JP5491891B2 - Device management apparatus and program - Google Patents

Description

  The present invention relates to a device management apparatus and a program for controlling a device in consideration of energy consumption in a house and user comfort.

  In recent years, HEMS (Home Energy Management System), which is a technology for reducing energy consumption (energy saving) in residential equipment, has attracted attention. In particular, air conditioners have been proposed not only to save energy but also to comfort, and to control the air conditioner so as to satisfy energy saving and comfort.

  PMV (predicted average report) is widely used as a comfort evaluation index for objectively evaluating comfort. In the technique described in Patent Document 1, it is described that PMV is used as an index of comfort due to operation of an air conditioner and that the comfort and energy saving are achieved by calculating the load of the air conditioner. In the technique described in Patent Document 1, a user inputs an amount of activity necessary for calculating PMV.

  In addition, in order to reduce the labor of input by the user, a configuration has been proposed in which the activity status of the user is detected by a human body information sensor (infrared sensor) and the amount of activity of the user is calculated (see Patent Document 2). ).

  On the other hand, in the field of BEMS (Building Energy Management System), which is a technology to save energy in facilities in large-scale buildings such as buildings, there is a demand for comfort by building users and a demand for energy saving by building managers. There has been proposed a technique for achieving both of the conflicting goals (see, for example, Patent Document 3). The technique described in Patent Literature 3 prevents the user from intentionally making excessive declarations by comparing the user's reported value with a comfort evaluation index (such as PMV). For example, a place that should be declared as “slightly hot” during cooling in summer is prevented from being overcooled by a user reporting it as “hot” or “very hot”.

JP-A-6-288595 Japanese Patent Laid-Open No. 6-333101 JP 2007-255835 A

  By the way, in the technology described in the above-mentioned patent document, energy saving and comfort are achieved at the same time, but energy consumption (for example, electricity charges) and comfort after the change of the operation state of the device It is not considered to determine the control state of the equipment after predicting the environmental change from both sides and evaluating the utility according to the user's values.

  The present invention has been made in view of the above reasons, and its purpose is to save energy and comfort so that the utility based on the user's values is maximized after the change of the operating state of the device. It is an object to provide a device management apparatus and a program that enable control to achieve both.

The invention of claim 1 is a device management apparatus that instructs the operation of a device that brings comfort to the user by consuming and operating a medium supplied from a supplier, and receiving the detection output of the sensor. A first means for estimating a degree of comfort representing the degree of the user's current comfort using the detection output of the sensor, and the current comfort of the user estimated by the first means The predicted value of the ex-post comfort level when the operating state of the device is changed based on the degree is calculated, and the subsequent consumption of the medium is predicted when the operating state of the device is changed based on the current operating state of the device. After the change of the operating state of the device according to the inference rule using the predicted value obtained for the comfort level and the consumption of the medium and the utility value given in advance based on the user's values Utility value Estimated, the utility value of the post determines the control content of the device which is the maximum, the determined control content and a second means for instructing a change of the operation state of the apparatus, said first means, said The sensor's detection output is used to express the objective comfort level as a comfort evaluation index, and the user's subjective comfort level is expressed as an evaluation value corresponding to the comfort evaluation index. It is characterized by estimating the current comfort level of the user from the sex evaluation index .

In the invention of claim 2, in the invention of claim 1, wherein the first means, using the reported values by the user as the evaluation value representing the degree of subjective comfort of the user, before Kikokoroyo qualification The present comfort level is expressed by a weighted sum of the index and the reported value.

According to a third aspect of the present invention, in the second aspect of the present invention, the first means is capable of changing a weighting factor used for calculating a weighted sum.

According to a fourth aspect of the present invention, in the second or third aspect of the present invention, the first means uses the latest operation on the device performed by the user based on the subjective comfort declaration value based on the user declaration. It is estimated from the contents of

The inventions of claim 5, comprising a device management apparatus for instructing the operation of the device that provides comfort to the user by operating by consuming medium supplied from the supply company, the detection output of the sensor input A first means for estimating a degree of comfort that represents a degree of a user's current comfort using a detection output of the sensor, and a current state of the user estimated by the first means A predicted value of the subsequent comfort level when the operating state of the device is changed based on the comfort level is obtained, and the subsequent consumption of the medium when the operating state of the device is changed based on the current operating state of the device. Afterward when the operating state of a device is changed in accordance with an inference rule that uses a prediction value obtained for the comfort level and consumption of the medium and a utility value given in advance based on the user's values. Utility value Estimated by, and a second means for utility value of the posterior determines the control content of the device which is the maximum, indicating the change of the operation state of the apparatus by the determined control content, the device is an air conditioner The first means has a function of estimating a current comfort level by adjusting the comfort evaluation index according to an elapsed time after entering the room.

The inventions of claim 6, comprising a device management apparatus for instructing the operation of the device that provides comfort to the user by operating by consuming medium supplied from the supply company, the detection output of the sensor input A first means for estimating a degree of comfort that represents a degree of a user's current comfort using a detection output of the sensor, and a current state of the user estimated by the first means A predicted value of the subsequent comfort level when the operating state of the device is changed based on the comfort level is obtained, and the subsequent consumption of the medium when the operating state of the device is changed based on the current operating state of the device. Afterward when the operating state of a device is changed in accordance with an inference rule that uses a prediction value obtained for the comfort level and consumption of the medium and a utility value given in advance based on the user's values. Utility value Estimated by, and a second means for utility value of the posterior determines the control content of the device which is the maximum, instructing a change of the operating state of the device by the determined control content, the device is an air conditioner said first means estimates the comfort of current by a Subscriber is adjusted according to the elapsed time after entering the comfort evaluation index calculated to be carried out an operation on the device, the user Has a function of estimating the current comfort level from the content of the operation when the device is operated.

According to a seventh aspect of the present invention, in the sixth aspect of the present invention, the first means includes a comfort evaluation index obtained by using the actual measured value of the current room temperature when the user performs an operation on the device and the utilization. The difference from the comfort evaluation index obtained using the set temperature required by the user is used as a correction value, and the value obtained by adding the correction value to the comfort evaluation index obtained using the actual measured value of the current room temperature is the current comfort It is used as a degree.

In the invention of claim 8, in the invention of any one of claims 5 to 7 , the first means adjusts the comfort evaluation index according to an elapsed time after entering the room, thereby allowing a predetermined time later. It has a function of estimating the comfort level.

In the invention of claim 9, in the invention of any one of claims 1 to 8 , the second means has a function of calculating a utility value specific to the mode according to the mode selected by the user, and as a mode, For events that cannot achieve both reduction in medium consumption and comfort, comfort-oriented mode with high utility value assigned to comfort, energy-saving priority mode with high utility value assigned to medium consumption reduction, and medium It is possible to select a moderate mode in which utility values are evenly assigned to the reduction of consumption and comfort.

In the invention of claim 10, in the invention of any one of claims 1 to 9 , when the second means changes the user's current comfort level estimated by the first means and the operation state of the device And the predicted value of the post-comfort level as a conditional probability table, and the predicted value of the consumption amount of the medium after the change of the current operating state of the device and the operating state of the device. It is characterized by calculating the subsequent utility value when the operating state of the device is changed according to the inference rule using the two conditional probability tables and the utility value given in advance based on the user's values To do.

According to an eleventh aspect of the present invention, in the invention according to any one of the first to tenth aspects, the sensor detects a user's activity in a non-contact manner with the user, and the first means determines the amount of activity of the user. A conditional probability table statistically determined using the relationship between the user's activity and the measured value of the sensor's detection output for the corresponding sensor's detection output is provided, and the amount of activity is reflected in objective comfort It is characterized by that.

According to a twelfth aspect of the present invention, in any one of the first to eleventh aspects, the first means and the second means are preliminarily assumed as the detection output of the sensor and the device. The control content that maximizes the utility value is obtained by giving the operating state and the assumed operation of the device as conditions, and the combination of the condition and control content is stored as a rule table, which is used for the actual detection output of the sensor. On the other hand, the control content is determined by applying the rule table.

According to a thirteenth aspect of the present invention, in the twelfth aspect of the present invention, the rule table is provided in a remote control device that instructs an operation of the device.

In the invention of claim 14 , in any one of claims 1 to 13 , the first means and the second means communicate with the sensor installed in a house via a wide area information communication network. A control content of the device provided in the server and determined by the second means is instructed via a wide area information communication network.

The invention of claim 15 is an equipment management device for instructing an operation of an air conditioner as equipment that operates by consuming electric power supplied from a supplier, and includes a sensor input unit to which a detection output of a sensor is inputted. A first means for estimating a comfort level representing a degree of a user's current comfort using a detection output of the sensor, and air conditioning based on the current comfort level of the user estimated by the first means Find the predicted value of the ex-post comfort when the operating state of the device is changed, and the predicted value of the ex-post electricity bill when the operating state of the air-conditioning unit is changed based on the current operating state of the air-conditioning unit. Estimate the subsequent utility value when the operating state of the air conditioner is changed according to the inference rule using the predicted value obtained for the degree of electricity and the electricity price and the utility value given in advance based on the user's values, after Determining the control content of the air conditioner utility value is maximum, and a second means for instructing a change of the operation state of the air conditioner by the determined control content, the first means, the detection output of the sensor Is used to express the objective comfort level as a comfort evaluation index, and the user's subjective comfort level is expressed as an evaluation value corresponding to the comfort evaluation index. The current comfort level of the user is estimated from the above.

The invention of claim 16 is a program for causing a computer to function as a device management apparatus that instructs the operation of a device that brings comfort to the user by operating by consuming a medium supplied from a supplier. A first means for estimating a comfort level indicating a degree of a user's current comfort using a detection output of a sensor input to the computer, and a user's current comfort level estimated by the first means The predicted value of the ex-post comfort level when the operating state of the device is changed based on the device, and the predicted value of the subsequent consumption of the medium when the operating state of the device is changed based on the current operating state of the device Change the operating state of the device according to the inference rules using the predicted value for the comfort level and consumption of the medium and the utility value given in advance based on the user's values A post-utility value is estimated, the control content of the device having the maximum post-utility value is determined, and the computer functions as a second means for instructing a change in the operating state of the device according to the determined control content The first means uses the detection output of the sensor to represent the objective comfort level as a comfort evaluation index, and the user's subjective comfort level corresponds to the comfort evaluation index. The user's current comfort level is estimated from the value and the comfort evaluation index .

According to the configuration of the inventions of claims 1 and 16, the predicted values relating to the subsequent comfort level and the medium consumption amount when the operation state of the device is changed are obtained, and these predicted values and the user's values are obtained. Since the control content is determined so that the subsequent utility value is maximized in accordance with the inference rules based on it, it is possible to achieve highly satisfactory control that can be said to achieve both energy saving and comfort for the user. In addition, since the control content that maximizes the utility value is given to the device in response to the detection output of the sensor, the user can operate the device in the optimum operating state according to the user's values without instructing the operating state of the device. Will be controlled.

