of body movements of the user during the nap period is less than or equal to a predetermined number, or when the magnitude of the body movement is less than or equal to a predetermined value.

When the reward calculation unit 160 determines that the quality of the nap is high (Yes in S211), the individual reward Fa is set as the first reward value (S212). When the reward calculation unit 160 determines that the quality of the nap is not high (No in S211), it sets the individual reward Fa to a second reward value smaller than the first reward value (S213).

FIG. 9B is a flowchart showing a process of calculating the individual reward Fb with respect to the time until falling asleep in the environment control device 100 according to the present embodiment. FIG. 9B shows details of the process in step S220 shown in FIG.

As shown in FIG. 9B, the reward calculation unit 160 determines whether or not the time until falling asleep is short based on the evaluation information (for example, the first evaluation information) (S221). The reward calculation unit 160 determines that the time to sleep is short when the time to sleep is shorter than a predetermined time, for example. The predetermined time may be a preset value, a value set by the user, or a value set according to the time until falling asleep measured in the past by the user. May be. The predetermined time is, for example, 5 to 30 minutes. When the reward calculation unit 160 determines that the time until falling asleep is short (Yes in S221), the individual reward Fb is set as the third reward value (S222). When the reward calculation unit 160 determines that the time until falling asleep is not short (No in S221), the individual reward Fb is set to the fourth reward value smaller than the third reward value (S223). The time until falling asleep may be the period from the end of the relaxing period shown in FIG. 3 until the user actually falls asleep, or the environmental control system 10 starts the environmental control for a nap. It may be a period from when the user actually falls asleep.

FIG. 9C is a flowchart showing a process of calculating the individual reward Fc for the recovery effect in the environment control device 100 according to the present embodiment. FIG. 9C shows details of the process in step S230 shown in FIG.

As shown in FIG. 9C, the reward calculation unit 160 determines whether or not the recovery effect on the nap is high based on the evaluation information (for example, the third evaluation information) (S231). The reward calculation unit 160 makes the above determination, for example, by comparing the action index information before and after the nap. For example, when the behavior index information is the type speed, the reward calculation unit 160 determines that the recovery effect is high when the type speed after the nap is faster than the type speed before the nap. When the reward calculation unit 160 determines that the recovery effect is high (Yes in S231), the individual reward Fc is set as the fifth reward value (S232). When the reward calculation unit 160 determines that the recovery effect is not high (No in S231), the individual reward Fc is set to the sixth reward value smaller than the fifth reward value (S233).

FIG. 9D is a flowchart showing a process of calculating the individual reward Fd for comfort in the environment control device 100 according to the present embodiment. FIG. 9D shows details of the process in step S240 shown in FIG.

As illustrated in FIG. 9D, the reward calculation unit 160 determines whether or not the feeling of comfort is high based on the evaluation information (for example, the first evaluation information) (S241). The reward calculation unit 160 determines whether or not the comfort level is high, for example, based on the input to the input device 300 by the user. The reward calculation unit 160 determines that the user feels comfortable when he or she obtains the evaluation information indicating that the user determines that the feeling of comfort is high and that the degree of comfort is equal to or higher than a predetermined degree. When the reward calculating unit 160 determines that the feeling of comfort is high (Yes in S241), the individual reward Fd is set as the seventh reward value (S242). When the reward calculation unit 160 determines that the comfort level is not high (No in S241), the individual reward Fd is set to the eighth reward value smaller than the seventh reward value (S243).

Note that at least one of the second reward value, the fourth reward value, the sixth reward value, and the eighth reward value may be a negative value.

It should be noted that although FIGS. 9A to 9D have described the example of the process of selecting one of the two reward values, the reward value may be determined stepwise. Explaining the case of FIG. 9A as an example, a plurality of reward values may be set according to the number of body movements and the magnitude of the body movements. In this case, the smaller the number of body movements and the smaller the body movements, the larger the reward value is set.

[3. Effects, etc.]
As described above, the environment control system 10 according to the present embodiment obtains the input information including the action index information indicating the action of the user, and the action value function (an example of the control content determination rule). In accordance with the input information, the determining unit 120 determines the control content for the nap, which is the control content for the environment control equipment 400 including the device that outputs light, and the environment control equipment 400 based on the determined control content. A second acquisition unit 150 that acquires evaluation information indicating the user's evaluation of the control content, and an update unit 180 that updates the action value function by machine learning using a value based on the evaluation information as a reward. Equipped with.

