Air conditioning system

Abstract

To provide an air conditioning system capable of properly estimating an air temperature in the neighborhood of a floor of a perimeter zone.SOLUTION: Floor neighborhood temperatures of positions different in at least directions of a perimeter zone are estimated on the basis of parameters relating to the floor neighborhood temperature of the perimeter zone, and an amount of blowout air of a variable air amount device facing the perimeter zone P is adjusted on the basis of the estimated floor neighborhood temperature. The estimation of the floor neighborhood temperature is performed by using at least a ceiling temperature grasped by a ceiling temperature sensor as the parameter.SELECTED DRAWING: Figure 5

Description

[Technical Field]

[0001] The present invention relates to an air conditioning system that provides air conditioning particularly for a perimeter zone. [Background technology]

[0002] When providing health air conditioning primarily for people in an office building or other building, the thermal conditions within the building vary. For example, perimeter zones, such as those near windows, are more susceptible to the effects of exterior loads than the interior zones within. This means that the amount of heat input varies significantly between rooms on the south side, which are more exposed to sunlight, and rooms on the north side, which are shaded even during the day. Even if the amount of heat input from the outside is the same, the thermal load may differ from room to room due to the exhaust heat from indoor equipment. To address these different thermal conditions and maintain optimal air conditions in each area, variable air volume single-duct air conditioning systems and floor-specific unit systems typically require the installation of multiple variable air volume units in the target space, with temperature sensors installed near each unit. The air volume of each variable air volume unit and the temperature of the air supplied by the air conditioner are adjusted based on the temperature sensor readings.

[0003] In air conditioning systems, particularly those designed for large-span spaces, temperature sensors are typically installed on or near the ceiling. In older spaces with many pillars, it was easy to install temperature sensors at the same level as the occupied area, but this is no longer possible. This is because people and equipment are often located near the floor, making them a nuisance. As a result, the indoor temperature measured by the air conditioning system can differ from the actual temperature measured by the people and equipment in the occupied area near the floor. In particular, during cold seasons, the air in the perimeter zone, especially near the inside of windows, is cooled by the outside air, creating a cold air flow known as a cold draft, which sinks downward and flows along the floor. This can make people in the perimeter zone or nearby areas feel uncomfortable. Because warm air has a smaller specific gravity than cold air, even if warm air is supplied from the ceiling outlet, if there is a large temperature difference between the warm air and the cold air caused by cooling by outside air, or if the air volume is insufficient, the inertial force of the air blown downward from the outlet will be canceled out by the upward buoyancy, and the air will not reach the floor. In other words, in seasons when the outside temperature is low, at foot level in the perimeter zone, for example, cold air will prevail over the warm air supplied from the variable air volume device, and it is expected that the supply of warm air to the occupied area will be insufficient.

[0004] To address this situation, it is effective to install a ventilation device known as a perimeter fan or air barrier fan, separate from the ceiling-mounted air outlet that supplies the perimeter zone with intake air controlled by the variable air volume device. The perimeter fan is installed, for example, near the floor near the window in the perimeter zone, creating an upward airflow from there. This prevents the indoor air from being cooled near the window, creating a cold draft that then flows along the floor into the room. Furthermore, by sending warm air controlled by the variable air volume device from the ceiling-mounted air outlet toward the perimeter fan, the warm air can be supplied to the perimeter zone's outer edge, maintaining a comfortable air environment for the occupants.

[0005] However, even with such a perimeter fan, it may not be sufficient to address the aforementioned cold drafts or insufficient supply of warm air to the occupied area. Since a temperature sensor for measuring indoor air temperature is installed near the ceiling, the measured value may differ from the air temperature near the floor. Therefore, if the air conditioning system is operated based on the temperature sensor's measurement, it is possible that sufficient warm air will not be supplied to the perimeter zone. For example, if the air outlet controlled by the variable air volume device is unable to supply a sufficient amount of warm air to the perimeter fan, or if there is a temperature difference between the upper and lower air in the room, the temperature measured by the temperature sensor near the ceiling will be higher than the temperature near the floor in the perimeter zone. As a result, the actual temperature near the floor near the perimeter zone may not be reflected in the operation of the variable air volume device, and even if a comfortable temperature is maintained near the ceiling, the temperature in the occupied zone may remain low. One possible solution to this situation would be to simply raise the set temperature on the variable air volume device, but if this were to be done, there would be a concern that the temperature in the occupied area would rise too much when, for example, the outside temperature rises as the season changes from winter to mid-season and the temperature difference between the top and bottom of the room becomes smaller.

[0006] For example, Patent Document 1 listed below describes a technique that can address these problems. [Prior art documents] [Patent documents]

[0007] [Patent Document 1] Patent Publication No. 2021-76348 Summary of the Invention [Problem to be solved by the invention]

[0008] The technology described in Patent Document 1 estimates the temperature near the floor in the perimeter zone and adjusts the supply air volume of the variable air volume device based on this. By estimating the temperature near the floor instead of the measurement value of a temperature sensor installed at ceiling height and using this value to operate the air conditioning system, it is possible to operate the system in accordance with the conditions near the floor.

[0009] Here, the temperature value near the floor is an estimated value instead of the measured value from a temperature sensor actually installed at that location, and it is of course desirable that this estimated value be as close to the actual value as possible. Even if the estimated temperature near the floor, which is close to the occupied area, is used instead of the measured temperature near the ceiling, if the estimated temperature deviates from the actual value, the same problems as above will ultimately occur, or conversely, there may be a case where excessive warm air is supplied.

[0010] In view of the above circumstances, the present invention aims to provide an air conditioning system that can suitably estimate the air temperature near the floor of the perimeter zone. [Means for solving the problem]

[0011] The present invention includes an air conditioner that delivers supply air; an air supply duct that guides air from the air conditioner to a target space; A plurality of air outlets that blow supply air into the target space; an air volume variable device for adjusting the volume of supply air blown out from one of the air outlets that supplies air to the perimeter zone; a perimeter fan provided in the perimeter zone for blowing indoor air upward; an intake port provided above the perimeter fan for drawing in at least a portion of return air to the air conditioner; a ceiling temperature sensor for measuring a ceiling temperature in the perimeter zone, and estimating the near-floor temperature at each position in at least different directions in the perimeter zone based on a parameter related to the near-floor temperature in the perimeter zone; The airflow rate of the variable air volume device facing the perimeter zone is adjusted based on the estimated near-floor temperature, The estimation of the near-floor temperature is performed using at least the ceiling temperature detected by the ceiling temperature sensor as a parameter. and further configured to use the amount of solar radiation as a parameter, The amount of solar radiation is grasped as an estimated value calculated based on the amount of extra-atmospheric global solar radiation during a shadow time set as a time when a shadow is cast on the pyranometer, while At other times, the actual measured value of the pyranometer is grasped. thing This relates to an air conditioning system characterized by the above. The present invention also provides an air conditioner that delivers supply air; an air supply duct that guides air from the air conditioner to a target space; A plurality of air outlets that blow supply air into the target space; an air volume variable device for adjusting the volume of supply air blown out from one of the air outlets that supplies air to the perimeter zone; a perimeter fan provided in the perimeter zone for blowing indoor air upward; an intake port provided above the perimeter fan for drawing in at least a portion of return air to the air conditioner; a ceiling temperature sensor for measuring the ceiling temperature in the perimeter zone; The system is configured to estimate near-floor temperatures at least at positions in different directions in the perimeter zone based on parameters related to near-floor temperatures in the perimeter zone, and adjust the blow-out air volume of the variable air volume device facing the perimeter zone based on the estimated near-floor temperatures; the estimation of the near-floor temperature is configured to be performed using at least the ceiling temperature detected by the ceiling temperature sensor as a parameter, and is configured to be performed using the presence or absence of solar radiation at the target position as a further parameter; The presence or absence of solar radiation is determined based on the shadow time at the target location. This relates to an air conditioning system characterized by the above.

[0012] The air conditioning system of the present invention can be configured to estimate the near-floor temperature using at least a past value of the ceiling temperature detected by the ceiling temperature sensor as a parameter.

[0013] The air conditioning system of the present invention can be configured to estimate the near-floor temperature further using the amount of solar radiation as a parameter, and to grasp the amount of solar radiation as an estimated value calculated based on extra-atmospheric global solar radiation during a shadow time set as a time when a shadow is cast on a pyranometer, while grasping the amount of solar radiation as an actual measured value by the pyranometer at other times.

[0014] In the air conditioning system of the present invention, the estimated value of the amount of solar radiation during the shadow time can be calculated by multiplying the amount of extra-atmospheric global solar radiation during the shadow time by a preset coefficient.

[0015] In the air conditioning system of the present invention, the estimated value of the amount of solar radiation during the shadow time can be calculated by multiplying the amount of solar radiation during the shadow time by a coefficient calculated by dividing the actual measured value of solar radiation before the shadow time by the amount of solar radiation outside the atmosphere.

[0016] In the air conditioning system of the present invention, the estimated value of solar radiation during the morning shadow time is calculated by multiplying the extra-atmospheric global solar radiation during the shadow time by a predetermined coefficient, and the estimated value of solar radiation during the afternoon shadow time is calculated by multiplying the extra-atmospheric global solar radiation during the shadow time by a coefficient calculated by dividing the actual measured value of solar radiation before the shadow time by the extra-atmospheric global solar radiation.