Further, according to the configuration of the invention of claim 1 , since the current comfort is evaluated in consideration of the objective comfort and the subjective comfort, the objective comfort evaluation by the sensor is performed. In addition to responding to the user's steady-state activity, control that also accommodates the user's unsteady-state activity is possible by incorporating the user's subjective comfort.

According to the configuration of the invention of claim 2 , since the degree of objective comfort and the degree of subjective comfort are quantified and integrated by weighted sum, the objectivity and use detected by the sensor regarding comfort are used. It is possible to evaluate the comfort combined with the subjectivity of the reporter.

As in the configuration of the invention of claim 3 , by making it possible to change the weighting coefficient used for calculating the weighted sum, the priority degree between the objective comfort and the subjective comfort can be adjusted.

According to the configuration of the invention of claim 4 , since the user's reported value is estimated by the latest operation performed by the user on the device, the user only operates the device without special consciousness. It is possible to determine the optimal control content.

According to the configuration of the invention of claim 5 , since the comfort evaluation index is adjusted according to the elapsed time after entering the room where the air conditioning apparatus is cooling and heating, the user sets the set temperature for the device. Even if the operation to change is not performed, the set temperature can be automatically adjusted so that the comfort is maintained regardless of the elapsed time from entering the room.

According to the configuration of the invention of claim 6, estimates the comfort of current by adjusting the comfort evaluation index in accordance with the elapsed time after entering the room heating and cooling by the air conditioner is being performed, the equipment When the operation to change the set temperature is performed, the current comfort level is estimated according to the operation, so the user can feel the subjective feeling while automatically adjusting the set temperature according to the comfort evaluation index. The set temperature can be adjusted accordingly.

According to the configuration of the seventh aspect of the invention, the evaluation value representing the degree of comfort corresponding to the set temperature is obtained by the comfort evaluation index obtained using the measured value of the room temperature and the set temperature requested by the user. Therefore, in calculating the evaluation value indicating the degree of comfort corresponding to the set temperature, it is possible to use the same procedure as the comfort evaluation index determined from the detection output of the sensor.

According to the configuration of the eighth aspect of the invention, since the elapsed time from entering the room is taken into account when estimating the comfort level after a predetermined time, the set temperature can be adjusted more appropriately.

According to the configuration of the ninth aspect of the invention, for the event where the reduction of the medium consumption and the comfort cannot be achieved at the same time, priority is given to either the reduction of the medium consumption or the comfort according to the user's values. It becomes possible to select whether to do.

According to the configuration of the invention of claim 10 , since the utility value is calculated stochastically by applying the conditional probability table to the consumption and comfort of the medium, the utility value can be calculated by combining complicated conditions. become.

According to the configuration of the invention of claim 11 , since the activity amount is estimated by the detection output of the sensor that is not in contact with the user, it is not necessary to use a special measuring means for measuring the activity amount, It becomes possible to detect the amount of activity without imposing a burden. Moreover, since the conditional probability table is statistically created based on the actual measurement result of the sensor, the conditional probability table can be created so that the accuracy in estimating the activity amount from the detection output of the sensor becomes high. .

According to the configuration of the twelfth aspect of the present invention, a combination of the expected detection output of the sensor and the control content that maximizes the utility value is created as a rule table, and this rule table is used as the actual detection output of the sensor. Since the control content is determined by application, the processing load can be greatly reduced compared to the case where the control content is determined by applying an inference rule every time the sensor output is obtained. Even with a processing capability of a degree, it is possible to perform control that achieves both comfort after changing the operating state of the device and reduction of the consumption of the medium.

According to the configuration of the thirteenth aspect of the present invention, since the rule table is provided in the remote control device provided in the device, the control content after the operation of the remote control device maximizes the utility value based on the user's values. Therefore, control with a high degree of satisfaction for the user is performed.

According to the configuration of the invention of claim 14 , since the control content is determined in the server that communicates with the sensor provided in the house via the wide area information communication network, the calculation for realizing the first means and the second means is performed. By providing the server with resources, the device does not require large computing resources, and it is possible to cope with equipment with a small processing capacity on the house side. In other words, since a device with a large calculation resource is not required when introducing the device, the user can enjoy the above functions at low cost.

According to the configuration of the invention of claim 15 , the apparatus is an air conditioner , and after the operation state is changed, there is a time difference until the user realizes the change, but the post-utility when the operation state is changed Since the control content is determined so as to maximize the value, the satisfaction of the user after the change of the control content can be increased. Further, since the utility value is evaluated based on the comfort and the electricity bill, the value of the user's sense of money is reflected in the utility value, and the meaning of the utility value becomes easy to understand.

1 is a block diagram illustrating a first embodiment. It is operation | movement explanatory drawing of the process part used for the same as the above. It is a schematic block diagram same as the above. It is a block diagram which shows the other form same as the above. FIG. 6 is a block diagram illustrating a second embodiment. It is operation | movement explanatory drawing of the process part used for the same as the above. (A) is a figure which shows the reported value of thermal sensation, (b) is a figure which shows the example of a function which approximated the reported value of thermal sensation. It is a figure which shows the report value of the thermal sensation verified by experiment. It is principal part operation explanatory drawing of FIG. It is operation | movement explanatory drawing which abbreviate | omitted a part of FIG. It is a figure which shows an experimental result regarding power consumption. It is a figure which shows an experimental result regarding the report value of a thermal sensation.

  In the following description, it is assumed that a communication network is prepared in a house as a facility, and a behavior pattern in a user's daily life is detected by a sensor, and the detected data is accumulated and analyzed. The present invention realizes the following functions by applying intelligent system technology for data accumulation and analysis to the HEMS field.

(Embodiment 1)
In this embodiment, the electricity supplied by the power company is exemplified as the medium supplied from the supplier, but the technical idea of the present invention is applied to other media such as gas, water, and heat. Can do.

  In the present embodiment, as shown in FIG. 3, a plurality of types (five types in the illustrated example) of sensors 21 to 25, a plurality of types (three types in the illustrated example) of devices 31, information input by the user, and a user A user interface 32 for presenting information to the device and a device management apparatus 10 for instructing the operation of the device 31. The device 31 may be anything that consumes at least one of the above-described media and creates a comfortable space, and there is no particular limitation on the environment in which the device 31 is used. That is, the technical idea of the present invention can be applied to equipment 31 used in stores, buildings, and factories as well as equipment 31 used in houses.

  When the device management apparatus 10 instructs the operation of the device 31, either a method of giving the instruction content to the device 31 as a control signal or a method of presenting the instruction content to the user through the user interface 32 is adopted. Below, the case where an instruction content is presented to the user through the user interface 32 will be described as an example. Specifically, the user interface 32 capable of screen display may be used to display the instruction content on the screen. Alternatively, the instruction content may be notified by voice using the user interface 32 capable of voice output.

  In addition, as equipment that brings convenience to the user by operating while consuming electricity as a medium, as shown in FIG. 3, an air conditioner (hereinafter referred to as “air conditioner”) 31a, a television receiver, and the like. There are various types such as 31b, lighting fixture 31c, water heater 31d, floor heating device 31e, etc., but in the following description, a case where an instruction for performing air conditioning by the air conditioner 31a is output from the device management device 10 will be described as an example.

  In the example shown in FIG. 3, as the sensors 21 to 25, a human sensor 21 and a sound sensor 22 for detecting the presence or absence of a person in a room where air conditioning by the air conditioner 31 a is effective, a temperature sensor 23 for measuring the indoor environment, and A humidity sensor 24 and a power amount sensor 25 for measuring the power consumption of the air conditioner 31a are provided.

  The human sensor 21 uses a pyroelectric infrared sensor, and outputs a square wave signal binarized according to the magnitude relationship between the output level of the pyroelectric infrared sensor and a specified threshold value. Therefore, the more the amount of human activity (or metabolic rate) detected by the human sensor 21, the greater the number of times the square wave signal rises (ON).

  The sound sensor 22 is provided to detect sounds such as sounds that interact with the environment (physical sounds, footsteps) and breathing sounds that are caused by the presence of a person. In particular, in the human sensor 21 using a pyroelectric infrared sensor, a square wave signal does not rise if a person is stationary even if a person is present in the detection region. The presence / absence of a person is determined by detecting an interaction sound with the environment (sounds, footsteps) or a person's breathing sound. Also for the sound detected by the sound sensor 22, a binarized square wave signal is generated in accordance with the magnitude relationship between the sound pressure level and the prescribed threshold value. As with the output of the human sensor 11, the output from the sound sensor 22 increases the number of times the square wave signal rises (ON) as the human activity (or metabolism) increases.

  The temperature sensor 23 has a known structure that includes a temperature sensitive element such as a thermistor and detects the room temperature, and the humidity sensor 24 also has a known structure that detects the relative humidity. Moreover, in order to detect the electric power consumption of only the air conditioner 31a, the electric energy sensor 25 uses what was integrated in the outlet which connects the air conditioner 31a. This outlet does not necessarily need to be a wall outlet, and may be an adapter type outlet connected to the wall outlet.

  In the above-described configuration, the device management apparatus 10, the sensors 21 to 25, and the equipment 31 can be directly connected. However, in this embodiment, a communication network using communication technology is provided in a house, and the sensor Assume that 21 to 25 and the facility 31 can communicate with the device management apparatus 10 via a communication network.

  Below, the apparatus management apparatus 10 is demonstrated concretely. The device management apparatus 10 enables the following operations by executing an appropriate program on a computer, and includes a sensor input unit 11 as an interface to which the sensors 21 to 25 are connected as shown in FIG. In addition, an input / output interface 12 for connecting the user interface 32 is provided.

  The sensor input unit 11 extracts the sensor output from the outputs of the sensors 21 to 25 at an appropriately set update time (for example, 5 minutes) and gives the extracted sensor output to the processing unit 13 described later. That is, the sensor input unit 11 outputs time-series data of detection outputs from the sensors 21 to 25. The sensor input unit 11 extracts the number of times of ON for each update time from the outputs of the human sensor 21 and the sound sensor 22, and compares the number of times of ON with a threshold value, thereby providing three types of “many”, “medium”, and “small” The output value selected from is supplied to the processing unit 13. The amount of human activity detected by the human sensor 21 and the sound sensor 22 is classified into three stages and given to the processing unit 13.

  In the sensor input unit 11, for the temperature sensor 23 and the humidity sensor 24, the sensor output for each update time described above is given to the processing unit 13. Alternatively, an average value of sensor outputs extracted a plurality of times within the update time is given to the processing unit 13. By adopting a configuration in which the sensor output is extracted a plurality of times within the update time and the average value is given to the processing unit 13 as in the latter case, it is possible to prevent an abnormal value of the sensor output from being input to the processing unit 13. Can do.