Thereby, the environment control system 10 can determine the control content based on the action value function learned in association with the action index information and the evaluation information. Therefore, according to the environment control system 10, it is possible to perform nap support with a high nap effect. Further, since the control unit 130 executes the control of the environment control device 400 with the control content determined by the determination unit 120, it is possible to reduce the time and effort for the user to operate the environment control device 400.

In addition, the control content determination rule includes a value function that determines the value of the control content, and the updating unit 180 updates the value function.

Thereby, the environment control system 10 can build the control content determination rule suitable for the user by updating the action value function. As a result, the environment control system 10 can realize a good sleep support suitable for the user.

The environment control device 400 further includes a device other than a device that outputs light, and includes another device capable of controlling the environment for nap, and the determination unit 120 includes the environment control device 400 including the other device. Determine the control content to control the.

Thereby, the environment control system 10 can perform environment control using each of the device that outputs light and the other device. Therefore, the environment control system 10 can execute nap support with a higher nap effect.

In addition, the environment control system 10 further includes a reward calculating unit 160 that calculates a reward based on the evaluation information, and a reward condition setting unit 170 that sets conditions for calculating the reward in the reward calculating unit 160.

Thereby, the reward calculating unit 160 and the reward condition setting unit 170 appropriately calculate the reward. Further, for example, the reward condition setting unit 170 sets a condition corresponding to each user, so that the reward calculation unit 160 can effectively calculate the reward for learning the action value function suitable for each user. it can.

The control content includes the first control content in the nap introduction period and the second control content in the nap period after the nap introduction period. The evaluation information is first evaluation information based on the user's body movement during the nap introduction period, second evaluation information based on the user's body movement during the nap period, and first evaluation information based on the difference between the user's action index information before and after the nap. 3 Evaluation information is included. Then, the reward calculation unit 160 calculates the reward by weighting and adding the individual rewards Fa to Fd, and the reward condition setting unit 170 sets the weight w for the weighted addition as a condition. The individual rewards Fb and Fd are examples of values based on the first evaluation information, the individual rewards Fa are examples of values based on the second evaluation information, and the individual rewards Fc are values based on the third evaluation information. This is an example.

Thereby, the reward calculation unit 160 calculates the reward using the first evaluation information to the third evaluation information, so that the reward can be calculated more accurately. Further, for example, the reward condition setting unit 170 sets a suitable weight w according to each user or each season, so that the reward can be calculated more accurately.

The environment control system 10 further includes a storage unit 190 that stores the control content determined by the determination unit 120 without associating it with the input information.

As a result, the determining unit 120 can determine the control content by reading the control content (for example, the previous control content) stored in the storage unit 190 even when the input information cannot be acquired. it can. That is, the environmental control system 10 can perform environmental control using the control content determined by the determination unit 120 even when the input information cannot be acquired.

The environment control system 10 further includes a display unit 140 (an example of an output unit) that outputs at least one of the input information, the control content, and the information based on the evaluation information to the user.

Thereby, the environment control system 10 can notify the user of at least one of the input information, the control content, and the information based on the evaluation information. For example, by outputting the control content to the user before starting the environmental control, the user can take a nap after confirming the control content. For example, since the user can change the control content presented by the environment control system 10 before the environmental control is started, the user can perform the environmental control with the control content including the subjectivity of the user. ..

Further, the updating unit 180 updates the control content determination rule based on the evaluation information acquired by the second acquisition unit 150 while controlling the environment control device 400 with the control content. The determination unit 120 further determines the control content for the input information during control again according to the control content determination rule updated while controlling the environment control device 400.

Thereby, the environment control system 10 can update the control content determination rule and change the control content according to the state of the user who is performing the environmental control. For example, in a case where the user cannot fall asleep during the nap introduction period (for example, a state where a large number of body movements continue during the nap introduction period), the control during the sleep introduction period is performed according to the state. You can change the content. In other words, the environment control system 10 can determine more appropriate control content for the user's state at that time by learning during the environment control, and thus realize better sleep support suitable for the user. You can

Further, the environment control system 10 further includes a sensor 200 (an example of a detection unit) that detects a user's action. And the 1st acquisition part 110 acquires the information based on the action which the sensor 200 detected as action index information.

Accordingly, the environment control system 10 can provide the user with a nap support based on the action index information detected by the sensor 200 without acquiring the action index information from the outside.