[0017] In the air conditioning system of the present invention, the amount of solar radiation can be estimated under the conditions that it is a shadow time and the weather is determined to be fine.

[0018] The air conditioning system of the present invention can be configured to estimate the near-floor temperature using the presence or absence of solar radiation at the target position as an additional parameter.

[0019] In the air conditioning system of the present invention, the presence or absence of solar radiation can be determined based on the shadow time at the target position.

[0020] In the air conditioning system of the present invention, the shadow time at a target location can be determined based on a sky factor chart.

[0021] The air conditioning system of the present invention can be configured to estimate the near-floor temperature in the perimeter zone using a temperature estimation model generated by machine learning.

[0022] In the air conditioning system of the present invention, the temperature estimation model can be configured to estimate the near-floor temperature by further using some or all of the parameters selected from the following parameters as explanatory variables: Outside temperature -Operating status of the air conditioner ·Outflow heat amount ·Wind direction ·wind speed ·Rainfall amount - Measurements from the floor-near temperature sensor that measures the floor-near temperature in the perimeter zone -Operating status of the perimeter fan The operating state of the variable air volume device facing the perimeter zone ·date ·time ·day of week The air conditioning system of the present invention can be configured so that the amount of solar radiation and the presence or absence of solar radiation at the target location are input to the temperature estimation model as mutually separate explanatory variables. [Effects of the Invention]

[0023] The air conditioning system of the present invention can provide the excellent effect of suitably estimating the air temperature near the floor of the perimeter zone. [Brief explanation of the drawings]

[0024] [Figure 1] 1 is a schematic diagram showing an example of the configuration of an air conditioning system to which the present invention is applied. [Figure 2] 1 is a schematic plan view showing an example of the arrangement of an air volume variable-volume device and a near-floor temperature sensor installed facing a perimeter zone in an air conditioning system to which the present invention is applied. FIG. [Figure 3] 1 is a schematic diagram showing an example of the overall configuration of an air conditioning system of the present invention applied to a multi-story building. [Figure 4] 1 is a schematic plan view showing an example of the arrangement of an air volume variable-volume device and a near-floor temperature sensor installed facing the perimeter zone of a reference floor in an air conditioning system to which the present invention is applied. FIG. [Figure 5] FIG. 2 is a conceptual diagram illustrating an example of an input / output configuration of a temperature estimation model. [Figure 6] 10 is a flowchart illustrating an example of a procedure for setting an operating condition of a variable air volume device or the like in an embodiment of the present invention. [Figure 7] 1 is a conceptual diagram illustrating the temperature difference between the top and bottom of the room air and temperature control based on that difference. [Figure 8] 10 is a graph illustrating an example of the relationship between a control deviation calculated based on an estimated value of the near-floor temperature and a temperature correction value. [Figure 9] 10 is a graph illustrating an example of the relationship between the outside air temperature and the upper limit of the supply air temperature. [Figure 10] 2 is a schematic diagram illustrating an example of the temperature distribution of indoor air in the vicinity of the perimeter zone of the air conditioning system of FIG. 1. FIG. [Figure 11] FIG. 3 is a block diagram conceptually illustrating control using a temperature correction value in the present embodiment. [Figure 12] 1 is a graph showing an example of fluctuations in the amount of solar radiation measured by a pyranometer (when no obstruction is present). [Figure 13] 10 is a graph showing another example of fluctuations in the amount of solar radiation measured by a pyranometer (when a shield is present). [Figure 14] 1 is a graph showing an example of the relationship between the amount of solar radiation measured by a pyranometer, the actual amount of solar radiation, and the estimated amount of solar radiation. [Figure 15] 1 is a flowchart illustrating an example of a procedure for executing a method for estimating solar radiation of the present invention. [Figure 16] 10 is a flowchart illustrating another example of the procedure for executing the method for estimating the amount of solar radiation of the present invention. [Figure 17] 10 is a flowchart illustrating yet another example of the procedure for executing the method for estimating the amount of solar radiation of the present invention. DETAILED DESCRIPTION OF THE INVENTION

[0025] Hereinafter, an embodiment of the present invention will be described with reference to the accompanying drawings.

[0026] FIG. 1 shows a schematic diagram of an example of an air conditioning system embodying the present invention. Conditioned air (supply air) A1 sent out from an air conditioner 1 is guided through an air supply duct 2 to a target space S, which is a room such as an office. Multiple air outlets 4 are installed in the ceiling 3 of the target space S, and the downstream side of the air supply duct 2 extending from each air conditioner 1 is connected to each air outlet 4. The conditioned air A1 circulating through the air supply duct 2 is supplied to the target space S from each air outlet 4. Although each air outlet 4 is shown as a single outlet, in reality each air conditioner has multiple air outlets.

[0027] An air volume variable device 5 is provided in front of each air outlet 4 in the supply air duct 2. The air volume variable device 5 is a device abbreviated as VAV (Variable Air Volume), and adjusts the air volume passing through it based on the measured values ​​by changing the opening of an internal damper and measuring the air speed and volume either before or after the damper. When the supply air A1 is supplied to the target space S, the air volume is adjusted by the air volume variable device 5.

[0028] The target space S can be divided into a perimeter zone P, such as near windows, which is susceptible to the effects of the exterior load, and an interior zone I, which is located inside the perimeter zone P and where the proportion of the total building heat load is relatively small. FIG. 1 shows one variable air volume device 5 at a position facing the interior zone I and one at a position facing the perimeter zone P, and also shows one air conditioner 1 (interior air conditioner 1a) that sends out supply air A1 to the variable air volume device 5 facing the interior zone I, and one air conditioner 1 (perimeter air conditioner 1b) that sends out supply air A1 to the variable air volume device 5 facing the perimeter zone P. (Note that this is a simplified schematic diagram, and it goes without saying that in an actual air conditioning system, the number and placement of variable air volume devices 5 and air conditioners 1, and the number of variable air volume devices 5 relative to an air conditioner 1, will be set appropriately depending on the size, shape, thermal conditions, etc. of the target space S. In particular, the air conditioner 1a for perimeter zone P and the air conditioner 1b for interior zone I are shown as separate units, but it goes without saying that this is included in the diagram, as it is common for the two air conditioners 1a, 1b to be combined into a single unit and the supply air duct and return air duct to branch off.)

[0029] A plurality of intake ports 7, 16 are provided on the ceiling 3 at positions corresponding to the interior zone I and the perimeter zone P, respectively. From each variable air volume device 5 facing the interior zone I or the perimeter zone P, supply air A1 sent from the interior air conditioner 1a or the perimeter air conditioner 1b, respectively, is supplied to the target space S through the air outlet 4, where it mixes with the air (room air) A2 in the target space S.

[0030] A portion of the indoor air A2, mainly in the interior zone I, is taken in as return air A3 from the intake port 7 facing the interior zone I. A portion of the return air A3 is discharged outside the return air duct 8 as exhaust air A4, and a portion is mixed with fresh outside air A0 taken in from the return air duct 8 and returned to the interior air conditioner 1a, where the temperature and humidity are adjusted and the air is sent out again as supply air A1.

[0031] Of the indoor air A2, a portion that is mainly in the perimeter zone P is taken in as return air A3 from the intake port 16 facing the perimeter zone P, and the entire amount returns to the perimeter air conditioner 1b through the return air duct 17, where it is temperature-adjusted and then sent out again as supply air A1. In this way, air circulates between the air conditioner 1 and the target space S. Note that while the interior zone I takes in outside air A0 to replace the air, the perimeter zone P is designed not to take in outside air A0 and to circulate all of the air. This is because it is expected that there will be fewer people in the perimeter zone P than in the interior zone I, and so less outside air is introduced to exhaust the carbon dioxide that increases with breathing than in the interior zone.

[0032] Such an air conditioning system responds to fluctuations in the air conditioning load within the target space S, for example, by using an air conditioning method called a per-floor unit method (a single-duct variable air volume method completed for each floor). Temperature sensors 9 that measure the temperature of indoor air A2 are installed at appropriate locations within the target space S (in this example, on the ceiling 3). (Hereinafter, these temperature sensors 9 that measure the temperature of indoor air A2 at or near the ceiling will be referred to as "ceiling temperature sensors." Also, below, the temperature of indoor air A2 at the height where the ceiling temperature sensors 9 are installed will be referred to as "ceiling temperature" for convenience, as needed.) Each variable air volume device 5 is equipped with a control device (VAV controller), which controls the variable air volume device 5 using PI control or the like so that the deviation between the measurement value of the ceiling temperature sensor 9 and the temperature set value of the variable air volume device 5 approaches zero, thereby supplying an appropriate volume of supply air A1 from the air outlet 4. For example, when heating, if the measurement value of the ceiling temperature sensor 9 is lower than the set value of the indoor temperature (temperature of indoor air A2), the supply amount of supply air A1 is increased, and if the set value and the measurement value are close, the supply amount is decreased.