  The output of the electric energy sensor 25 may be handled in the same manner as the outputs of the temperature sensor 23 and the humidity sensor 24. However, the output of the electric energy sensor 25 is integrated at the sensor input unit 11, and the integrated value at the update time is input to the processing unit 13 instead of the instantaneous value for the power consumption of the air conditioner 31a.

  By the way, in the present embodiment, the utility value under the condition of interest is maximized in consideration of the comfort level indicating the degree of user comfort associated with the use of the air conditioner 31a and the power consumption associated with the use of the air conditioner 31a. The purpose is to determine the operating state of the air conditioner 31a. In the present embodiment, an example in which the user gives an instruction using the user interface 32 will be described as the focused condition, but the condition may be set in the processing unit 13 in advance.

  The utility value in the present embodiment is determined by three factors: the condition that the user is paying attention to, the comfort level for the user, and the power consumption in the air conditioner 31a.

  The condition that the user pays attention to is given from the user interface 32 to the mode selection unit 14 through the input / output interface 12, and the mode selection unit 14 gives the condition instructed from the user interface 32 to the processing unit 13 as a mode. In this embodiment, three types of modes of “comfort-oriented mode”, “energy-saving-oriented mode”, and “medium mode” can be selected. These modes are selected based on the user's values.

  The power consumption of the three elements used for calculating the utility value is generally estimated using a plurality of types of parameters such as the indoor volume, the operating state of the air conditioner 31a, the cooling capacity of the air conditioner 31a, and the indoor temperature. However, in this embodiment, by designating the operating state of the air conditioner 31a, the subsequent operation when the operating state of the air conditioner 31a is changed using only the relationship between the current operating state and the designated operating state. Assume that power consumption can be predicted.

  For the degree of comfort that represents the degree of comfort for the user, the operating status of the air conditioner 31a is changed by combining the comfort rating index (PMV = Predicted Mean Vote) and the information that the user has reported about comfort / discomfort. The current comfort level before the estimation is estimated, and the subsequent comfort level when the operating state of the air conditioner 31a is changed is predicted based on the estimated current comfort level.

  The calculation of the utility value is performed by the processing unit 13, and the processing unit 13 estimates the subsequent utility value when the operating state of the air conditioner 31a is changed based on the predicted value of the comfort level and the power consumption. A calculation unit 15 is provided.

  As described above, the subsequent power consumption in the air conditioner 31a can be predicted based on the relationship between the current operation state and the designated operation state, and a condition to be focused on for obtaining the utility value is instructed from the user interface 32. Therefore, the condition for predicting the subsequent power consumption and the condition to be focused on for obtaining the utility value are instructed from the user interface 32.

  The processing unit 13 also includes a comfort level estimation unit 16 that estimates the current level of comfort for the user based on the comfort evaluation index and the user's report, information for calculating the comfort evaluation index, and an air conditioner. And an activity amount estimation unit 17 that estimates an activity amount (metabolic amount) of the user as information for changing the operation of 31a.

  In this embodiment, factors that are related to each other are expressed by nodes in the influence diagram, a conditional probability table is set in the nodes in the influence diagram, and the probability that a probabilistic inference rule is applied to given data A configuration for performing propagation calculation is adopted. The utility value calculation unit 15, the comfort level estimation unit 16, and the activity amount estimation unit 17 constituting the processing unit 13 can be expressed as shown in FIG. 2 using an influence diagram.

  The influence diagram is expressed by nodes represented by three types of figures, a rectangle, a circle (ellipse), and a rhombus, and links represented by solid or broken arrows connecting the nodes.

  The rectangular node is a decision node and represents an item for selecting an action. In the present embodiment, the node represents an instruction related to the operating state of the air conditioner 31a. The elliptical node is an opportunity node, which represents an uncertain event that cannot be controlled by user decision making, and includes a conditional probability table for representing the uncertain event. However, in the illustrated example, an elliptical node surrounded by a double line indicates a node from which a unique output can be obtained by substituting an input value into a mathematical expression.

  In the example shown in the figure, elliptical nodes surrounded by double lines are provided, and these nodes do not have their output values stochastically determined for their input values, but are uniquely output values for their input values. Means an opportunity node. The diamond-shaped node is a value node and calculates a utility value for the input based on the value for the user. In the present embodiment, “comfort-oriented mode”, “energy-saving-oriented mode”, and “medium mode” can be selected according to the user's values, and the utility value calculation unit 15 is specific to the mode depending on the selected mode. The utility value is calculated.

  The solid line link is a normal link, and represents that the node connecting the start ends of the arrows affects the node connecting the end ends of the arrows. A broken line link is a notification link, and indicates that information generated in a node connected to the start end of the arrow is notified to the node connected to the end of the arrow.

  By the way, the utility value calculation unit 15 provided in the processing unit 13 evaluates by replacing the power consumption with the electricity bill. As shown in FIG. 2, when the utility value calculation unit 15 is represented by an influence diagram, the value node N35 for calculating the utility value includes an opportunity node N33 regarding the comfort level for the next 5 minutes, and an electricity bill for the next 5 minutes. Is connected to the opportunity node N34. The opportunity node N33 related to the comfort level for 5 minutes is connected to the opportunity node N26 related to the current comfort level and a decision node N31 representing the operation content of the air conditioner 31a. Further, a decision node N31 representing the operation content of the air conditioner 31a and an opportunity node N32 related to the current operating state of the air conditioner 31a are connected to the opportunity node N34 related to the electricity bill for 5 minutes. Here, “the next 5 minutes” means the subsequent 5 minutes when the operating state of the air conditioner 31a is changed.

  First, the relationship between the opportunity nodes N33 and N34 and the value node N35 will be described. The “comfort level (for 5 minutes)” that is an output variable of the opportunity node N33 is selected from four values of {become better, stay better, stay worse, get worse}. Further, the “electric charge (for 5 minutes)” that is an output variable of the opportunity node N34 is selected from three values of {high, cheap, zero}.

  In the value node N35, utility values of 0 to 100 are assigned in advance to all combinations of output values of the two opportunity nodes N33 and N34. However, in the value node N35, different utility values are assigned in advance as shown in Tables 1 to 3 according to the “comfort-oriented mode”, “energy saving-oriented mode”, and “medium mode” selected by the user interface 32. It is.

  The utility value of each mode is given based on the user's values, and the user selects which mode's utility value is used. In other words, the “comfort-oriented mode” is provided for users who prioritize comfort even if the electricity price increases, and “energy saving” is provided for users who prioritize reduction of electricity charges even at the expense of comfort. The “important mode” is provided. In addition, a user who needs only moderate comfort and reduced electricity charges may select the “medium mode”.

  By the way, about other combinations of comfort and an electricity bill, since utility values differ according to a user's values, utility values differ in each of Tables 1 to 3.

  However, regardless of which of Tables 1 to 3 is adopted, the comfort is either (becomes better) or remains good, and the combination that the electricity rate is (zero) is the most preferable for the user and is comfortable It can be said that the combination of either (because it is bad) or (becomes worse) and the electricity price is (high) is the most unfavorable for the user.

  Therefore, in any of Tables 1 to 3, the utility value for the former combination is set to 100 (best), and the utility value for the latter combination is set to 0 (worst). In addition, if the electricity bill is (cheap), it goes without saying that comfort (becomes better) (becomes better) takes precedence over (becomes bad) (becomes worse). However, how much the difference between utility values is set depends on the user's values (ie, the selected mode).

  When it is impossible to achieve both reductions in electricity charges and comfort (that is, electricity charges increase if comfort is pursued and comfort decreases if electricity charges are pursued), user values Accordingly, the utility value is assigned so that one of the reduction of the electricity bill and the comfort is prioritized.

  In the comfort-oriented mode shown in Table 1, the comfort level is given priority over the reduction of the electricity price, so even if the electricity price is (high), if the comfort level is (improves) (stays good), it is relatively high. Utility values are assigned. On the other hand, even if the electricity rate is (zero), a relatively low utility value is assigned if the comfort level is (bad) (becomes worse).

  In the energy saving priority mode shown in Table 2, priority is given to reducing electricity charges over comfort levels, so if the electricity price is (zero), even if the comfort level is (bad) (becomes worse) A high utility value is assigned to, and conversely, a relatively low utility value is assigned if the electricity rate is (high) even if the comfort level is (becomes better) (still remains good).

  In the case of the moderate mode shown in Table 3, the reduction in electricity charges and the comfort level are handled to the same extent. Therefore, when the comfort level is (becomes better) (while it is good) and the electricity rate is (high), and when the comfort level is (becomes bad) (becomes worse) and the electricity rate is (zero) The same utility value is assigned. That is, utility values are assigned equally.

  By the way, the electricity bill for the next 5 minutes is predicted by the relationship between the current operating state of the air conditioner 31a and the future operating state. That is, the opportunity node N34 is predicted by a combination of output values of the decision node N31 that determines the operation of the air conditioner 31a and the opportunity node N32 that is the latest operating state of the air conditioner 31a. The output value of the decision node N31 is selected from five values {+1, -1, keep, on, off}, and the output value of the opportunity node N32 is selected from four values {high, normal, low, zero}. The output value (that is, the predicted value) of the opportunity node N34 is selected from two values {on, off}.

  +1 and −1 at the opportunity node N31 mean that the set temperature is raised by 1 [° C.] and 1 [° C.], respectively, keep means that the set temperature is not changed, and on and off are the operations of the air conditioner 31a. It means stop. The same applies to the opportunity node N32, and “on” and “off” mean operation and stop of the air conditioner 31a.

  The opportunity node N34 is set as shown in Table 4 in which the probabilities in the case where the electricity rate is {high, cheap, zero} are set for all combinations of the output values of the decision node N31 and the opportunity node N32. Has a conditional probability table of contents. Therefore, when the operating state of the air conditioner 31a is changed, it is possible to predict a change in the electricity bill according to Table 4 from the output values of the decision node N31 and the opportunity node N32.

  On the other hand, the comfort level for the next 5 minutes is predicted based on the relationship between the current comfort level and the future operating state of the air conditioner 31a. That is, the opportunity node N33 is predicted by a combination of output values of the decision node N31 that determines the operation of the air conditioner 31a and the machine node N26 that estimates the current comfort level that is the output result of the comfort level estimation unit 16. As described above, the output value of the decision node N31 is selected from the five values {+1, -1, keep, on, off}. On the other hand, as an output value (that is, a predicted value) of the opportunity node N26, an integer index is used as in the PMV. In this embodiment, as the output value of the opportunity node N26, six types of five integer values of −2 to +2 and N / A that is an unusable value are used. That is, the output value of the opportunity node N26 is selected from six values {-2, -1, 0, +1, +2, N / A}.

  Therefore, in the opportunity node N33, the comfort level becomes {good, good, bad, bad} for all combinations of output values of the opportunity node N26 and the decision node N31. A conditional probability table having contents as shown in Table 5 in which probabilities are set is provided. By using this conditional probability table, when the operating state of the air conditioner 31a is changed, the comfort level changes from the output values of the opportunity node N26 and the decision node N31 according to Table 5 (part of all combinations are described). Can be predicted.