The environment control system 10 further includes an input device 300 (an example of an input unit) that receives an operation for evaluating a nap from a user. And the 2nd acquisition part 150 acquires the information based on the user's operation which the input device 300 acquired as evaluation information.

Thereby, the environment control system 10 can update the action value function based on the information based on the user's operation acquired by the input device 300 without acquiring the evaluation information from the outside.

As described above, the environment control method of the environment control system 10 according to the present embodiment follows the first acquisition step (S110) of acquiring input information including the behavior index information indicating the behavior of the user and the control content determination rule. A determination step (S120) of determining the control content of the environment control device 400 including the device that outputs light, which is the control content for napping the user from the input information, and the environment based on the determined control content. A control step (S130) for controlling the control device 400, a second acquisition step (S140) for acquiring evaluation information indicating the user's evaluation of the control content, and a control content determination by machine learning using a value based on the evaluation information as a reward. And an updating step (S160) of updating the rule.

Thereby, the same effect as that of the environment control system 10 can be obtained.

(Modification 1 of Embodiment 1)
In this modification, another configuration of the environment control system will be described with reference to FIG. FIG. 10 is a block diagram showing the configuration of the environment control system 10a according to this modification. The environment control system 10a is different from the environment control system 10 of the first embodiment in the configuration of the environment control device 100c.

As shown in FIG. 10, the environment control system 10a includes an environment control device 100c and an environment control device 400. The environment control system 10a further includes a sensor 200 and an input device 300 in addition to the environment control device 100 of the first embodiment. In this case, the sensor 200 functions as a detection unit that is a part of the environment control device 100c. In addition, the input device 300 functions as an input unit that is a part of the environment control device 100c.

In this modification, for example, it is assumed that the user is sitting on a chair and working, and takes a nap while sitting, or a massage chair. In this case, the chair is provided with the environment control device 100c including the sensor 200 and the input device 300, so that the convenience is improved. For example, since the environment control device 100c is provided in the chair, the environment control device 100c appropriately determines the control contents even when the user moves to any place while sitting on the chair during a nap. can do. The environment control device 100c is not limited to being provided in the chair. Further, the environment control device 100c may be configured to further include the environment control device 400.

(Modification 2 of Embodiment 1)
Hereinafter, the environment control system according to this modification will be described with reference to FIGS. 11 and 12. Since the configuration of the environment control system is the same as that of the environment control system 10 of the first embodiment, the description thereof will be omitted.

In the first embodiment, the control content determination rule is described as an example represented by the action value function, but it may be represented by a neural network for estimating the value of each of the plurality of control content from the input information. Details of the neural network will be described with reference to FIG.

FIG. 11 is a conceptual diagram showing an example of a neural network in the environment control device according to the present embodiment.

As shown in FIG. 11, this neural network is a multi-layer artificial neural network, and is a mathematics for estimating the value Qai of a plurality of control contents ai (i=1 to n) in the physical state s indicated by the input information. It is a model.

FIG. 12 is a diagram showing a specific example of a plurality of control contents in this modification. Here, a case where the environment control device 400 is a lighting device will be described.

As shown in FIG. 12, the plurality of control contents a1 to an are a plurality of control content sets including a plurality of control contents of the lighting device. For example, the control contents a1 to an are the setting conditions of the 14 parameters shown in FIGS. 3 and 4. For example, the control content a1 is that the nap introduction period is "5 minutes", the maximum brightness of the lighting device is "3.5%", the minimum brightness of the lighting device is "0.1%", and the lighting device is It includes that the change target value of the maximum value of the brightness of is 2%. The value of such a plurality of control contents a1 to an in each physical state s is estimated by the neural network.

The operation in such an environment control system will be described below.

The first acquisition unit 110 acquires input information.

The determining unit 120 estimates the value of each control content from the input information based on the neural network, and determines the control content for nap based on the estimated value of the estimated value. For example, the determination unit 120 selects the control content having the highest value from the plurality of control content.

The control unit 130 executes environmental control based on the determined control content.

Then, the second acquisition unit 150 acquires the evaluation information.

The update unit 180 updates the value of the plurality of control contents based on the input information and the evaluation information. At this time, the value based on the evaluation information is used as the reward in the reinforcement learning.