[0033] The operating status of each device that makes up the air conditioning system is monitored and operated by a control device 10. The control device 10 is equipped with a controller that controls the operation of each air conditioner 1 (interior air conditioner 1a and perimeter air conditioner 1b), and acquires values ​​such as the required air volume ratio of each variable air volume device 5 and the set temperature in the target space S, and based on these, determines the temperature, air volume, etc. of the supply air A1 supplied from each air conditioner 1.

[0034] The control device of each variable air volume device 5 acquires the ceiling temperature as a measurement value from each ceiling temperature sensor 9, calculates the required air volume based on the deviation from the set temperature, and adjusts the opening of the damper inside the variable air volume device 5 based on the deviation from the air volume measured and calculated by an anemometer or the like provided in the variable air volume device 5, supplying supply air A1 at an appropriate volume from the air outlet 4. That is, for example, during heating, if the measurement value of the ceiling temperature sensor 9 is lower than the target value, the supply volume of supply air A1 is increased, and if the measurement value is close to the target value, the supply volume is reduced.

[0035] In each air conditioner 1, supply air A1 sent out from the variable air volume device 5 is sent out to each variable air volume device 5, and the amount of supply air A1 sent out here is the total of the required air volumes of one or more variable air volume devices 5 located downstream. Therefore, the control device 10 determines the supply air volume for each air conditioner 1 by performing calculations based on each required air volume ratio, which is the ratio of the required air volumes of each variable air volume device 5, as an input signal.

[0036] In addition to the air volume control described above, actual air conditioning systems also perform a control (load reset control) that adjusts the supply air temperature (the temperature of supply air A1 sent out from the air conditioner 1) according to the requested air volume ratio. In a variable air volume air conditioning system, it can be said that, in principle, the requested air volume ratio from each air volume variable device 5 increases when the load is heavy, and decreases when the load is light. Therefore, not only the supply air volume of supply air A1 in the air conditioner 1 but also the supply air temperature are appropriately changed according to the requested air volume ratio from each air volume variable device 5. For example, during heating, if the requested air volume is smaller than the rated air volume of the air volume variable device 5, the set value of the supply air temperature is lowered. Various methods can be used for load reset control, such as a method of changing the set value of the supply air temperature based on the opening information of the air volume variable devices 5 so that the opening of each air volume variable device 5 falls within a predetermined opening range, or a method of weighting the deviation between the set temperature of the air volume variable devices 5 and the measured temperature (the measured temperature of the air at an appropriate location in the air conditioning system; for example, the temperature of return air A3) and changing the set value of the supply air temperature based on that weighting.

[0037] In this way, in an air conditioning system such as that shown in FIG. 1, the supply air volume and supply air temperature are automatically changed according to the indoor load.

[0038] In such an air conditioning system, the temperature of the indoor air A2 is measured by the ceiling temperature sensor 9, and so there may be a discrepancy between the temperature of the indoor air A2 that a person in the target space S is actually in contact with and the temperature of the indoor air A2 measured by the control device 10. In the example shown in Fig. 1, the ceiling temperature sensor 9 is placed at the height of the ceiling 3, but due to outdoor air conditions and the control conditions of each device, such as the variable air volume device 5 and the air conditioner 1, a difference will arise between the temperature of the indoor air A2 near the floor where the person is located and the temperature of the indoor air A2 around the ceiling temperature sensor 9 (near the ceiling).

[0039] In particular, in winter when the outside air temperature is low, as described above, cold drafts may occur, causing discomfort to people in the perimeter zone P or nearby.

[0040] Therefore, in this embodiment, a perimeter fan 6 is provided near the floor of the perimeter zone P in order to correct the temperature distribution of the indoor air A2 in the perimeter zone P. The perimeter fan 6 is configured to blow the indoor air A2 in the target space S upward along the window surface. By operating this perimeter fan 6, a flow of indoor air A2 is formed from near the floor of the perimeter zone P upward, preventing the indoor air A2 near the window surface, which has been cooled by outside air, from becoming a cold draft and heading toward the interior zone I inside the perimeter zone P, and instead blowing it into the upper air inlet 16. Furthermore, by blowing indoor air A2 toward the perimeter fan 6 from the nearby variable air volume device 5, in combination with the operation of the perimeter fan 6, it is possible to supply high-temperature indoor air A2 to near the outer edge of the perimeter zone P.

[0041] Furthermore, in this embodiment, as shown in Figures 1 and 2, of the air volume variable devices 5 located near the perimeter zone P, optional near-floor temperature sensors 11 are provided near some of the air volume variable devices 5 in a plan view (in Figure 1, the position of the near-floor temperature sensors 11 in a side view is indicated by a dashed line). The near-floor temperature sensors 11 are installed on or near the floor surface at a height where people and the like are present in the perimeter zone P, and measure the temperature of the surrounding air. These near-floor temperature sensors 11 are used as parameters for a temperature estimation model that estimates the near-floor temperature, and the measured value is used as one parameter to be used for other near-floor positions rather than for the near-floor position in the target space in which they are installed.

[0042] Hereinafter, in this specification, the temperature of the indoor air A1 near the floor will be referred to as the "near-floor temperature." In this specification, the "near-floor temperature" refers to the temperature of the air near the floor, between 0 cm and 10 cm, or the temperature of the air in the occupied area, between 10 cm and 170 cm. Furthermore, the "ceiling temperature" refers to the temperature of the air at a height of 170 cm or more above the floor and below the ceiling 3. In other words, the ceiling temperature sensor 9 may be installed at the height of the ceiling 3, or at a lower position (as long as it is at a height of 170 cm or more above the floor).

[0043] Here, the near-floor temperature sensor 11 that measures the near-floor temperature may be installed at a height of 0 cm or more but less than 10 cm when measuring the temperature of indoor air A1 on the floor surface or nearby, or at a height of 10 cm or more but less than 170 cm (occupied area) when measuring the temperature of indoor air A1. Alternatively, in the latter case, the near-floor temperature sensor 11 may be installed at a height of 0 cm or more but less than 10 cm, and the temperature at the desired height may be calculated based on the measurement values ​​from the ceiling temperature sensor 9 and the near-floor temperature sensor 11, with the difference being apportioned by the distance between the two sensors.

[0044] The installation position of the near-floor temperature sensor 11 in plan view is preferably inside the installation position of each perimeter fan 6 (see FIG. 1) (i.e., the near-floor temperature sensor 11 is located further inside the perimeter fans 6 when viewed from the window or wall). The perimeter fans 6 are usually positioned so that they take in room air A2 from the side facing the inside of the target space S and send it upward. For this reason, if the near-floor temperature sensor 11 is located inside the perimeter fans 6, the room air A2 near the floor surface taken in by the perimeter fans 6 will move and come into contact with the near-floor temperature sensor 11, allowing the near-floor temperature sensor 11 to preferably obtain the near-floor temperature.

[0045] For ease of explanation, Fig. 1 simply illustrates two air conditioners 1, one target space S, two air outlets 4, two variable air volume devices 5, one perimeter fan 6, and a total of two air inlets 7, 16; however, this is merely a schematic diagram for explanation purposes, and an actual air conditioning system may have a different number of air conditioners 1 and variable air volume devices 5 installed, or may be configured to direct supply air A1 to multiple target spaces S. The number of air outlets 4 installed per target space S may also be changed as appropriate depending on the size of the target space S, etc. In an actual air conditioning system, various other devices and sensors will be installed in addition to the devices shown here, but configurations that are not directly related to the spirit of the present invention have been omitted from the illustration as appropriate.

[0046] Such an air conditioning system can be installed throughout a building, as shown in Fig. 3, for example. On each floor, as shown in Fig. 2, a near-floor temperature sensor 11 is installed in a position near a portion of the variable air volume device 5 facing the perimeter zone P in a plan view. In the example shown here, one near-floor temperature sensor 11 is installed on each side of the perimeter zone P in a plan view (note that the "variable air volume device 5 facing the perimeter zone P" here roughly means "the variable air volume device 5 that supplies air to the perimeter zone P"). The temperature conditions in the perimeter zone P are easily affected by exterior loads (wind direction, sunlight, etc.) via windows and walls facing the outside air, and vary greatly depending on the orientation of the windows and walls even within the same building or target space S. However, it is thought that the temperature conditions will be roughly similar within the same room and facing the same direction relative to the building.

[0047] In other words, the perimeter zones P of each target space S on each floor can be divided into four main systems in principle according to the temperature characteristics that differ depending on the direction, and for example, a target space S that has windows and walls facing northeast and windows and walls facing northwest can be considered to have two perimeter zones: a northeast perimeter zone P and a northwest perimeter zone P (of course, this is not the case for floors or target spaces S with special shapes, and it is also possible to imagine cases where they can be considered to have more perimeter zones P). Then, a temperature estimation model M, described below, is generated for each perimeter zone P corresponding to each of these systems.

[0048] The above-described division of perimeter zones by direction and the creation of a temperature estimation model based on this division are merely examples. When actually generating a temperature estimation model, the division of perimeter zones and the creation of a temperature estimation model can be appropriately set depending on the temperature conditions at the site, etc.