  As described above, when an opportunity node N33 predicts a change in comfort level for the next 5 minutes and an opportunity node N34 predicts an increase or decrease in the electricity price for the next 5 minutes, By adding the values (the above-described modes), it is possible to estimate the utility value using any one of Tables 1 to 3. That is, the post-utility value when the operating state of the air conditioner 31a is changed according to the inference rule using the post-mortem prediction value obtained for the comfort level and the electricity rate, respectively, and the preferential utility value based on the user's values. Is estimated.

  By the way, in the present embodiment, when determining the operation of the air conditioner 31a, not only the operation content of the air conditioner 31a but also information on the presence or absence of a person in a room (for example, a living room) where the temperature is adjusted by the air conditioner 31a is used. That is, the decision node N31 uses the output value of the opportunity node N23 related to the latest operation of the air conditioner 31a and the output values of the nodes N12 and N13 related to the detection information of the human sensor 21 and the sound sensor 22 used in the activity amount estimation unit 17. The output value is determined.

  In addition, the current comfort level of the user in the room is estimated by the comfort level estimation unit 16, and the comfort level estimation unit 16 estimates the current comfort level of the user using the PMV and the user's own declaration. To do. That is, the comfort level estimation unit 16 includes an output value of the opportunity node N11 in which the activity amount estimation unit 17 estimates the activity amount (metabolic amount) based on the outputs of the human sensor 21 and the sound sensor 22, and the temperature sensor 23. The output value of the opportunity node N24 that calculates the PMV using the output value of the opportunity node N21 and the output value of the opportunity node N22 having the humidity sensor 23, the opportunity node N23 that uses the latest operation of the air conditioner 31a as the output value, and the opportunity From the output value of the opportunity node N25 that estimates the user's thermal sensation estimated from the output value of the node N23, the current comfort level of the user is estimated at the opportunity node N26.

  First, the activity amount estimation unit 17 will be described. The output value of the opportunity node N11 provided in the activity amount estimation unit 17 reflects the activity amount of the user, and is selected from three values {moving, static, absent}. That is, the presence / absence of the user in the room where the temperature is adjusted by the air conditioner 31a, and the movement of the user when it is present are used as the output value of the activity amount.

  The human sensor 21 and the sound sensor 22 are used for detecting the amount of activity. That is, the number of rises of the square wave signal output from the human sensor 21 and the sound sensor 22 is counted every 5 minutes, and the opportunity nodes N12 and N13 classify the number of times into three stages, and the output value is { It is selected from three values of many, medium, and small}. The output values of the opportunity nodes N12 and N13 give an indication of the amount of activity of the user. Here, the probability propagation calculation is performed by setting the prior probability values as shown in Table 6 in the opportunity node N11 and the conditional probability tables as shown in Table 7 and Table 8 in the opportunity nodes N12 and N13, respectively. The amount of activity can be estimated.

  The conditional probability tables in Tables 7 and 8 can be determined statistically using actual values obtained by the human sensor 21 and the sound sensor 22. That is, it becomes possible to determine the conditional probability table using the characteristics of individual user behavior. In the present embodiment, the number of rising times of the square wave signal is counted every 5 minutes, but this time interval can be set appropriately.

  Next, the comfort level estimation unit 16 will be described. The comfort level estimation unit 16 calculates PMV at the opportunity node N24. As is well known, calculation of PMV requires six variables: activity (metabolism), clothing, air temperature, average radiation temperature, average wind speed, and relative humidity. In the present embodiment, by using the temperature and humidity measured by the temperature sensor 23 and the humidity sensor 24, values to be substituted for variables other than the activity amount are calculated.

That is, since it is empirically known that the clothing amount Y [clo] is related to the temperature X [° C.] by the following equation, the clothing amount Y is approximated with the room temperature measured by the temperature sensor 23 as the temperature X. Ask for.
Y = −0.0377X + 1.4864
As for the average radiation temperature, a radiation thermometer is required for accurate measurement. However, it is usually possible to substitute the average radiation temperature by adding 0.5 to 1.0 [° C] to the room temperature. Are known. Therefore, in this embodiment, a value obtained by adding 0.5 [° C.] to the temperature detected by the temperature sensor 23 is used instead of the average radiation temperature. Furthermore, for the average wind speed, an anemometer is required for accurate measurement, but it is empirically known that it is 0.1 to 0.2 [m / s] if indoors. In the present embodiment, a value of 0.1 [m / s] is substituted.

  As the air temperature and relative humidity, values measured by the temperature sensor 23 and the humidity sensor 24 are used. For the amount of activity, the output value of the opportunity node N11 that maximizes the probability value obtained by the probability propagation calculation using the conditional probability table at the opportunity nodes N12 and N13 is selected, and the output value of the opportunity node N11 is { Depending on which value of dynamic, static, absent}, the activity (metabolism) for each value is 3.5 [mets], 1.0 [mets], and an unusable value N, respectively. / A is assigned. In short, the output value of the opportunity node N11 having the maximum probability is 3.5 [mets] if {moving}, 1.0 [mets] if {static}, N / A if {absent}. Is used for activity.

  As described above, the PMV can be calculated by using the detection results obtained by the human sensor 21, the sound sensor 22, the temperature sensor 23, and the humidity sensor 24. Since the opportunity node N21 having the temperature sensor 23, the machine node N22 having the humidity sensor 22, and the opportunity node N24 for calculating the PMV are determined by measurement or calculation, these opportunity nodes N21, N22, N24 are A double ellipse is shown to indicate that it is not based on probability propagation calculation. The output value of the opportunity node N24 for calculating the PMV is selected from an integer value of −3 to +3 and N / A used for the output value when the activity amount is N / A. That is, the output value of the opportunity node N24 is selected from eight of {-3, -2, -1, 0, +1, +2, +3, N / A}. Here, -3 indicates that it is cold, and +3 indicates that it is hot.

  As described above, the PMV value can be objectively determined based on the measurement results obtained by the human sensor 21, the sound sensor 22, the temperature sensor 23, and the humidity sensor 24. However, PMV is an effective index in a steady state where a person in the room continues the same state, but shows a degree of comfort in an unsteady state immediately after a person enters the room or after exercise. There is a problem that it is not necessarily appropriate as an index.

  In the present embodiment, the comfort level estimation unit 16 uses the comfort level (declared value) that is the degree of comfort reported by the user so that the user's subjectivity is reflected in the determination of comfort. is there. By using the reported value, which is a subjective sensation, together with PMV, which is an objective index, to determine comfort, the above-described problem due to determination using only PMV is avoided. Here, in the present embodiment, by using the operation of the air conditioner 31a by the user as a declaration of comfort by the user, the user can declare the comfort without being aware of it. Yes.

  For example, when the user performs an operation to lower the temperature setting of the air conditioner 31a during the cooling operation of the air conditioner 31a, it is estimated that the user feels hot, and when the user performs an operation to increase the temperature setting. It is estimated that the user feels cold. In other words, the temperature setting increase / decrease is replaced with the declared value.

  Although the user can operate the air conditioner 31a (operation on the operation state) using the user interface 32 shown in FIG. 1, a remote controller (not shown) can also be used. As the remote control device, a device that uses infrared rays as a transmission medium and can perform operation, stop, temperature setting, air volume setting, wind direction setting, and the like of the air conditioner 31a may be used. However, here, attention is paid only to instructions for raising and lowering the set temperature and for operating and stopping.

  Therefore, the output value of the opportunity node N23 is selected from five values of {up, down, keep, on, off}. {Up} increases the set temperature, {down} decreases the set temperature, {keep} maintains the set temperature, {on} operates the air conditioner 31a, {off} stops the air conditioner 31a Each means Prior probabilities are set for each output value of the opportunity node N23 as shown in Table 9.

  At the opportunity node N25, the user's thermal sensation is estimated according to the conditional probability table shown in Table 10 from the operation on the operating state of the air conditioner 31a. The output value of the opportunity node N25 is selected from three values {-1, 0, +1}. Here, {−1} represents “cold”, {+1} represents “hot”, and {0} represents “neither”.

  When the PMV is calculated by the opportunity node N24 as described above and the user's thermal sensation is estimated at the node N25, the current value of the user at the opportunity node N26 is obtained using the output values of both the opportunity nodes N24 and N25. Estimate comfort. In other words, the opportunity node N26 evaluates the user's current comfort level by integrating the objective comfort level related to cold and warm and the subjective comfort level.

Here, an example in which a weighted sum is used to integrate comfort levels is shown. That is, assuming that the weighting factor for PMV is α and the weighting factor for the user's thermal sensation is (1−α), the current comfort level AM is obtained by the following equation.
AM = α (PMV) + (1−α) (warm / cool sensation)
However, (PMV) is an output value of the opportunity node N24, and (temperature / cool feeling) is an output value of the opportunity node N25. If the weighting factor α is changed, it is possible to select whether to give priority to the objective comfort or the subjective comfort of the user. For example, if α = 0, only the user's subjective comfort is reflected, and if α = 1, the user's subjective comfort is not reflected, and the comfort is evaluated only by PMV. Become.

  In this embodiment, it is assumed that the weighting factor α is set to 0.5. That is, comfort is estimated using objective comfort and user's subjective comfort on an equal basis. Table 11 shows values calculated by rounding off the decimal point of the calculated value of the comfort level AM when the weighting coefficient α is set to 0.5. In the opportunity node N26, a conditional probability table obtained by converting the format of Table 11 is set, and the current comfort level is obtained by probability propagation calculation.

  Now, using the influence diagram shown in FIG. 2, detection information (opportunity nodes N12, N13, N21, N22) by the human sensor 21, sound sensor 22, temperature sensor 23, humidity sensor 24, and the air conditioner 31a by the user. Appropriate values are set for the latest operation information (opportunity node N32) and the latest operation state of the air conditioner 31a (decision node N31), and the mode is selected using the user interface 32, and the above-described prior probability is set. When the probability propagation calculation is performed using the conditional probability table, the operation of the air conditioner 31a that maximizes the utility value at the value node N35 can be simulated.

  An example of the result of executing the simulation is shown below. The simulation conditions are as follows: when the air conditioner 31a is in the cooling operation, the room temperature detected by the temperature sensor 23 is 29 [° C.], the relative humidity detected by the humidity sensor 24 is 50 [%], and the average radiation temperature is 29.degree. 5 [° C.], and the average wind speed is 1.0 [m / s]. The amount of clothes is set to 0.39 [clo] by the above-described approximate calculation using room temperature.

  Under the above-mentioned conditions, when the activity amount is each value {moving, static, N / A} (that is, the activity amount is 3.5 [mets], 1.0 [mets], each value of N / A) PMV is calculated. Here, a conditional probability table as shown in Table 12 is set for the opportunity node 24 for calculating the PMV.

  Tables 13 to 15 show the results of simulations using the above-described conditions in the comfort-oriented mode, the energy-saving-oriented mode, and the intermediate mode.