Further, the updating unit 180 updates the parameter (for example, the weight w) of the neural network based on the updated value. That is, the updating unit 180 learns the parameters of the neural networks of a plurality of layers by inputting the updated value of each control content as a teacher signal. The updating unit 180 learns the determination of the control content adapted to the user by the reinforcement learning based on the value of the plurality of control content.

The so-called deep reinforcement learning is performed by internally repeating the process of updating the values of a plurality of control contents and updating the parameters of the neural network based on the updated values. The deep reinforcement learning is not particularly limited, and a conventional technique may be used. Therefore, detailed description of deep reinforcement learning is omitted.

(Embodiment 2)
In this embodiment, another configuration of the environment control system will be described with reference to FIG. FIG. 13 is a block diagram showing the configuration of the environment control system 10b according to this embodiment.

As shown in FIG. 13, the environment control system 10b according to the present embodiment includes a plurality of individual environment control systems 10c and a server apparatus 500 communicably connected to each of the plurality of individual environment control systems 10c.

Each of the plurality of individual environment control systems 10c has, for example, a configuration similar to that of the environment control system 10 of the first embodiment, but in FIG. 13, configurations other than the first acquisition unit 110 and the storage unit 190 are omitted. .. The number of the plurality of individual environment control systems 10c included in the environment control system 10b is not particularly limited.

The server apparatus 500 receives the input information acquired by the first acquisition unit 110 from each of the plurality of individual environment control systems 10c and the learning result stored in the storage unit 190 (for example, control content, reward, and updated). At least one of the action value functions) and centrally manage it. The server device 500 stores the acquired information in the storage unit 510. As a result, it is possible to share the learning result of each of the plurality of individual environment control systems 10c.

Note that at least one environment control device 100 among the plurality of environment control devices 100 included in each of the plurality of individual environment control systems 10c may function as a server device. That is, the environment control system 10b does not have to include the server device 500 separately from the individual environment control system 10c. In this case, each of the plurality of individual environment control systems 10c is communicably connected to each other and communicates at least one of the input information, the evaluation information, and the learning result with each other. Then, the updating unit 180 updates the action value function based on at least one of the input information acquired from the other individual environment control system 10c and the learning result.

As described above, the environment control system 10b according to the present embodiment includes the individual environment control system 10c including the first acquisition unit 110, the determination unit 120, the control unit 130, the second acquisition unit 150, and the update unit 180. Have multiple. Each of the plurality of individual environment control systems 10c is communicably connected to each other, and communicates at least one of the input information and the learning result with each other. Then, the updating unit 180 updates the control content determination rule based on at least one of the input information and the learning result acquired from the other individual environment control system 10c.

Thereby, the updating unit 180 can update the action value function of the own device based on the input information and the like acquired by the other individual environment control system 10c. Therefore, the accuracy of learning in the own device is improved, and a more appropriate action value function can be obtained.

The environment control system 10b includes a plurality of individual environment control systems 10c, and a server device 500 communicably connected to each of the plurality of individual environment control systems 10c.

Thereby, at least one of the input information, the evaluation information, and the learning result can be centrally managed. Further, when the server device 500 includes a learning unit, the action value function can be updated by machine learning based on the input information and the evaluation information acquired from each of the plurality of individual environment control systems 10c. .. Therefore, learning accuracy is improved, and a more appropriate action value function can be obtained. In this case, the server device 500 preferably includes a processor or the like that can perform processing at a higher speed than the environment control device 100.

(Other embodiments)
As described above, the environment control system and the environment control method according to one or more aspects of the present invention have been described based on the embodiments and modifications (hereinafter, also referred to as embodiments). However, the present invention is not limited to this embodiment and the like. As long as it does not depart from the gist of the present invention, various modifications that can be conceived by those skilled in the art may be applied to the present embodiment, or a configuration constructed by combining components in different embodiments may be one or more of the present invention. It may be included in the range of the aspect.

For example, in the above-described embodiments and the like, an example in which the determination unit determines one control content has been described, but the present invention is not limited to this. When the environmental control system includes two or more environmental control devices, the determination unit may control the two or more environmental control devices with one control content, or with the control content determined for each of the two or more environmental control devices. You may control. For example, when the environmental control system includes two or more lighting devices as the two or more environmental control devices, the two or more lighting devices may be controlled by one control content or may be controlled by different control content. Good.

Further, in the above-described embodiment and the like, an example in which the nap period is entered after the nap introduction period determined based on the input information is shown, but the present invention is not limited to this. For example, when the control unit determines from the action index information (for example, body movement) that the user slept during the nap introduction period, the control unit shifts to the nap period during the nap introduction period, and the environmental control device is controlled by the control content in the nap period. You may control.