[0049] 2 only shows the air volume variable device 5 and near-floor temperature sensor 11 installed in the perimeter zone P, but in reality, air volume variable devices 5 are also appropriately placed in the interior zone I, and of course, devices such as the perimeter fan 6 and ceiling temperature sensor 9 are also placed as shown in FIG. 1. Also, FIG. 3 shows a simplified building and air conditioning system as an overall view of the air conditioning system, and one air conditioner 1 (perimeter air conditioner 1b) and one variable air volume device 5 are shown on each floor, but it goes without saying that the number of floors in an actual building may differ from those shown here, and that more air conditioners 1 and variable air volume devices 5 may be installed on each floor. Of course, the floor plan of each floor and the placement of the variable air volume devices 5 in each target space S may differ from the configuration shown in FIG. 2.

[0050] On some of the floors that make up the building, near-floor temperature sensors 11 are installed near all of the variable air volume devices 5 that face the perimeter zone P, as shown in Fig. 4. This is to collect actual measured values ​​of near-floor temperatures that will be used as learning data D (see Fig. 3) for generating a temperature estimation model M, which will be described later.

[0051] As shown in FIG. 4 , the floors (hereinafter referred to as “reference floors”) on which near-floor temperature sensors 11 are installed for all variable air volume devices 5 facing the perimeter zone P may be, for example, specific floors among multiple floors with a common floor plan. This is because it is expected that the temperature conditions will be roughly similar if the floor plan is common. For example, if the 11th to 40th floors of a building have the same floor plan and the target floors are the 11th to 40th floors, the 25th floor may be set as the reference floor, and the measured near-floor temperatures collected on the 25th floor may be used as the learning data D to generate a temperature estimation model M, which may then be used to estimate the near-floor temperatures on the other floors from the 11th to 40th floors. Note that even if the floor plan is common, significant differences in height may result in corresponding differences in the temperature conditions. In such a case, for example, the 18th and 32nd floors may be set as different reference floors. That is, a first temperature estimation model is generated based on learning data collected using the 18th floor as a first reference floor and used to estimate near-floor temperatures on the other floors from the 11th to the 24th floors, and a second temperature estimation model is generated based on learning data collected using the 32nd floor as a second reference floor and used to estimate near-floor temperatures on the other floors from the 25th to the 40th floors. Alternatively, if the temperature situation changes significantly across a specific floor due to the influence of shadows from nearby buildings, for example, different temperature estimation models may be applied above and below that specific floor. In this way, the setting of the reference floor and the target to which a certain temperature estimation model is applied may be determined appropriately taking into account various conditions.

[0052] As shown in FIG. 3, the air conditioning on each floor is operated by an air conditioner 1, a variable air volume device 5, a perimeter fan 6, a ceiling temperature sensor 9, a near-floor temperature sensor 11, and a control device 10. The operating status of the air conditioning on each floor is monitored and controlled by a central monitoring device 12 connected to the control device 10 on each floor. The central monitoring device 12 is a device that comprehensively monitors the air conditioning system and other systems (electrical systems, etc.) of the entire building. A temperature estimation unit 13 is further connected to the central monitoring device 12, and a model generation unit 14 is connected to the temperature estimation unit 13. The temperature estimation unit 13 and the model generation unit 14 are information processing devices such as personal computers. As will be described later, the temperature estimation unit 13 estimates the near-floor temperature in the perimeter zone P (see FIGS. 1 and 2) on each floor. The model generation unit 14 generates a temperature estimation model M that the temperature estimation unit 13 uses to estimate the near-floor temperature. The temperature estimation model M is generated based on the learning data D and is a model that estimates the near-floor temperature at a certain location using parameters related to the near-floor temperature as explanatory variables. That is, for example, the near-floor temperature at multiple locations in a perimeter zone P on the north side of a specific target space S on a certain floor is estimated based on various parameters described below.

[0053] Furthermore, in the air conditioning system of this embodiment, solar radiation is used as one of the parameters for estimating the near-floor temperature (more specifically, as one of the explanatory variables input to the temperature estimation model M to estimate the near-floor temperature), and the measured value of solar radiation is input to the central monitoring device 12 from a pyranometer 100 installed on the roof of the building (see FIG. 3).

[0054] Furthermore, the air conditioning system of this embodiment includes a solar radiation estimator 101 for estimating solar radiation. The solar radiation estimator 101 calculates an estimated value of solar radiation at a target time under specific conditions based on measured values ​​of solar radiation input to the central monitoring device 12, extra-atmospheric global solar radiation, and the like. The model generator 14 and the temperature estimation model M use the estimated value calculated by the solar radiation estimator 101 instead of actual measurements from the pyranometer 100 as necessary to generate the temperature estimation model M and estimate near-floor temperatures at each location. Note that although the solar radiation estimator 101 is shown here as part of the temperature estimator 13, it may be, for example, an external device connected to the central monitoring device 12, or the solar radiation estimator 101 may be configured as part of the central monitoring device 12. The estimation of solar radiation by the solar radiation estimator 101 will be described in detail later.

[0055] The temperature estimation model M can use, for example, the following parameters to estimate the temperature near the floor. Outdoor temperature (values ​​obtained at multiple heights (e.g., rooftop, high-rise, and low-rise) may be used) Solar radiation (actual solar radiation measured by the pyranometer 100, or estimated solar radiation calculated by the solar radiation estimation unit 101) - Presence or absence of sunlight (is the target location in the sun or shade? This parameter will be explained in detail later) The operating status of the air conditioner 1 (off / cooling / heating, target value of supply air temperature, supply air volume. The value of the air conditioner 1 responsible for air conditioning of the location where the near-floor temperature is to be estimated may be used, or the value of another air conditioner 1 may be used.) Actual measured supply air temperature for air conditioner 1 - Heat outflow (the amount of heat flowing from the target space S to the outside. Can be calculated based on the temperature difference between the inside and outside of the room) Operating status of variable air volume device 5 facing perimeter zone P (on / off, set temperature, required air volume) Ceiling temperature sensor 9 measurement value (current value, and values ​​obtained in the past) Wind direction (values ​​obtained at multiple heights (e.g., rooftop, high-rise, and low-rise) may be used) Wind speed (values ​​obtained at multiple heights (e.g., rooftop, high-rise, and low-rise) may be used) ·Rainfall amount Measurement value of the near-floor temperature sensor 11 (measurement value of the near-floor temperature sensor 11 installed at a position other than the position where the near-floor temperature is to be estimated) Perimeter fan 6 operating status (on / off, airflow rate) Date (month and / or day) Time (hours and / or minutes) Day of the week (Sun, Mon, Tues, Wed, Thurs, Fri, Sat, or whether it's a holiday or a weekday)

[0056] An example of the input / output configuration of the temperature estimation model M is shown in Figure 5. The input explanatory variables are the outside air temperature (average value for the past 30 minutes) obtained from the central monitoring device 12, the amount of solar radiation (average value for the past 30 minutes), the presence or absence of solar radiation, the operating status of each air conditioner 1 in the perimeter zone P and the interior zone I, the actual measured values ​​of the supply air temperatures therein, the operating status of the variable air volume device 5 facing the perimeter zone P, the amount of heat flowing out, and the measurements of the ceiling temperature sensors 9 installed in various locations. These are used as input values, and the estimated value of the near-floor temperature at each target location is output.

[0057] Such a temperature estimation model M is created for each system unit of variable air volume device 5 so that the near-floor temperature at the corresponding position can be output for each perimeter air conditioner 1b. In the example shown here, a temperature estimation model M is assumed to be used for an air conditioning system equipped with a total of eight variable air volume devices 5, and is configured to output eight estimated near-floor temperatures corresponding to the positions of the eight variable air volume devices 5. Note that for positions where near-floor temperature sensors 11 are installed, it is sufficient to use the measured values ​​from those sensors, so there is no need to estimate the near-floor temperature using a temperature estimation model.

[0058] Here, in the air conditioning system of this embodiment, which controls the variable air volume device 5 based on an estimated value of the near-floor temperature as described below, research by the present inventors has revealed that if the measured values ​​of the ceiling temperature sensors 9 installed in various locations are used as explanatory variables to be input into the temperature estimation model M, the near-floor temperature at the target location can be estimated with particularly high accuracy.

[0059] The air conditioning system of this embodiment is intended to estimate the near-floor temperature and control it to approach a target temperature (near-floor temperature target value). However, the direct control target is the air temperature near the ceiling (ceiling temperature). By manipulating the set value of this ceiling temperature (ceiling temperature target value), the near-floor temperature is indirectly maintained at a suitable temperature. As will be explained in detail later, this type of control first sets the ceiling temperature target value based on the near-floor temperature estimated at a certain point in time. The air volume of the variable air volume device 5 is adjusted accordingly, and the supply air is sent toward the perimeter fan 6 and supplied to the perimeter zone P, thereby adjusting the near-floor temperature in the perimeter zone P. This process is repeated until the near-floor temperature gradually approaches the target value. In other words, the temperature control performed near the ceiling is reflected in the near-floor temperature with a certain time lag. As such, since the near-floor temperature is affected by the ceiling temperature through the operation of the air conditioning system itself, the measurement value of the ceiling temperature sensor 9 is an effective explanatory variable for estimating the near-floor temperature. Furthermore, when estimating the current near-floor temperature, using the measurement value of the ceiling temperature sensor 9 at a certain point in time in the past is particularly effective in accurately estimating the near-floor temperature, as that value (the ceiling temperature at that time) has a significant influence on the current near-floor temperature value.