  As is clear from the simulation results shown in Tables 13 to 15, the number of operations (that is, “−1”) for lowering the set temperature by 1 [° C.] in the comfort-oriented mode → the intermediate mode → the energy-saving-oriented mode is In the energy saving emphasis mode, the air conditioner 31a is stopped under any conditions.

  The device management apparatus 10 described above can be realized by a single computer such as a personal computer, but may be provided in a server SV provided separately from the house M as shown in FIG. In addition to the air conditioner 31a as a facility, the house M is provided with a human sensor 21, a sound sensor 22, a temperature sensor 23, and a humidity sensor 24, and a user interface 32 that includes a touch panel and can be interactively operated. It is done. Further, a remote control device 33 using infrared rays as a transmission medium is provided for instructing the operating state of the air conditioner 31a. An electric energy sensor 25 is inserted in the power supply path from the commercial power supply AC to the air conditioner 31a, and the power consumption in the air conditioner 31a is detected.

  By the way, since the device management apparatus 10 is provided in the server SV, the various sensors 21 to 25, the user interface 32, and the remote control device 33 described above are connected to a wide area information communication network NT such as the Internet via the communication means 42. ing. On the other hand, the server SV is provided with a communication means 41 for connecting the device management apparatus 10 to the wide area information communication network NT.

  In the configuration shown in FIG. 4, data detected by the sensors 21 to 25 arranged in the house M is transmitted to the server SV via the wide area information communication network NT, and the device management apparatus 10 provided in the server SV The operation of the air conditioner 31a is selected so that the utility value after 5 minutes is maximized. In calculating the utility value, values corresponding to the mode instructed from the user interface 32 are used. The current operating state of the air conditioner 31a (corresponding to the opportunity node N32) is detected by the electric energy sensor 25 and notified to the server SV, and the latest operation of the air conditioner 31a (corresponding to the opportunity node N23) is the operation content of the remote control device 33. Is notified to the device management apparatus 10 by notifying the server SV.

  The operation for maximizing the utility value is notified to the remote control device 33 via the wide area information communication network NT, and the air conditioner 31a is controlled via the remote control device 33. That is, in the embodiment shown in FIG. 4, a configuration is adopted in which the device management apparatus 10 automatically controls the air conditioner 31a. However, a configuration in which the user himself / herself operates the remote control device 33 by returning the operation content by the device management apparatus 10 to the user interface 32 and proposing the operation content to the user may be adopted.

  If this configuration is adopted, the server SV performs a probability propagation calculation with a large required calculation resource, and thus there is an advantage that a computer having a large calculation capability is not required for the house M. However, when a computer having sufficient computing resources exists in the house M, the above-described processing can be performed without using the server SV.

  Furthermore, the utility value is maximized in the device management apparatus 10 on condition that the data expected to be detected by the sensors 21 to 25, the assumed operating state of the device 31, and the assumed operation of the device 31. An operation (operation state) may be obtained in advance, and a rule table in which a combination of conditions and operations is a rule may be created in the device management apparatus 10. In this case, if the rule table is transferred to the user interface 32 or the remote control device 33 of the house M, the data detected by the sensors 21 to 25 is collated with the rule table without passing through the wide area information communication network NT. The device 31 can be controlled.

  In the embodiment described above, in the device management apparatus 10, the comfort level obtained using the detection outputs of the sensors 21 to 24, the operation on the device 31, and the current operation state of the device 31 (detection output of the electric energy sensor 25). The utility value based on the user's values is calculated by the probability propagation calculation using the power consumption in the device 31 based on the type of the sensors 21 to 24 detecting the comfort and the operation state of the device 31. The method for detecting is not limited to the technique described in the embodiment.

  In the above-described embodiment, the control of the air conditioner 31a that adjusts the hot and humid environment is illustrated as an example of the facility that creates the comfortable space. However, the scope of the present invention is not limited to this example, and the light The present invention can also be applied to light control or color control of a lighting fixture that generates an environment, or control of a device that generates a sound environment. In controlling these facilities, an index that objectively evaluates the environment, an index that is subjectively evaluated by the users in the environment, and an appropriate sensor for objective evaluation and subjective evaluation Will be chosen. Moreover, it is also possible to apply the technical idea of the present invention not only to equipment that is subject to HEMS, but also to equipment that is subject to BEMS. Furthermore, in the embodiment, the example in which the probability propagation operation is used as an inference rule using the predicted value obtained for the comfort level and the electricity price and the utility value given in advance based on the user's values is shown. The expression format of values and utility values and the expression format of inference rules are arbitrary.

(Embodiment 2)
In the first embodiment, the comfort level estimation unit 16 uses the declared value as the subjective feeling of the user together with the PMV that is an objective index for the determination of comfort. In the case where the air conditioner described in 1 is used as a device, a case will be described in which a technique for adjusting PMV in consideration of an elapsed time since a user enters a room where air conditioning is performed by the air conditioner will be described. Moreover, although the following example demonstrates the time of air_conditioning | cooling, the same technique is applicable also at the time of heating.

  It is assumed that the temperature of the air conditioner is set so that the temperature is appropriate when the user is stationary, and the room is maintained at a substantially constant temperature. Here, when the user enters the room from outside the room where the temperature is higher than room temperature, immediately after entering the room, heat is stored in clothes and the human body and the body surface temperature is high. It is on the high temperature side, and if time passes after entering the room, it is easily expected that the heat will be dissipated and gradually approached comfortably. Similarly, at the time of heating, the reported value is on the low temperature side immediately after entering the room, and is expected to gradually approach comfort as time passes after entering the room. If the room is maintained at a substantially constant temperature, the reported value is considered to be substantially constant when a sufficiently long time has passed since entering the room.

  For example, for a specific user, the room temperature is adjusted (cooled) so that the declared comfort value (thermal feeling) becomes 0 when the staying time in the room becomes sufficiently long Then, immediately after entering the room, the declared value of comfort by the user becomes a value larger than 0, and the reported value of comfort approaches 0 with the passage of time from entering the room. Fig.7 (a) has shown the change of the reported value of comfort after entering a room. Declared values were reported every 10 minutes.

  In the illustrated example, it can be seen that after 30 to 60 minutes have passed since entering the room, the degree of comfort becomes substantially constant and the reported value converges to almost zero. In FIG. 7A, a region between two alternate long and short dash lines indicates a comfortable range (that is, −0.5 ≦ PMV ≦ 0.5). Moreover, the upper broken line in FIG. 7 (a) shows the change in the reported value when the outside air temperature is 30 to 35 [° C.], and the lower broken line is the declaration when the outside air temperature is 25 to 30 [° C.]. The change in value is shown.

  In order to verify the change in thermal sensation shown in FIG. 7 in detail, an experiment was conducted with 10 healthy men in their 20s to 50s as subjects. When the outside air temperature is about 30 [° C.], the subject walks for 12 minutes outdoors at a speed of about 5 [km / s] (about 1 km walking), and then the room temperature is 24 [° C.]. I was allowed to enter the room that was adjusted to. After entering the room, I was seated in a chair and reported a feeling of warmth in seven steps, -3 to +3, every 5 minutes.

  FIG. 8 shows the experimental results. In FIG. 8, the average value of the reported values is shown, and the reported value is approximated by a quadratic curve indicating curve A. That is, the reported value y with respect to the elapsed time x [minutes] from entering the room is represented by the relationship y = 0.0002x2-0.0387x + 1.0574. However, since the amount of calculation increases when the reported value y is approximated by a quadratic curve, a polygonal line approximation of the quadratic curve is performed and represented by two straight lines as shown in FIG.

  By the way, in order to estimate the comfort that the user feels for the environment, the PMV can be used as in the first embodiment, and when the PMV is used, the user's thermal feeling can be objectively evaluated as the comfort. Can do. However, in PMV, although it is possible to properly evaluate the thermal feeling in a state where the user stays in a certain thermal environment for a sufficiently long time (hereinafter referred to as “steady state”), the room where air conditioning is performed In a thermal environment that is not in a steady state, such as immediately after entering a room, it is not appropriate as an index of thermal sensation felt by the user.

  Therefore, in the first embodiment, the change in the degree of comfort (subjectivity) is defined as the declared value (comfort level), and the detection results obtained by the human sensor 21, the sound sensor 22, the temperature sensor 23, and the humidity sensor 24 are used. A technique of correcting the calculated PMV with a report value is adopted.

  On the other hand, this embodiment adjusts PMV using a change in a sense of the degree of comfort with the passage of time after entering the room at the time of cooling and heating separately from the comfort level which is a subjective sense of the user. This embodiment is different from the first embodiment in that a function is provided and proper temperature adjustment is performed by this function even if the user does not report. That is, in the present embodiment, based on the actual measurement result as shown in FIG. 7A, the change in the degree of comfort with time is corrected by a correction value γ (= f (t) as shown in FIG. 7B. And a technique for adjusting the PMV by adding a correction value γ to the PMV.

The correction value γ is expressed by the following equation.
γ = − (γ1 / t1) t + γ1 (0 ≦ t <t1)
γ = 0 (t1 ≦ t)
γ1 and t1 are determined based on the actual measurement results. Using the above experimental results, γ1≈1 and t1≈32 [minutes].

  In order to adjust the PMV as described above, the device management apparatus 10 of the present embodiment, as shown in FIG. 5, enters the user's room by the sensor input unit 11 in addition to the configuration of the first embodiment shown in FIG. 1. A clock unit 18 is provided for measuring the elapsed time after the detection. Further, the comfort level estimation unit 16 adjusts the value of PMV by applying the above formula to the elapsed time from the user entering the room, which is timed by the clock unit 18. Here, the output of the human sensor 21 is used for detecting the user's entry into the room.

  The operations of the utility value calculation unit 15, the comfort level estimation unit 16, and the activity amount estimation unit 17 in the present embodiment can be represented in the form of FIG. 6 by using an influence diagram. As can be seen from a comparison between FIG. 2 and FIG. 6, in the influence diagram of the present embodiment, opportunity nodes N27 and N28 are added, and the opportunity nodes N33a, N33b, N34a and N34b are substituted for the opportunity nodes N33 and N34. Is provided. Although the wording described in each node is different between FIG. 2 and FIG. 6, the nodes denoted by the same reference numerals have substantially the same function.

  The opportunity node N27 has an elapsed time since entering the room as an output value, and corresponds to the clock unit 18. The opportunity node N28 outputs a correction value γ (= f (t)) corresponding to the elapsed time output from the opportunity node N27. The output of the opportunity node N28 is input to the opportunity node N26 and the opportunity node 33a for use in estimating the comfort level after a predetermined time from determining the current comfort level and the set temperature of the air conditioner. The functions of the opportunity nodes N26 and N33a are the same as those of the opportunity nodes N26 and N33 in the first embodiment, respectively.