Further, in the above-described embodiments and the like, an example in which the evaluation information is determined based on the action index information has been described, but the present invention is not limited to this. The evaluation information may be determined based on the physiological index information in addition to the behavior index information. Specifically, the first evaluation information may be generated based on the user's body movement and physiological index information during the nap introduction period. The second evaluation information may be generated based on the user's body movement and physiological index information during the nap period. Further, the third evaluation information may be generated based on the difference between the behavior index information of the user and the difference between the physiological index information before the nap introduction period and after the nap period. The physiological index information is an index relating to the electroencephalogram, autonomic nerves, and respiration of the user, and includes information such as heart rate, blink, pupil change, skin temperature, respiratory rate, and blood flow. The physiological index information is acquired using an electrode attached to the user's body, a biometric sensor, and a sensor such as a camera.

Further, in the above-described embodiments and the like, an example in which the first acquisition unit acquires the input information directly from the sensor has been described, but the present invention is not limited thereto. For example, the first acquisition unit may acquire the output signal of the sensor and acquire the input information by processing the acquired output signal. Obtaining the input information by processing the output signal by the first obtaining unit is included in obtaining the input information from the sensor. The same applies when the first acquisition unit acquires input information from the input device.

Further, in the above-described embodiments and the like, the example in which the second acquisition unit directly acquires the evaluation information from the input device has been described, but the present invention is not limited to this. For example, the second acquisition unit may acquire the output signal of the input device and process the acquired output signal to acquire the evaluation information. Acquiring the evaluation information by processing the output signal by the second acquisition unit is included in acquiring the evaluation information from the input device. The same applies when the second acquisition unit acquires the evaluation information from the sensor.

Further, the method of communication between the devices included in the environment control system according to the above-described embodiments and the like is not particularly limited. When wireless communication is performed between the devices, the wireless communication method (communication standard) is, for example, short-range wireless communication such as ZigBee (registered trademark), Bluetooth (registered trademark), or wireless LAN (Local Area Network). is there. Alternatively, the wireless communication method (communication standard) may be communication via a wide area communication network such as the Internet. Wired communication may be performed between the devices instead of wireless communication. The wired communication is, specifically, a communication using a power line communication (PLC) or a wired LAN.

Further, in the above-described embodiments and the like, each device included in the environment control system may be realized by a single device, or may be realized by a plurality of devices connected to each other. For example, the environment control device may be realized by cloud computing.

Further, a part or all of the constituent elements included in the environment control system in the above-described embodiments and the like may be configured by one system LSI (Large Scale Integration). For example, the environment control device may include a system LSI including a first acquisition unit, a determination unit, a control unit, a second acquisition unit, and a learning unit. The system LSI is a super-multifunctional LSI manufactured by integrating a plurality of components on one chip, and specifically, a microprocessor, a ROM (Read Only Memory), a RAM (Random Access Memory), and the like. It is a computer system configured to include. A computer program is stored in the ROM. The system LSI achieves its function by the microprocessor operating according to the computer program.

Although the system LSI is used here, it may also be called IC, LSI, super LSI, or ultra LSI depending on the degree of integration. Further, the method of circuit integration is not limited to LSI, and it may be realized by a dedicated circuit or a general-purpose processor. A programmable programmable gate array (FPGA) that can be programmed after the LSI is manufactured, or a reconfigurable processor that can reconfigure connection and setting of circuit cells inside the LSI may be used.

Furthermore, if integrated circuit technology comes out to replace LSI's as a result of the advancement of semiconductor technology or a derivative other technology, it is naturally also possible to carry out function block integration using this technology. The application of biotechnology is possible.

Further, one aspect of the present invention may be not only such an environment control system, but also an environment control method having a step of a characteristic component included in the environment control system. Further, one aspect of the present invention may be a computer program that causes a computer to execute the characteristic steps included in the environment control method. Further, one embodiment of the present invention may be a computer-readable non-transitory recording medium in which such a computer program is recorded.

Further, in the above-described embodiments and the like, each component may be configured by dedicated hardware, or may be realized by executing a software program suitable for each component. Each component may be realized by a program execution unit such as a CPU or a processor reading and executing a software program recorded in a recording medium such as a hard disk or a semiconductor memory. Here, the software that realizes the environment control system and the like of the above-described embodiments is the following program.