[0060] In the example shown in FIG. 5, current and past values ​​from eight ceiling temperature sensors 9 are input as explanatory variables into the temperature estimation model M, and estimates of near-floor temperatures at eight locations are calculated. Past values ​​are calculated in five-minute increments, covering up to 30 minutes prior. Note that the number of ceiling temperature sensor measurements input into the temperature estimation model (eight in FIG. 5) and the number of target estimates (near-floor temperatures) output (also eight in FIG. 5) are merely examples and may be changed as appropriate depending on the number of ceiling temperature sensors installed in an actual system and the number of locations for which near-floor temperatures are to be estimated. Furthermore, the number of past values ​​to input, the number of minutes prior to use, and the interval between past values ​​may be set as appropriate when designing the air conditioning system or model.

[0061] Here, when estimating the near-floor temperature at a certain location, as mentioned above, it is also possible to use the actual measurement value of the near-floor temperature sensor 11 installed at another location. For example, when estimating the near-floor temperature at a location on the right side of Figure 2, it is possible to use the measurement value of the near-floor temperature sensor 11 installed on the same side of the same room (within the perimeter zone P of the same system) as an explanatory variable. However, in the target space S, such as a room in an office building, partitions are installed on the floor, and equipment that emits heat is located, so the thermal conditions often differ significantly depending on the location, even in the same perimeter zone within the same room. As a result, if the actual measurement value of the near-floor temperature sensor 11 at another location is input as an explanatory variable, the correlation between the near-floor temperature at the target location and the actual measurement value at that other location is low, and the output estimated value may deviate from the actual value. On the other hand, the thermal conditions near the ceiling are less likely to change or vary from location to location due to factors such as partitions or heat exhaust from equipment compared to those near the floor. Therefore, although it depends on the actual conditions in the target space S, measurements from ceiling temperature sensors 9 installed on or near the ceiling 3 are likely to be more useful as explanatory variables. For example, by using a model that takes into account the trends and patterns of ceiling temperature, such as the temperature estimation model M shown in Figure 5, it is possible to accurately estimate the near-floor temperature. However, for example, if you want to estimate the near-floor temperature at a certain location and it is known that the thermal conditions at that location are similar to the thermal conditions at another location where a near-floor temperature sensor 11 is installed, it may be effective to use the measurements from the near-floor temperature sensor 11 at that other location as explanatory variables. Furthermore, while using the current and past values ​​of the ceiling temperature sensor 9 as explanatory variables, measurements from near-floor temperature sensors 11 at other locations may also be used to further improve accuracy.

[0062] The inventors of the present application have also found that the near-floor temperature can be estimated more accurately by further using values ​​related to solar radiation (measured or estimated solar radiation, or the presence or absence of solar radiation at the target location) as explanatory variables, as described below. Here, when considering both the amount of solar radiation and the presence or absence of solar radiation, for example, a value obtained by multiplying the measured or estimated amount of solar radiation by the presence or absence of solar radiation at the target location (zero or one) can be used as an explanatory variable. Alternatively, the presence or absence of solar radiation may be prepared as a data table separate from the amount of solar radiation and used as separate explanatory variables. In particular, in the latter case, by inputting the amount of solar radiation, which is a quantitative variable, and the presence or absence of solar radiation, which is a qualitative variable, into the temperature estimation model M as separate explanatory variables, complex simulations are not required, reducing the effort required for creating the model and preparing the data, and enabling the construction of a highly accurate temperature estimation model M. Alternatively, only either the measured or estimated amount of solar radiation or the presence or absence of solar radiation at the target location may be used as an explanatory variable related to solar radiation.

[0063] Of the above parameters, the outdoor temperature, rainfall, and solar radiation may also affect the predicted value of the near-floor temperature with a time lag. Therefore, for these parameters, the actual measured values ​​prior to the time point to be predicted (for example, 10 to 30 minutes prior) may also be used as parameters.

[0064] The operating state of the perimeter fan 6 can also be reflected in the temperature estimation model M. In this case, taking into consideration that the thermal situation in the perimeter zone P is particularly greatly affected by the operating state of the perimeter fan 6, instead of using the operating state of the perimeter fan 6 as an explanatory variable, or in addition to using the operating state of the perimeter fan 6 as an explanatory variable, the temperature estimation model M itself may be used differently depending on the operating state of the perimeter fan 6. That is, for example, the temperature estimation model M when the operating state of the perimeter fan 6 is on (temperature estimation model M ON ) and the temperature estimation model M when it is off (temperature estimation model M OFF) are generated, and a temperature estimation model M ON , M OFF The two are switched.

[0065] Furthermore, instead of the parameters exemplified here, equivalent parameters with different units or definitions or related parameters can be used. Furthermore, in addition to the parameters exemplified above, some other parameter related to the near-bed temperature may also be used. Furthermore, it goes without saying that processing such as normalization or centering of values ​​may be performed as appropriate.

[0066] The learning data D is a set of data related to the operation of air conditioning accumulated in the central monitoring device 12 over an appropriate period of time (for example, from a few days to a year) from the control devices 10 on each floor (particularly the reference floor) and other sensors (not shown), and is stored in the temperature estimation unit 13. The model generation unit 14 generates a temperature estimation model M based on the learning data D and stores it in the temperature estimation unit 13. The temperature estimation unit 13 also includes a correction value calculation unit 15. The correction value calculation unit 15 calculates a temperature correction value based on the near-floor temperatures at each location in the perimeter zone P on each floor calculated using the temperature estimation model M. The temperature correction value is a correction value that is added to the set temperature for each variable air volume device 5 facing the perimeter zone P. The role of this temperature correction value will be explained again later.

[0067] The generation of the temperature estimation model M by the model generation unit 14 will now be described. Various types of models, such as linear regression, ridge regression, gradient boosting, and random forest, can be used as the temperature estimation model M. However, if the temperature estimation model M is generated using a multi-input, multi-output neural network, for example, the near-floor temperature can be estimated with high accuracy. In particular, if a two-layer linear neural network is used, high-accuracy estimation is possible even in a short learning period of about two weeks without overlearning. Alternatively, high-accuracy estimation is also possible using a three-layer neural network or a time-series neural network such as LSTM.

[0068] The learning data D used to generate the temperature estimation model M is a data set that records various parameters (for example, the parameters exemplified above) related to the near-floor temperature at various times or positions when an air conditioning system such as that shown in Figures 1 to 3 is actually operated, and actual measured values ​​of the near-floor temperature. The actual measured values ​​of the near-floor temperature are obtained from near-floor temperature sensors 11 installed in various locations in the perimeter zone P of the reference floor. Some of the parameters related to the near-floor temperature can be obtained from the air conditioner 1, the ceiling temperature sensor 9, the control unit of the variable air volume device 5, etc.

[0069] The learning data D records the above parameters at a plurality of points in time, as well as actual measurements of the near-floor temperatures at those points in time acquired on the reference floor.

[0070] The collection of the learning data D may be performed, for example, on the condition that all of the following conditions are met: Whether heating operation is in progress. In the perimeter zone P, the cold air problem described above occurs only when the outdoor temperature is low and heating operation is in progress. In other words, it is only during heating operation that it is necessary to estimate the temperature near the floor, calculate a temperature correction value based on that, and operate using that temperature correction value. Therefore, only parameters recorded during heating operation are used as learning data D. Whether the outdoor temperature is lower than a preset threshold. The above problem occurs only when the outdoor temperature is lower than the indoor air temperature by a certain amount, so estimating the near-floor temperature and calculating the temperature correction value are only necessary when the outdoor temperature is low. Therefore, learning data D is collected only when the outdoor temperature is below the threshold. Whether the variable air volume device 5 is in operation. If the variable air volume device 5 itself is off, it will not operate using the temperature correction value, so there is no need to estimate the temperature near the floor or calculate the temperature correction value. Therefore, if the variable air volume device 5 is not in operation, learning data D is not collected.

[0071] However, this is not the case when the amount of acquired learning data is small, and learning data D may be collected as appropriate even when at least some of the above conditions are not met.

[0072] The model generation unit 14 performs machine learning using the training data D to generate a temperature estimation model M that estimates the near-floor temperature at each location in the perimeter zone P on each floor based on various parameters. When generating the temperature estimation model M using a neural network, the model generation unit 14 appropriately selects and discards multiple parameters from the various parameters described above recorded in the training data D, and also performs processing and weighting such as normalization and centering on each parameter as necessary to form a pattern of options. The near-floor temperature estimated from the formed pattern is compared with the actual near-floor temperature acquired by the near-floor temperature sensor 11 on the reference floor, and the pattern is repeatedly modified, until a highly accurate temperature estimation model M is generated through machine learning. The generated temperature estimation model M is a model that estimates the near-floor temperature at a certain location in the target air conditioning system based on parameters related to the near-floor temperature.