  However, the opportunity node N26 of the present embodiment adds the comfort level correction value γ output from the opportunity node N28 to the PMV output from the opportunity node N24, so that the user with the passage of time since entering the room is added. A function to estimate the current comfort level is added. In the present embodiment, as the output value of the opportunity node N26, eight types of seven integer values of −3 to +3 and N / A which is an unusable value are used. That is, the output value of the opportunity node N26 is selected from {-3, -2, -1, 0, +1, +2, 3, N / A}.

  By the way, in this embodiment, since the technique which adjusts PMV with the correction value γ with the passage of time from entering the room is employed, the period until the time from the entry until the correction value γ becomes zero ( When 0 ≦ t <t1), the PMV is automatically corrected with time. That is, the PMV is automatically adjusted in consideration of the change in subjective feeling.

  In the same manner as in the first embodiment, when a subjective feeling regarding the degree of comfort due to the operation of the remote control device 33 (see FIG. 4) is added to this technique, a correction value is used in the period from the entry to the time t1. γ competes with an instruction from the remote control device 33. In order to avoid such competition, the present embodiment employs the following technology.

  Here, before describing a technique for avoiding competition, a technique for adjusting the PMV by operating the remote control device 33 will be described again. During cooling, an operation of lowering the temperature of the air conditioner 31a (see FIG. 4) is performed when the user feels hot, and an operation of increasing the temperature of the air conditioner 31a is performed when the user feels cold. Therefore, it can be said that the set temperature of the air conditioner 31a by the user reflects an evaluation based on a subjective feeling regarding the degree of comfort of the user regarding the room temperature.

  Therefore, the difference between the PMV value obtained using the current measured value of the room temperature and the PMV value obtained using the set temperature of the room temperature required by operating the remote control device 33 is obtained as the correction value Δ. If the correction value Δ is added to the PMV value obtained by using the actual measured value of the current room temperature, it is possible to adjust the PMV to reflect the subjective feeling related to the degree of comfort of the user.

As in the first embodiment, the PMV variables are represented by the function g of the following equation, with the amount of activity, the amount of clothes, the air temperature, the average radiation temperature, the average wind speed, and the relative humidity as in the first embodiment.
PMV = g (activity, clothes, air temperature, average radiation temperature, average wind speed, relative humidity)
If the value of PMV obtained using the actual measured value of the current room temperature (current room temperature) is PMVe, PMVe is expressed by the following equation.
PMVe = g (activity, clothes, current room temperature, average radiation temperature, average wind speed, relative humidity)
On the other hand, if it is considered that the degree of comfort (the user's thermal sensation) due to the subjective sense of the user is expressed by the PMV value (PMVu) corresponding to the temperature set by the remote control device 33, PMVu can be expressed as It is represented by
PMVu = g (activity, clothes, set temperature, average radiation temperature, average wind speed, relative humidity)
As the average radiation temperature, a value obtained by adding 0.5 [° C.] to the temperature set by the remote control device 33 is used.

When PMVe, which is an objective evaluation value of comfort determined by the indoor environment, and PMVu, which is a subjective evaluation value of comfort, determined by a user's report (operation of the remote controller 33) are obtained. The correction value Δ is obtained as Δ = PMVe−PMVu. In the present embodiment, PMVu is used as a thermal sensation declaration value at the opportunity node N25 when the set temperature is changed by the remote control device 33, and the current comfort level AM estimated as the output value of the opportunity node N26 in this case is calculated. Is defined by the following equation. Since the comfort level AM is obtained by calculation of PMV, it is possible to share an algorithm, and as a result, description of the program becomes easy.
AM = PMVe + Δ (= PMVe−PMVu) = 2 PMVe−PMVu
As a typical scene where the user operates the remote control, it can be assumed immediately after entering the air-conditioned environment in summer and after that. The following experiment was conducted to confirm the calculation algorithm hypothesis of the comfort level AM. After 10 healthy male subjects in their 20s and 50s walked about 1 km at the normal speed for 12 minutes in the summer, set the temperature to 22 [° C], 24 [° C] and 26 [° C] respectively. After entering the maintained air-conditioning environment, get seated in a chair and increase the current setting temperature by 1 [° C], “2 [° C] increase”, “3 [° C] increase” or “as is” Seven options, “decrease 1 [° C.],” “decrease 2 [° C.]” and “decrease 3 [° C.]” were investigated by a questionnaire system. As a result, for the case of maintaining 24 [° C.], a similar change tendency with the passage of time between the estimated value of thermal sensation based on the above hypothesis and the thermal sensation value actually reported by the subject. However, the difference between the two is about 0.5, indicating that there is room for improvement in this algorithm.

  In the present embodiment, the degree of comfort AM obtained by correcting the objective evaluation value PMVe of the comfort determined by the indoor environment using the subjective evaluation value PMVu determined by the operation of the remote control device 33 is adopted, or the user enters the room Whether to use a value obtained by adjusting PMV (= PMVe) using a correction value γ that changes in accordance with the elapsed time is switched depending on whether or not the remote controller 33 is operated.

  That is, when the set temperature is changed using the remote control device 33, the comfort level AM is adopted. If there is no change in the set temperature by the remote control device 33 during the period from the user's entry to the time t1, the time PMV adjusted according to the course is used.

  When the current comfort level is estimated as described above, as in the first embodiment, in the utility value calculation unit 15, the subsequent utility value when the operating state of the air conditioner 31a is changed is the comfort level and the power consumption. Calculated based on the predicted value of (electricity charge).

  When the set temperature of the air conditioner 31a is changed, the comfort level after the change of the set temperature (after the fact) changes. The comfort level after a predetermined time from the change (hereinafter referred to as 5 minutes) can be estimated from the indoor environment (room temperature and humidity), the set temperature of the air conditioner 31a, and the elapsed time from entering the room. If the comfort level after 5 minutes from the change of the set temperature is estimated, the change in the comfort level after 5 minutes can be evaluated by comparison with the current comfort level.

  The PMV value is used for the comfort level after a predetermined time (for example, after 5 minutes) after changing the set temperature of the air conditioner 31a. However, among the six variables for obtaining the PMV, the amount of activity, the amount of clothing, the average wind speed, and the relative humidity are values obtained when the objective evaluation value PMVe is obtained, and the temperature is the set temperature of the air conditioner 31a. The average radiation temperature is a value obtained by adding 0.5 [° C.] to the set temperature. The PMV value calculated in this way should be estimated in consideration of the point in time that has elapsed since the user entered the room when the preset temperature has been changed and the predetermined time has passed. When changing the set temperature immediately after entering the room or before entering the steady state, it is necessary to correct the previously calculated PMV value. Table 16 shows an example of the correction value ε defined by the elapsed time since the user entered the room when the timing for changing the set temperature of the air conditioner 31a is not a steady state.

  Table 16 shows that the amount of activity is 1.0 [mets], the amount of clothes is 0.5 [clo], the average wind speed is 0.1 [m / s], and the relative humidity is 50 [%] after entering the room. An example of the correction value ε corresponding to the elapsed time is shown. When changing the set temperature of the air conditioner 31a between 20-30 [° C.] and calculating the PMV value using the value obtained by adding 0.5 [° C.] to the set temperature as the average radiation temperature, the steady state is shown in Table 16. The value shown as is obtained.

  As described above, the subjective comfort level changes until a steady state is reached after a sufficiently long time (time t1 = 30 to 60 minutes) immediately after entering the room. The correction value ε also changes during a period when the steady state is not reached as will be described later. During cooling, since it is hot immediately after entering the room and the comfort level gradually approaches 0, the correction value ε immediately after entering the room is set to +1, for example, and is defined as +0.8 after 5 minutes and +0.6 after 10 minutes. This correction value ε is added to the comfort level obtained for the steady state. Furthermore, the comfort level after adding the correction value ε is rounded to an integer value (the right column of each column represents the comfort level after rounding).

  The comfort level is calculated every unit time (for example, every 5 minutes), and the comfort level after a predetermined time (for example, 5 minutes) is calculated based on the set temperature of the air conditioner 31a as described above. It becomes possible to predict a change in the comfort level later. As the value of the change in the comfort level after a predetermined time (the output value of the node N33b), four values of {become good, remain good, remain bad, get worse} are used. However, the value of {N / A} is also added as impossible.

  The opportunity node N33b representing the change in the comfort level after a predetermined time is uncertain and the output value is determined probabilistically. Therefore, the opportunity node N33b includes all of the current comfort level and the comfort level after the predetermined time. For the combinations, a conditional probability table having contents as shown in Table 17 in which the probabilities when the comfort level is {better, better, worse, worse, N / A} is set. . By using this conditional probability table, when the operating state of the air conditioner 31a is changed, the comfort level is calculated according to Table 17 (part of all combinations are described) from the current comfort level and the comfort level after a predetermined time. It becomes possible to predict changes.

  By the way, when the set temperature of the air conditioner 31a is changed, the power consumption (unit power consumption) also changes. That is, in the opportunity node N34a representing the unit power consumption after changing the set temperature of the air conditioner 31a, each value of the unit power consumption {high, normal, low, zero} with respect to the set temperature of the air conditioner 31a. A conditional probability table having contents as shown in Table 18 in which probabilities are set is provided. By applying this conditional probability table, the opportunity node N34a can predict a change in unit power consumption after changing the set temperature of the air conditioner 31a.

  On the other hand, since the current unit power consumption of the air conditioner 31a is known, the current unit power consumption (output value of the opportunity node N32) and the unit power consumption after changing the set temperature estimated at the opportunity node N34a Is used to predict a change in the electricity rate after a predetermined time at the opportunity node N34b. For the opportunity node N34b, as shown in Table 19, the probability is set when the electricity price after a predetermined time becomes {higher, no change, cheaper}.

  As described above, the opportunity node N33b predicts a change in the comfort level after a predetermined time (in the example, 5 minutes), and the opportunity node N34b predicts a change in the electricity rate after the predetermined time (in the example, after 5 minutes). Thus, as in the first embodiment, it is possible to obtain a utility value according to whether the user attaches importance to comfort or electricity charges. That is, the “comfort-oriented mode”, “energy-saving-oriented mode”, and “medium mode” can be selected according to the user's values, and the utility value calculation unit 15 obtains the utility value according to the selected mode. As shown in Tables 20 to 22, utility values are assigned in advance for each mode.

  Table 20 shows utility values in the comfort-oriented mode, Table 21 shows utility values in the energy-saving-oriented mode, and Table 22 shows utility values in the medium mode. The concept of the utility value is the same as Tables 1 to 3 in the first embodiment.

  In order to evaluate the calculation of the utility value in the device management apparatus described above, it is assumed that the user does not operate the remote control device 33, and an influence diagram as shown in FIG. 9 which is a part of the influence diagram shown in FIG. The simulation was performed.