That is, this program is a first acquisition step of acquiring input information including action index information indicating a user's action in the computer, and the control content for the nap from the input information according to the control content determination rule, A decision step for deciding a control content for controlling an environmental control device including a device for outputting light, a control step for executing an environmental control by controlling the environmental control device based on the decided control content, and a control content An environment control method including a second acquisition step of acquiring evaluation information indicating an evaluation in a user's nap and an updating step of updating a control content determination rule by machine learning using a value based on the evaluation information as a reward is executed.

10, 10a, 10b environment control system 10c individual environment control system 100, 100c environment control device 110 first acquisition unit 120 determination unit 130 control unit 140 display unit (output unit)
150 Second Acquisition Unit 160 Reward Calculation Unit 170 Reward Condition Setting Unit 180 Update Unit 190 Storage Unit 200 Sensor (Detection Unit)
300 Input device (input section)
400 environment control device 500 server device

Claims (13)

A first acquisition unit that acquires input information including behavior index information indicating the behavior of the user;
According to a control content determination rule, from the input information, a determination unit for determining the control content of the environmental control device including the device that outputs light, which is the control content for napping the user,
A control unit for controlling the environment control device based on the determined control content;
A second acquisition unit that acquires evaluation information indicating the user's evaluation of the control content;
An environment control system comprising: an updating unit that updates the control content determination rule by machine learning using a value based on the evaluation information as a reward. The control content determination rule includes a value function that determines the value of the control content,
The environment control system according to claim 1, wherein the updating unit updates the value function. The environment control device is a device other than the device that outputs the light, including other devices capable of controlling the environment for the nap,
The environment control system according to claim 1, wherein the determination unit determines the control content of the environment control device including the other device. further,
A reward calculation unit that calculates the reward based on the evaluation information,
The environment control system according to any one of claims 1 to 3, further comprising: a reward condition setting unit that sets a condition in the calculation of the reward in the reward calculating unit. The control content includes a first control content in a nap introduction period and a second control content in a nap period after the nap introduction period,
The evaluation information includes first evaluation information based on the body movement of the user during the nap introduction period, second evaluation information based on the body movement of the user during the nap period, and before the nap introduction period and the nap period. The third evaluation information based on the difference of the action index information of the user after
The reward calculation unit calculates the reward by weighting and adding a value based on the first evaluation information, a value based on the second evaluation information, and a value based on the third evaluation information,
The environment control system according to claim 4, wherein the reward condition setting unit sets a weight when performing weighted addition as the condition. The environment control system according to claim 1, further comprising a storage unit that stores the control content determined by the determination unit without being associated with the input information. The environment control system according to claim 1, further comprising an output unit that outputs at least one of the input information, the control content, and information based on the evaluation information. The update unit updates the control content determination rule based on the evaluation information acquired by the second acquisition unit while controlling the environment control device with the control content,
The said determination part re-determines the control content with respect to the said input information during the said control again according to the said control content determination rule updated while controlling the said environmental control apparatus. Environmental control system. Furthermore, a detection unit for detecting the action of the user is provided,
The environment control system according to claim 1, wherein the first acquisition unit acquires information based on the action detected by the detection unit as the action index information. Furthermore, an input unit for accepting an operation for evaluating the nap from the user is provided,
The environment control system according to claim 1, wherein the second acquisition unit acquires, as the evaluation information, information based on the operation received by the input unit. A plurality of individual environment control systems having the first acquisition unit, the determination unit, the control unit, the second acquisition unit, and the update unit,
Each of the plurality of individual environment control systems are communicably connected to each other, and mutually communicate at least one of the input information and the learning result,
The update unit updates the control content determination rule based on at least one of the input information and the learning result acquired from another individual environment control system. Environmental control system. A plurality of individual environment control systems,
The environment control system according to claim 11, comprising a server device communicatively connected to each of the plurality of individual environment control systems. A first acquisition step of acquiring input information including behavior index information indicating the behavior of the user;
According to a control content determination rule, a determination step of determining, from the input information, the control content for napping the user, the control content of an environmental control device including a device that outputs light,
A control step of controlling the environment control device based on the determined control content;
A second acquisition step of acquiring evaluation information indicating the evaluation of the user with respect to the control content;
An updating step of updating the control content determination rule by machine learning using a value based on the evaluation information as a reward.

Images