[0073] In the air conditioning system described above, an example of a procedure for operating the variable air volume device 5 facing the perimeter zone P using the estimated near-floor temperature and a temperature correction value calculated based on this will be described with reference to the flowchart in Figure 6.

[0074] The temperature estimation unit 13 (see FIG. 3) acquires information about the operating status of each part of the air conditioning system via the central monitoring device 12 (step S1). The information acquired here includes, for example, the supply air temperature (temperature of supply air A1) in each air conditioner 1, the actual measured temperature of indoor air A2 acquired by ceiling temperature sensors 9 installed in the target space S on each floor, the set temperatures in each target space S and variable air volume devices 5, the blowing air volume (damper opening) of each variable air volume device 5, the outdoor air temperature measured by a temperature sensor (not shown), the date and time, the amount of solar radiation measured by pyranometer 100, the presence or absence of solar radiation at each position, etc. The near-floor temperature at each position in the perimeter zone P of each target space S estimated using the temperature estimation model M is also acquired in step S1.

[0075] Next, based on the acquired information, it is determined whether or not to operate each air conditioner 1 constituting the air conditioning system using a temperature correction value at the variable air volume device 5 located downstream of it (step S2). In this step S2, whether or not to operate using a temperature correction value is determined based on, for example, whether or not all of the following conditions are met (note that the conditions listed below are merely examples, and other conditions may also be used): Heating operation is in progress. In the perimeter zone P, the above-mentioned cold air problems only occur when the outside temperature is low and heating operation is in progress, so operation using the temperature correction value is only performed when heating operation is in progress. The outdoor temperature is lower than a preset threshold. The above problem occurs only when the outdoor temperature is relatively low compared to the indoor air temperature, so operation using the temperature correction value is performed only when the outdoor temperature is below the threshold. Note that the threshold used here can be set to, for example, two levels, and the higher threshold is used when changing the temperature correction value from enabled to disabled, and the lower threshold is used when changing it from disabled to enabled, which is preferable because it prevents excessively frequent switching.

[0076] Once it has been determined whether or not to use a temperature correction value for each air conditioner 1, the use of a temperature correction value and specific numerical values ​​for each downstream variable air volume device 5 are set for each individual air conditioner 1. First, for one air conditioner 1, it is determined in step S3 whether or not it was decided in the previous step S2 that the downstream variable air volume device 5 will operate using a temperature correction value.

[0077] If none of the variable air volume devices 5 downstream of the target air conditioner 1 are to operate using a temperature correction value, then each variable air volume device 5 downstream of that air conditioner 1 is set to not use a temperature correction value. In step S4, one variable air volume device 5 is set to not use a temperature correction value (temperature correction value = 0). In the next step S5, it is determined whether any variable air volume devices 5 downstream of the target air conditioner 1 have not yet had a temperature correction value set. If there is an unset variable air volume device 5, then the process moves on to another variable air volume device 5 (step S6), and sets the temperature correction value = 0 for that variable air volume device 5 (step S4). This process is repeated until temperature correction values ​​have been set for all variable air volume devices 5 downstream of the target air conditioner 1, and then the process moves on to step S7.

[0078] In step S7, an upper limit value for the supply air temperature is set for the target air conditioner 1. The upper limit value for the supply air temperature in the air conditioner 1 is set by the central monitoring device 12, and normally the value set in this central monitoring device 12 can be used as is, but in this embodiment, the setting of this upper limit value is changed only in specific cases. How the upper limit value for the supply air temperature is changed will be explained later when explaining step S16, but in step S7, the value set in the central monitoring device 12 is used. Once step S7 is complete, the process proceeds to step S17, which will be described later.

[0079] If it is determined in step S3 that it has been decided that operation using a temperature correction value will be performed for at least some of the variable air volume devices 5 downstream of the target air conditioner 1, the process proceeds to step S8. In this step S8, it is determined individually whether or not to enable setting of the temperature correction value for each variable air volume device 5 located downstream of the target air conditioner 1. The determination in step S3 is made, for example, based on whether or not all of the following conditions are met: The variable air volume device 5 in question is not experiencing any malfunctions. The corresponding variable air volume device 5 is turned on.

[0080] Once it has been determined whether the temperature correction value is enabled or disabled for each variable air volume device 5, the process proceeds to step S9. In step S9, it is determined whether operation using the temperature correction value was enabled for a particular variable air volume device 5 in the previous step S8. If the temperature correction value for that variable air volume device 5 has been set to be disabled, the process proceeds to step S10, where the temperature correction value is set to 0 for that variable air volume device 5. If it is determined in step S9 that the temperature correction value for that variable air volume device 5 has been set to be enabled, the process proceeds to step S11, where the temperature correction value is calculated.

[0081] The calculation of this temperature correction value will be explained below, but first, the significance of the temperature correction value will be explained. As mentioned above, the temperature correction value is a correction value added to the set temperature for each variable air volume device 5 facing the perimeter zone P. This is set to correct the difference between the measured temperature and the perceived temperature, taking into account the fact that the ceiling temperature sensor 9 is installed at a height above the floor. In the air conditioning system shown in FIG. 1, the actual measurement value of the indoor air A2 is recognized as the measurement value of the ceiling temperature sensor 9 installed at the height of the ceiling 3. However, since people are located near the floor in the target space S, as explained above, a discrepancy is likely to occur between the perceived temperature perceived by people and the temperature recognized as the measurement value of the ceiling temperature sensor 9, especially in the perimeter zone P.

[0082] During heating, the air conditioning system, which is composed of the control device 10, the air conditioner 1, and the variable air volume device 5, adjusts the temperature (ceiling temperature T1) measured by the ceiling temperature sensor 9 until it reaches the set value (room temperature set value T SP If the ceiling temperature T1 is lower than the indoor temperature setting T SP However, the actual temperature of the indoor air A2 near the floor is often lower than the ceiling temperature T1. SP Even if the temperature is close enough to the floor, there may be a difference between the temperature near the floor (T2) and the ceiling temperature T1, which may result in a situation where people near the floor are not able to get a satisfactory level of warmth. To correct this, in this embodiment, the estimated value of the temperature near the floor T2 (estimated temperature near the floor T 2E ) is calculated (step S1), and the target value of the near-floor temperature T2 (target near-floor temperature T 2SP The temperature correction value (β) is calculated according to the deviation between these values ​​(control deviation E2), and the set temperature value (ceiling temperature target value T 1SP The original set value (room temperature set value T SP ) and the temperature correction value β are added to obtain the ceiling temperature target value T 1SP is calculated, and each device is operated using this as a target value, thereby correcting the temperature situation near the floor surface.

[0083] The management of the temperature near the floor using such a temperature correction value β can also be explained by a diagram such as that shown in Figure 7. In Figure 7, the left vertical line indicates the temperature near the floor T2, and the right vertical line indicates the ceiling temperature T1. For example, when the indoor temperature setting value T SP If the ceiling temperature T1 is 23°C and the air conditioning system is operated based on this, the temperature difference between the ceiling and the upper and lower areas near the floor will cause the near-floor temperature T2 to be, for example, 20°C. 2E ) is the indoor temperature setting value T SP (or the reference value T0 explained next) 2SPSince it is lower than the temperature deviation E2, the temperature correction value β is calculated according to this difference (corresponding to the control deviation E2), and this is used as the indoor temperature setting value T SP The setting value (ceiling temperature target value T 1SP ) is set on the ceiling side and set at 26°C. As a result, the temperature difference near the floor is compensated for, resulting in a comfortable temperature (23°C).

[0084] However, the temperature difference is not uniform and can change depending on the conditions. For example, if sunlight enters the perimeter zone P and warms the floor, or if the outside temperature rises and cold air intrusion decreases, the temperature difference will become smaller. In this case, if the estimated near-floor temperature T 2E and the indoor temperature setting value T SP If the temperature correction value β is set to a uniform value (for example, 3°C) without using the deviation from the set value, the temperature near the floor will rise too much. Conversely, if the temperature difference is large, the temperature near the floor may not be sufficient even if the temperature correction value β is added to the set value. Therefore, as explained above, the temperature correction value β is set to the estimated temperature near the floor T 2E and the indoor temperature setting value T SP If the temperature correction value is calculated each time based on the deviation from the set value, the temperature conditions near the floor surface can be maintained appropriately. In other words, instead of setting a uniform temperature correction value to the set value, the temperature correction value can be adjusted according to the temperature conditions near the floor surface.

[0085] An example of a procedure for calculating the temperature correction value β by the correction value calculation unit 15 will be described below. First, a control deviation (referred to as E2) that is the basis for the temperature correction value β is calculated by the following formula. [Number 1] Control deviation E2 = target temperature near the floor T 2SP - Estimated floor temperature T 2E ……(1) Near-floor temperature target value T 2SP =Reference value T0+k·α-γ ……(2)

[0086] The control deviation E2 is the target value of the near-floor temperature T 2SP and the estimated temperature near the floor T 2E This is the deviation between the two, and the air conditioning system is operated in a way that eliminates this deviation.