  In the simulation, all combinations of the current comfort level at the opportunity node N26 and the current unit power consumption of the air conditioner 31a at the opportunity node N32 are set in advance, and the probability inference is performed, thereby executing the air conditioner 31a. The maximum value of the utility value was obtained by setting the temperature. In this simulation, assuming the cooling time, as the variables for obtaining PMV, the activity amount is 1.0 [mets], the clothing amount is 0.5 [clo], the average wind speed is 1.0 [m / s], and the relative humidity The air temperature is the set temperature of the air conditioner 31a, and the average radiation temperature is a value obtained by adding 0.5 [° C.] to the set temperature of the air conditioner 31a.

  Tables 23 to 25 show examples of simulation results immediately after entering a room in each mode of “comfort-oriented mode”, “energy-saving mode”, and “medium mode”, respectively.

  In the simulation results of the comfort-oriented mode shown in Table 23, the set temperature is 24 [° C.] on the higher temperature side than when the current comfort level is comfortable (PMV = 0) regardless of the current unit power consumption. ing.

  In the simulation results of the energy saving emphasis mode shown in Table 24, the air conditioner 31a is operated only on the lower temperature side than the case where the current unit power consumption is above the standard and the current comfort level is comfortable, and in other cases The air conditioner 31a is turned off. In short, the cooling operation by the air conditioner 31a is not performed as much as possible, and the increase in the electricity bill is suppressed.

  In the intermediate mode simulation results shown in Table 25, when the unit power consumption is low and the current comfort level is higher than the comfortable level, the air conditioner 31a is selected to be turned off. This can be interpreted as recommending that the air conditioner 31a be turned off once when the cooling load of the air conditioner 31a is not increased.

  Tables 23 to 25 show the simulation results immediately after entering the room. Tables 26 to 28 show the simulation results in the steady state after a sufficiently long time has passed since the entry.

  Table 26 shows simulation results in the comfort-oriented mode, and a constant temperature (27 [° C.]) higher than that immediately after entering the room is selected regardless of the unit power consumption. Table 27 shows simulation results in the energy saving priority mode. In the temperature range that can be tolerated, the cooling is turned off as much as possible to suppress the generation of the electricity bill. Table 28 shows the simulation results in the intermediate mode, and the air conditioner 13a is turned off only when the cooling function of the air conditioner 31a is not so necessary (when the current unit power consumption is low). Regardless of the mode, in this simulation result, the set temperature when operating the air conditioner 31a is 27 [° C.].

  In the present embodiment, with respect to the control of the air conditioner 31a, it is possible to select an operation considering the comfort level according to the elapsed time from entering the room, an operation considering the temperature setting by the remote control device 33, and calculating the utility value. The point of using the degree of comfort and the change in electricity rate is the main difference from the first embodiment, and the other configurations and operations are the same as those of the first embodiment.

  The results of evaluating the effectiveness of this embodiment through experiments will be described below. In the influence diagram shown in FIG. 6, as described above, the current comfort level AM (the output value of the opportunity node N26), the objective evaluation value of comfort (the output value of the opportunity node N24), and the comfort level are shown. From the subjective evaluation value (the output value of the opportunity node N25). Also, the subjective evaluation value of comfort determined by the indoor environment is adjusted using a correction value γ that changes according to the elapsed time since the user entered the room, or a quadratic curve in FIG.

  By the way, as described above, in order to obtain the current comfort level AM (the output value of the opportunity node N26), it is necessary to combine many variables. That is, the variables include the amount of activity of the user, the amount of clothing of the user, the average radiation temperature, the average wind speed, the relative humidity, the set temperature by the operation of the remote control device 33, the elapsed time since entering the room, and the indoor air temperature. included. Therefore, if all variables are made variable, the number of combinations of variables becomes enormous, and a combinational explosion of variables occurs, making it difficult to evaluate the effectiveness.

  Therefore, in the experiment described below, some variables were evaluated by reducing the number of combinations of variables as fixed values or values having a certain relationship with other variables. In addition, a variable having a fixed value or a value having a certain relationship with the values of other variables was selected so as not to greatly affect the evaluation result. Specifically, the amount of activity, the amount of clothes, and the average wind speed were fixed values, and the average radiation temperature was a value obtained by adding 0.5 [° C.] to the indoor air temperature. The room was cooled using an air conditioner 31a.

  In order to set the amount of activity to a fixed value, the user (subject) decided to sit still on the chair without moving in the room after entering the room. The amount of activity in this case was 1.0 [mets]. The subjects were men, and the clothes were trunks, T-shirts, short-sleeved shirts, trousers and socks. The amount of clothes in this state corresponds to 0.5 [clo]. The average wind speed was set to 0.1 [m / s].

  Further, the temperature setting by the remote control device 33 is not performed, and the set temperature of the air conditioner 31a is set to a constant value (24 [° C.]).

  Based on the above conditions, there are two types of variables that are variable for determining the current comfort level AM: indoor air temperature and indoor relative humidity. Here, the indoor air temperature was measured by the temperature sensor 23, and the indoor relative humidity was measured by the humidity sensor 24 (see FIG. 3).

  By setting the above-described conditions, information regarding the human sensor 21, the sound sensor 22, and the remote control device 33 (see FIGS. 3 and 4) in the influence diagram shown in FIG. 6 is no longer necessary. That is, the opportunity nodes N12, N13, N23, and N25 and the links input to the opportunity nodes N12, N13, and N25 are no longer necessary. Thereby, FIG. 6 can be rewritten as shown in FIG.

  Since the influence diagram shown in FIG. 10 does not include the operation of the remote control device 33 by the user (subject), the experimenter set the temperature of the air conditioner 31a. That is, the experimenter determined the temperature of the air conditioner 31a every 5 minutes according to the elapsed time after the subject entered the room, and set the determined temperature as the set temperature of the air conditioner 31a. The set temperature of the air conditioner 31a is variable from 20 [° C.] to 30 [° C.] in increments of 1 [° C.]. In determining the temperature of the air conditioner 31a, the temperature was determined in accordance with the passage of time since entering the room and the mode described above.

  Further, instead of the temperature sensor 23 and the humidity sensor 24, a thermometer and a hygrometer were arranged on a desk installed adjacent to the chair on which the subject is seated. Therefore, as the indoor air temperature and relative humidity, values obtained by the experimenter reading the indicated values of the thermometer and the hygrometer were used.

  In order to estimate the PMV output from the opportunity node N24, as described above, the amount of activity (1.0 [mets]), the amount of clothes (0.5 [clo]), the average wind speed (0.1 [m / s]) is a fixed value, and an average radiation temperature obtained by adding 0.5 [° C.] to room temperature measured with a thermometer is used. Therefore, at the opportunity node N26, the current comfort level AM is obtained by adding the correction value γ obtained at the opportunity node N28 to the PMV value obtained at the opportunity node N24.

  On the other hand, the comfort level after a predetermined time estimated at the opportunity node N33a (after 5 minutes in the illustrated example) was approximately evaluated based on the set temperature of the air conditioner 31a. That is, the indoor air temperature after a predetermined time is assumed to be the set temperature of the air conditioner 31a, and the average radiation temperature after the predetermined time is obtained by adding 0.5 [° C.] to the set temperature of the air conditioner 31a. . In addition, since the amount of activity, the amount of clothes, and the average wind speed are fixed values, they do not change after a predetermined time. It was assumed that the relative humidity did not change after a predetermined time.

  In the opportunity node N33a, the PMV value after a predetermined time was estimated in the same procedure as the opportunity node N26 using the above-described conditions, and the value of the change in the comfort level after the predetermined time was obtained. That is, the opportunity node N33a estimates the PMV value after a predetermined time, and calculates the comfort level after the predetermined time by adding a correction value γ corresponding to the elapsed time after entering the room to this value. Further, at the opportunity node N33b, the value of the change in the comfort level after a predetermined time with respect to the current comfort level is obtained. As described above, the output value of the opportunity node N33b is a value selected from {Improved, Good, Bad, Bad, N / A}.

  In the actual experiment, the comfort level was not calculated at the opportunity nodes N26 and N33a, but a model represented by a simplified influence diagram as shown in FIG. 9 was used. The output values of the opportunity nodes N26 and N33a in this model were prepared at each timing for determining the set temperature of the air conditioner 31a (here, every 5 minutes after the subject entered the room). In this case, the conditional probability table (see Table 17) used in the opportunity node N33a at each timing is different.

  Further, a rule for executing a simulation at each timing in advance and determining the set temperature of the air conditioner 31a from the output value of the node N26 (current comfort level) and the output value of the node N32 (current unit power consumption of the air conditioner). Groups were created. The simulation was performed for each mode of “comfort-oriented mode”, “energy saving-oriented mode”, and “medium mode”, and a rule group corresponding to each mode was created.

  That is, the set temperature of the air conditioner 31a with respect to the combination of the output values of the opportunity nodes N26 and N32 changes the set temperature of the air conditioner 31a and obtains the output value (utility value) of the node N35. Temperature was used. The change range of the set temperature of the air conditioner 31a was 20 [° C.] to 30 [° C.], and the unit of change was 1 [° C.] increments. Further, the set temperature of the air conditioner 31a includes the output value {−3, −2, −1, 0, +1, +2, 3, N / A} of the opportunity node N26 and the output value {high, normal, For all combinations with low, zero}. Thus, the rule which determines the preset temperature of the air-conditioner 31a with respect to the combination of the output value of opportunity node N26, N32 was determined, and the rule group was created by matching preset temperature for every combination.

  Therefore, in evaluating the effectiveness of the present embodiment, when the set temperature of the air conditioner 31a is adjusted using the model shown in the influence diagram shown in FIG. 9, the set temperature of the air conditioner 31a is constant at 24 [° C.]. An experiment was conducted to compare the case of maintaining the temperature. The test subjects were five men in their 20s to 50s.

  The power consumption of the air conditioner 31a was measured by connecting a power sensor to an outlet that supplies power to the air conditioner 31a. Moreover, the indicator which shows the instruction | indication value of a power sensor with a thermometer and a hygrometer is arrange | positioned on the desk arrange | positioned adjacent to the chair where a test subject sits, and the thermometer, the hygrometer, and the indicator value of the indicator are 5 Read every minute. Since the output value of the opportunity node N26 is determined from the indicated values of the thermometer and the hygrometer, and the output value of the opportunity node N32 is determined from the indicated value of the power sensor, the set temperature of the air conditioner 31a is determined by the above-described rule group. That is, when adjusting the set temperature of the air conditioner 31a after entering the room, the set temperature of the air conditioner 31a is adjusted according to the above-described rule group.

  11 and 12 show the results of the experiment. The temperature shown on the graph in FIG. 11 and FIG. 12 shows the set temperature of the air conditioner 31a determined using the influence diagram shown in FIG. 9 as a model. That is, the set temperature is set to 23 ° C. immediately after entering the room, the set temperature is gradually increased until the elapsed time from entering the room reaches 25 minutes, the set temperature is lowered after 25 minutes, and the set temperature is increased after 75 minutes. 27 [° C.]. In addition, FIG. 12 has shown using the average value of the thermal sensation report by five test subjects. 11 and 12, the result A shows the case where the set temperature of the air conditioner 31a is adjusted using the influence diagram of FIG. 9 as a model, and the result B keeps the set temperature of the air conditioner 31a constant at 24 [° C.]. Shows the case.