[0087] Target value T of the temperature near the bed 2SP is a value set as the target value of the temperature T2 near the bed, which is the control target of this system, and is calculated based on the reference value T0 and other values as shown in equation (2). The reference value T0 is the set temperature value in the variable air volume device 5 determined in the central monitoring device 12, for example, 23°C. α is a correction value added to the set temperature in the variable air volume device 5 by the operation of the person in the target space S, and can take, for example, two values of 0 and 1 (the initial value is 1). k is a coefficient that determines the weighting of α when calculating the control deviation E2. γ is a value set as an adjustment parameter in the air conditioning system in consideration of the difference in air temperature between the height close to the bed (the height where the air temperature is estimated and the area to be operated) and the living area. Indoor temperature set value T SP is calculated by adding the correction value α from the indoor side to the reference value T0 determined centrally, for example, in the form of the reference value T0 + k·α. Therefore, further considering the adjustment parameter γ, the target value T of the temperature near the bed is set as shown in the above equation (2). 2SP As the adjustment parameter γ, values such as 0.5°C or 1°C are set in consideration of the actual air situation in the perimeter zone P.

[0088] After calculating the control deviation E2, for example, according to the relationship shown in the following mathematical formula, the effective control deviation E 2eff is obtained. [Equation 2] E 2eff = E2 + δ (E2 ≤ -δ) E 2eff = 0 (-δ < E2 < δ) E 2eff = E2 - δ (δ ≤ E2)

[0089] δ is a value set as the temperature dead zone. That is, centered on the point where E2 = 0, in the range of -δ < E2 < δ, E 2eff = 0, and in other ranges, that is, in the range where the absolute value of the control deviation E2 is greater than or equal to the threshold value (δ), the effective control deviation E linearly changes according to the increase or decrease of the control deviation E2 2effThis is effective in preventing frequent switching of the operating state when the absolute value of the control deviation E2 is small.

[0090] The effective control deviation E 2eff The temperature correction value β is determined by the relationship shown in Figure 8. In the example shown here, PI control is used, and the effective control deviation E 2eff When the temperature is -0.5℃ or more, the effective control deviation E 2eff The temperature correction value β is determined so as to be proportional to the reference value T0 in the central monitoring device 12 (in the illustrated example, the proportionality coefficient is 1.25). 2SP The estimated temperature near the floor, T 2E If the temperature is low, the temperature correction value β is added to the set temperature of the variable air volume device 5 at the corresponding position to set the set temperature (ceiling temperature target value T 1SP ) and increase the amount of air blown out.

[0091] In the example shown here, the upper limit of the temperature correction value β is set to +2°C, and the effective control deviation E 2eff If the temperature is 2°C or higher, the temperature correction value is uniformly +2°C. 2eff The calculation formula for the effective control deviation E 2eff The relationship between the temperature correction value β and the control deviation E2 or the effective control deviation E 2eff , a temperature correction value β may be determined.

[0092] The target temperature near the floor, T, is calculated using the above procedure. 2SP and ceiling temperature target value T 1SP For example, it is expressed as follows: The following equation 3 is for the case where k=1 and γ=1. [Number 3] T 2SP =T0+α-1 ……(3) T 1SP =T0+α+β ……(4)

[0093] Once the temperature correction value has been set for the target variable air volume device 5 (steps S10 and S11), the process proceeds to step S12. In step S12, it is determined whether there is a variable air volume device 5 downstream of the currently target air conditioner 1 for which a temperature correction value has not yet been set. If there is an unset variable air volume device 5, the target is changed to another variable air volume device 5 (step S13), and a temperature correction value is set for that device (steps S9 to S11).

[0094] Once the temperature correction values ​​have been set for all variable air volume devices 5 downstream of the target air conditioner 1, the process proceeds to step S14. In step S14, it is determined again whether any of the variable air volume devices 5 downstream of the target air conditioner 1 are currently operating with the temperature correction value enabled. If the determination here is NO, the process proceeds to step S15. In step S15, the upper limit of the supply air temperature for the target air conditioner 1 is set to the same value as the setting in the central monitoring device 12.

[0095] If the determination in step S14 is YES, that is, if there is a variable air volume device 5 operating with the temperature correction value enabled downstream of the target air conditioner 1, the process proceeds to step S16. In step S16, an upper limit value for the supply air temperature is set for the target air conditioner 1 separately from the set value in the central monitoring device 12.

[0096] As mentioned above, the problem of cold drafts in the perimeter zone P occurs when the outdoor air temperature is low. For this reason, in this embodiment, operation is performed using a temperature correction value, with the outdoor air temperature being lower than a threshold value as one of the conditions. Meanwhile, in the air conditioning system shown in FIGS. 1 to 3, the amount of heat supplied to the target space S is determined by the blown air volume of each variable air volume device 5 and the supply air temperature from the air conditioner 1. In other words, if the supply air temperature of the air conditioner 1 is high, the blown air volume required to supply the same amount of heat will be correspondingly smaller. If the blown air volume from the variable air volume device 5 facing the perimeter zone P is reduced, the difference in air temperature between the regions of the perimeter zone P will increase, increasing the possibility of a cold draft occurring.

[0097] Therefore, when operating using a temperature correction value, in step S16, the upper limit of the supply air temperature is set based on the outside air temperature, for example, using the relationship shown in Figure 9. In this example, when the outside air temperature is 10°C or lower, the upper limit of the supply air temperature is reduced. When the outside air temperature is between 10°C and 0°C, the supply air temperature is reduced linearly with changes in the outside air temperature (in the case of Figure 9, the proportionality coefficient is 0.6). When the outside air temperature is 0°C or lower, the upper limit of the supply air temperature is set uniformly to 25°C. This prevents the supply air temperature from rising too much when the outside air temperature is low, ensuring the volume of air blown from the variable air volume device 5 facing the perimeter zone P. Note that because the determination that the outside air temperature is lower than the threshold value was already made in the previous step S2, it is sufficient to specifically set the upper limit of the supply air temperature in step S16.

[0098] Once the upper limit of the supply air temperature has been set (step S7, S15, or S16), the process proceeds to step S17. In step S17, it is determined whether any downstream variable air volume devices 5 have air conditioners 1 remaining for which operation settings using the temperature correction value have not been completed. If so, the process moves to another air conditioner 1 (step S18), and the downstream variable air volume devices 5 are configured for operation using the temperature correction value and an upper limit of the supply air temperature is set (steps S3 to S16). Once the settings of the downstream variable air volume devices 5 and upper limit of the supply air temperature have been completed for all air conditioners 1, the process proceeds to step S19. In step S19, the settings related to the operating conditions determined in steps S2 to S18 are input to the central monitoring unit 12. The central monitoring unit 12 operates each part of the air conditioning system based on the recorded settings.

[0099] In this way, the air conditioning system of this embodiment estimates the near-floor temperature at each position in the perimeter zone P. If the estimated near-floor temperature for a certain position is low, a temperature correction value is added to the set temperature of the air volume variable device 5 facing that position (step S11), and the set temperature is raised and the system operates. The air volume variable device 5 appropriately adjusts the blown airflow in accordance with the raised set temperature, increasing the amount of air sent from the air outlet 4 to the perimeter zone P. Furthermore, if the outside air temperature is low, the upper limit of the supply air temperature from the air conditioner 1 is lowered (step S16), ensuring a sufficient blown airflow from the air volume variable device 5. This allows the indoor air A2 (see FIG. 1) supplied from the air volume variable device 5 to reach the vicinity of the floor and the perimeter fan 6. In this way, the temperature distribution near the floor is corrected based on the estimated near-floor temperature, as shown in FIG. 10. That is, the warm air (room air A2) that reaches the perimeter fan 6 from the variable air volume device 5 is sent to the upper intake port 16 and taken in as exhaust air A4, creating an upward flow of warm air, which blocks the cold air that occurs at the outermost perimeter of the perimeter zone P and prevents cold drafts from entering the interior. In the perimeter zone P and the interior zone I further inside, the warm air fills up to the floor level, achieving optimal air conditioning.

[0100] Such control is possible, for example, by providing near-floor temperature sensors 11 at positions corresponding to all of the variable air volume devices 5 located near the perimeter zone P for the target space S on all floors of a building, as shown in Figure 4, and measuring the near-floor temperature at each location. However, if near-floor temperature sensors 11 were installed in the entire perimeter zone P on each floor of the building, the number of near-floor temperature sensors 11 would be enormous, which would increase installation costs and make management cumbersome. In this embodiment, by estimating the near-floor temperature in the perimeter zone P using software, it is possible to reduce the cost of installing sensors and control the airflow rate from the variable air volume device 5 based on the near-floor temperature.

[0101] As explained above, this embodiment employs a control method using a temperature correction value, and this control method can be conceptualized as shown in the block diagram of Fig. 11. First, in a variable air volume type air conditioning system (see Fig. 1), the direct control target is the temperature of the indoor air A2 near the ceiling 3 (ceiling temperature T1) (shown as block B1 in Fig. 11), which is grasped as the measurement value of the ceiling temperature sensor 9 installed near the ceiling 3 (block B2). This measurement value and the target ceiling temperature T 1SP (Block B3) (Block B4) to calculate the deviation (control deviation) between them, and the variable air volume device 5 (Block B5) adjusts the damper opening (Block B6) to eliminate this control deviation, thereby adjusting the amount of indoor air A2 supplied.