  FIG. 11 shows the transition of power consumption due to the operation of the air conditioner 31a over a time period of 150 minutes after the subject enters the room. As is apparent from the figure, when the set temperature is constant (result B), the power consumption is substantially constant. On the other hand, when the set temperature is adjusted based on the model shown in FIG. 9 (Result A), power consumption temporarily increases immediately after entering the room, but sufficient time (about 45 minutes) has elapsed. As a result, the power consumption is lower than when the set temperature is kept constant at 24 [° C.]. With respect to the total power consumption for 150 minutes, a reduction effect of about 45% was obtained as compared with the case where the set temperature was set to 24 [° C.].

  FIG. 12 shows the transition of the reported value of thermal sensation by the subject over time after the subject entered the room. The broken line in the figure shows a range where the reported value of thermal sensation is ± 0.5, and this range represents a comfortable range that is neither hot nor cold. As is clear from the figure, the time required to reach the comfortable range after entering the room is more constant when the set temperature is adjusted based on the model shown in FIG. 9 (result A) at 24 [° C.]. Is shorter than the result (result B). In addition, when the set temperature is constant, after about 90 minutes have passed since entering the room, the reporting notice is -1 or less. That is, when the set temperature is constant, it is evaluated that it is slightly cold. On the other hand, when the set temperature is adjusted, the temperature is maintained so as not to deviate from the comfortable range.

  According to the above experiment, at the time of cooling, by adjusting the set temperature of the air conditioner 31a using the model based on the influence diagram shown in FIG. 9, it is possible to achieve both comfort related to thermal sensation and reduction of power consumption. The result is that it becomes possible. That is, in order to determine the set temperature of the air conditioner 31a during cooling, it was effective to perform control based on the influence diagrams shown in FIGS. In short, when the user enters the room from outside, the comfort temperature and the power saving can be achieved by adjusting the set temperature of the air conditioner 31a as described above.

  In the experiment described above, the effect on the set temperature of the air conditioner 31a during cooling was verified. However, even during heating, the correction value γ corresponding to the elapsed time after entering the room and the conditional probability table used in the opportunity nodes N33a and N34a were changed. Then, the same technology can be applied. In the above-described experiment, fixed values are used for the activity amount and the clothing amount. However, if the activity amount and the clothing amount are detected by detecting the behavior and clothing of the user in the room, the set temperature of the air conditioner 31a is further increased. It becomes possible to control accurately.

10 equipment management device 11 sensor input unit (first means)
14 mode selection unit 15 utility value calculation unit (second means)
16 Comfort estimation part (1st means)
17 Activity amount estimation unit (first means)
18 Clock part 21 Human sensor (sensor)
22 Sound sensor (sensor)
23 Temperature sensor (sensor)
24 Humidity sensor (sensor)
25 Electric energy sensor (sensor)
31 Equipment 32 User Interface 33 Remote Control Device NT Wide Area Information Communication Network SV Server

Claims (16)

A device management device that instructs the operation of a device that brings comfort to the user by operating by consuming a medium supplied from a supplier, and includes a sensor input unit to which a detection output of a sensor is input, A first means for estimating a comfort level representing a degree of a user's current comfort using the detection output of the sensor, and an operation of the device based on the current comfort level of the user estimated by the first means Obtain a predicted value of the subsequent comfort when the state is changed, obtain a predicted value of the subsequent consumption of the medium when the operating state of the device is changed based on the current operating state of the device, Estimating the subsequent utility value when the operating state of the device is changed according to the inference rule using the predicted value obtained for the consumption amount of the medium and the utility value given in advance based on the user's values, After the fact Use value determines the control content of the device which is the maximum, and a second means for instructing a change of the operation state of the apparatus by the determined control content, the first means using the detection output of the sensor The objective comfort level is expressed by the comfort evaluation index, and the subjective comfort level of the user is expressed by an evaluation value corresponding to the comfort evaluation index. Equipment management device characterized by estimating the current comfort level of the user. The first means uses a reported value by the user as the evaluation value representing the degree of subjective comfort of the user, and represents the current comfort level by a weighted sum of the comfort evaluation index and the reported value. The device management apparatus according to claim 1. 3. The apparatus management apparatus according to claim 2, wherein the first means is capable of changing a weighting coefficient used for calculating a weighted sum . It said first means, the declaration value of subjective comfort based on reported users, users claim 2 or 3, wherein the estimating the contents of the most recent operation for the apparatus that has made Equipment management equipment. A device management device that instructs the operation of a device that brings comfort to the user by operating by consuming a medium supplied from a supplier, and includes a sensor input unit to which a detection output of a sensor is input, A first means for estimating a comfort level representing a degree of a user's current comfort using the detection output of the sensor, and an operation of the device based on the current comfort level of the user estimated by the first means Obtain a predicted value of the subsequent comfort when the state is changed, obtain a predicted value of the subsequent consumption of the medium when the operating state of the device is changed based on the current operating state of the device, Estimating the subsequent utility value when the operating state of the device is changed according to the inference rule using the predicted value obtained for the consumption amount of the medium and the utility value given in advance based on the user's values, After the fact And a second means for instructing a change in the operating state of the device according to the determined control content, wherein the device is an air conditioner, and the first means. , the device management apparatus characterized by having the function of estimating the comfort of the current by adjusting in accordance with the elapsed time after entering the comfort evaluation index. A device management device that instructs the operation of a device that brings comfort to the user by operating by consuming a medium supplied from a supplier, and includes a sensor input unit to which a detection output of a sensor is input, A first means for estimating a comfort level representing a degree of a user's current comfort using the detection output of the sensor, and an operation of the device based on the current comfort level of the user estimated by the first means Obtain a predicted value of the subsequent comfort when the state is changed, obtain a predicted value of the subsequent consumption of the medium when the operating state of the device is changed based on the current operating state of the device, Estimating the subsequent utility value when the operating state of the device is changed according to the inference rule using the predicted value obtained for the consumption amount of the medium and the utility value given in advance based on the user's values, After the fact Use value determines the control content of the device which is the maximum, the determined control content and a second means for instructing a change of the operation state of the apparatus, the apparatus is an air conditioner, said first means If the user does not operate the device, the comfort index is adjusted according to the elapsed time after entering the room to estimate the current comfort level, and if the user operates the device, A device management apparatus having a function of estimating the current comfort level from the contents . The first means includes a comfort evaluation index obtained using a comfort evaluation index obtained using an actual measured value of the current room temperature and a set temperature requested by the user when the user performs an operation on the device. 7. The apparatus according to claim 6 , wherein a difference between the index and a correction value is used as a correction value, and a value obtained by adding the correction value to a comfort evaluation index obtained using an actual measurement value at a current room temperature is used as a current comfort level. Management device. The said 1st means has a function which estimates the comfort degree after predetermined time by adjusting the said comfort evaluation index | exponent according to the elapsed time after entering a room , The feature of Claims 5-7 The device management apparatus according to any one of the above. The second means has a function of calculating a utility value specific to the mode according to the mode selected by the user. Comfort-oriented mode with high utility value assigned, energy-saving-oriented mode with high utility value assigned to reduce media consumption, and medium mode with equally assigned utility value for medium consumption reduction and comfort The device management apparatus according to claim 1 , wherein the device management apparatus is selectable . The second means includes, as a conditional probability table, the user's current comfort level estimated by the first means and a predicted value of the subsequent comfort level when the operating state of the equipment is changed. And the predicted value of the consumption of the medium after the change of the operating state of the device as a conditional probability table, based on two conditional probability tables and user values in advance The apparatus management apparatus according to claim 1 , wherein a subsequent utility value when the operation state of the apparatus is changed is calculated according to an inference rule using the utility value given to the apparatus. The sensor detects a user's activity without contact with the user, and the first means is an actual value of the user's activity and the sensor's detection output with respect to the sensor's detection output corresponding to the user's activity amount. The apparatus according to claim 1 , further comprising: a conditional probability table statistically determined using a relationship between the activity amount and the activity amount reflected in objective comfort. Management device. The first means and the second means provide utility values by providing, as a condition, data expected as detection output of the sensor in advance, an assumed operating state of the equipment, and an assumed operation of the equipment. It is characterized by determining the control content that maximizes the control content, holding the combination of the condition and the control content as a rule table, and determining the control content by applying the rule table to the actual detection output of the sensor The apparatus management apparatus of any one of Claims 1-11 . 13. The device management apparatus according to claim 12, wherein the rule table is provided in a remote control device that instructs an operation of the device. The first means and the second means are provided in a server that communicates with the sensor installed in a house via a wide area information communication network, and the control content of the device determined by the second means is The device management apparatus according to claim 1, wherein the device management apparatus is instructed via a wide area information communication network . A device management device that instructs the operation of an air conditioner as a device that operates by consuming electric power supplied from a supplier, and includes a sensor input unit to which a detection output of the sensor is input, and the detection output of the sensor The first means for estimating the degree of comfort representing the degree of the user's current comfort using, and the operating state of the air conditioner changed based on the current degree of comfort of the user estimated by the first means Obtain the predicted value of the ex-post comfort level, obtain the predicted value of the ex-post electricity charge when the air conditioner operating state is changed based on the current operating state of the air conditioner, and calculate the comfort level and the electric charge respectively. Estimate the subsequent utility value when the operating state of the air conditioner is changed according to the inference rules using the predicted value and the utility value given in advance based on the user's values. is there Second control means for determining the control content of the control device and instructing the change of the operating state of the air conditioner according to the determined control content, wherein the first means is objective using the detection output of the sensor The degree of comfort is expressed by the comfort evaluation index, and the subjective comfort level of the user is expressed by an evaluation value corresponding to the comfort evaluation index. Equipment management device characterized by estimating comfort level . The computer, a program for functioning as a device management apparatus for instructing the operation of the device that provides comfort to the user by operating by consuming medium supplied from the supply company, is entered into the computer A first means for estimating a comfort level representing a degree of a user's current comfort using a detection output of the sensor, and an operation state of the device based on the current comfort level of the user estimated by the first means the calculated predicted value comfort post when changing obtains a predicted value of consumption of post of said medium when changing the operational state of the device based on the current operating state of the apparatus, comfort and the post utility when changing the operational state of the apparatus in accordance with inference rules using a utility value given in advance on the basis of the predicted value with the user values determined respectively for the consumption of the medium The estimates, utility value of the post determines the control content of the device which is the maximum, the determined control contents cause the computer to function as a second means for instructing a change of the operation state of the apparatus, the first The means represents the degree of objective comfort using the detection output of the sensor as a comfort evaluation index, and represents the subjective comfort level of the user as an evaluation value corresponding to the comfort evaluation index, A program characterized by estimating a user's current comfort level from the value and the comfort evaluation index .

Images