[0102] Furthermore, the air conditioner 1 (block B7), which supplies conditioned supply air A1 to the air volume variable device 5 (block B5), adjusts the fan air volume (block B8) according to the air volume required by the air volume variable device 5, and sends out an appropriate volume of supply air A1 (block B6) to the downstream air volume variable device 5. Additionally, the air conditioner 1 (block B7) adjusts the opening of the refrigerant flow path (block B9) using load reset control to adjust the temperature of supply air A1.

[0103] In this way, the ceiling temperature T1 (block B1), which is the object of control, is adjusted through the air volume and temperature of the supply air A1. The ceiling temperature T1 (block B1) is further influenced by disturbances (block B10), but the resulting ceiling temperature T1 (block B1) is detected by the ceiling temperature sensor 9 (block B2), and the target ceiling temperature T 1SP (Block B3; see equation (3) above) (Block B4), and control is performed in a direction that also eliminates fluctuations due to disturbances.

[0104] The control method using the temperature correction value can be said to be cascade control in which another loop (the loop outside the area surrounded by the dashed line, called the major loop) is added to surround the loop formed by blocks B1 to B10 (called the minor loop, shown as the area surrounded by the dashed line in Figure 11).

[0105] That is, the near-floor temperature T2 (block C1) is set as another control target related to the ceiling temperature T1 (block B1), and this is estimated using the temperature estimation model M, and the output near-floor temperature estimated value T 2E (Block C2) is the target temperature near the floor T 2SP (Block C3; see formula (4) above) (Block C4). Then, the temperature correction value β (Block C5) calculated based on the deviation between these two values ​​(control deviation E2) is compared with the ceiling temperature target value T 1SP (See equation (3) above.) In the minor loop, the ceiling temperature target value T 1SP The temperature of the indoor air A2 near the ceiling 3 (ceiling temperature T1; block B1) is adjusted based on the temperature (block B3), and this adjusts the temperature of the indoor air A2 near the floor (near-floor temperature T2; block C1). The near-floor temperature T2 (block C1) is also affected by disturbances (block C6), but the control by the major loop containing the minor loop works in the direction of eliminating this disturbance (block C6) by PI control via a temperature correction value. In other words, when a disturbance (block C6) occurs due to a change in the exterior load, the estimated near-floor temperature T 2E (Block C2) changes, and the target temperature near the floor T 2SP A deviation (control deviation) E2 occurs with respect to (Block C3), and the temperature correction value β as the manipulated variable is changed in the direction to cancel out the deviation E2, and is input to the minor loop.

[0106] With this type of control method, it is possible to appropriately correct the temperature situation near the floor by adding a temperature correction value based on the temperature near the floor to the target set temperature, without making any changes to the system (minor loop) that has been established as a variable air volume air conditioning system.

[0107] This type of cascade control also has advantages in terms of controllability. If a control configuration were adopted in which disturbances (blocks B10 and C6) were eliminated using a single loop, the control period would be at least five minutes, considering the data acquisition period in the central monitoring device 12 and the time required to estimate the near-floor temperature, resulting in a large deviation calculated within the loop. Therefore, as in this embodiment, cascade control is configured, consisting of a minor loop that controls the air volume of the variable air volume device 5 and the supply air temperature of the air conditioner 1, and an outer major loop that controls the near-floor temperature estimate. While the control period of the major loop is approximately five minutes, the control period of the minor loop is several seconds to several tens of seconds, and the deviation calculated in block B4 is small. Because the disturbance (block B10) input to the minor loop is controlled within the minor loop, its impact on the control variable of the major loop is small. In this way, the influence of disturbances can be effectively suppressed, improving controllability.

[0108] It should be noted that the system configuration of the air conditioning system, the generation of the temperature estimation model M, the estimation of the near-floor temperature, the procedure for calculating the temperature correction value, etc. described here are merely examples. As long as the temperature estimation model M can be generated from the various parameters related to the near-floor temperature as described above, the near-floor temperature can be estimated using this, and the operation of the variable air volume device 5 can be adjusted, the system configuration, various procedures, etc. can be modified in various ways.

[0109] For example, although the configuration illustrated here is one in which the temperature estimation unit 13 is connected to the central monitoring device 12, and the model generation unit 14 is connected to the temperature estimation unit 13, the connection relationship between the devices that make up the system can be changed as appropriate. In addition to controlling the operating status of each air volume variable device 5 via the central monitoring device 12, for example, a temperature correction value may be set directly from the temperature estimation unit 13 for each air volume variable device 5. Furthermore, instead of correcting the set temperature, the blown air volume may be corrected directly.

[0110] Furthermore, when controlling the air volume blown out from the variable air volume device 5, the supply air temperature from the air conditioner 1, and the operating conditions of each part, it goes without saying that other factors not described above may be taken into account as appropriate.

[0111] Here, estimation of the amount of solar radiation used as a parameter will be described. In the air conditioning system of this embodiment, as described above, the amount of solar radiation is used to control the air conditioning. At this time, it is naturally important to grasp as accurate a value as possible for the amount of solar radiation. The amount of solar radiation can be measured by a pyranometer 100 (see FIG. 3). However, for example, as shown in FIG. 3, if there are obstructions such as those indicated by symbols O1 and O2 around the pyranometer 100, shadows may be cast on the pyranometer 100 by the obstructions O1 and O2 depending on the location of the obstructions O1 and O2 and the season, which may cause the measured value of the amount of solar radiation to deviate from the actual amount of solar radiation.

[0112] Therefore, the air conditioning system of this embodiment is provided with a solar radiation amount estimation unit 101 for estimating the amount of solar radiation, and under certain conditions, the estimated value calculated by the solar radiation amount estimation unit 101 is adopted as the amount of solar radiation instead of the actual value measured by the pyranometer 100. The estimation of the amount of solar radiation by this solar radiation amount estimation unit 101 will be described below.

[0113] The graph in Figure 12 shows an example of the amount of solar radiation measured by the pyranometer when it is assumed that there is no obstruction around the pyranometer. Each curve in the graph shows the change over time in the amount of solar radiation observed on a different day (dates D1 to D3). It is assumed that the weather was sunny throughout the day on each of the days on which measurements were taken. When there is no obstruction around the pyranometer, the amount of solar radiation changes in a mountain shape as shown by each curve.

[0114] However, if there are obstructions O1 and O2 around the pyranometer 100 as shown in Fig. 3, a shadow is cast at the position of the pyranometer 100, which may affect the actual measured value of the amount of solar radiation measured by the pyranometer 100. An example of fluctuations in the actual measured value of the amount of solar radiation in such a case is shown in Fig. 13. Note that here, it is assumed that the obstructions O1 and O2 are, for example, other buildings or elevator shafts, but of course other devices, structures, terrain, etc. can also be obstructions.

[0115] For example, on date D1, the sun is high enough in the sky that the amount of solar radiation is not affected by surrounding objects and exhibits the same fluctuations as when there is no obstruction (see Figure 12). However, on another date, D2, which is closer to the winter solstice than date D1, a shadow caused by obstruction O1 is cast at the position of pyranometer 100 in the afternoon (here, from around 1:30 PM to around 2:00 PM), reducing the amount of solar radiation measured. On another date, D3, which is even closer to the winter solstice, a shadow caused by obstruction O2 is cast at the position of pyranometer 100 in the morning (here, from around 8:30 AM to around 9:30 AM), reducing the amount of solar radiation measured. Furthermore, the time period during which the shadow caused by obstruction O2 is cast is also extended in the afternoon (from around 1:00 PM to 2:40 PM).

[0116] During the time when the shadow is cast on the pyranometer 100 (hereinafter, in this specification, the time when the shadow is cast on the target position is referred to as the "shadow time"), the actual amount of solar radiation is, for example, as shown by the dashed line in Fig. 13. However, due to the influence of the obstructions O1 and O2, a value that is less than the actual amount of solar radiation is grasped as the actual measured value during this shadow time.

[0117] Therefore, using the method described below, during the shadow hours, the amount of solar radiation is estimated as close as possible to the actual value as shown by the dashed line in Figure 13, and this estimated value is used as the amount of solar radiation to operate the air conditioning system.

[0118] The solar constant is defined as the energy received per unit area and time by a surface perpendicular to the sunlight at the top of the atmosphere (about 8 km above the Earth's surface) when the Earth is at an average distance (r) from the Sun, and is approximately 1,367 W / m 2 ]. The extra-atmospheric global solar radiation Q0 corresponding to a point on Earth (longitude λ, latitude φ) at a certain time (month a, day b, c hour d minute) is calculated by multiplying this solar constant by the distance from the sun of the target point at the target time (r * ) and the angle that the ground at the target point makes with respect to sunlight (solar altitude α; a value different from α in equations 1 to 4 above. The same applies to δ and k, which appear below). In other words, the extraatmospheric global solar radiation Q0 can be expressed by the following equation. Q0=1367(r / r * ) 2 sin(α) ……(I)

[0119] The geocentric solar distance r / r in the above formula (I) * (The distance r between the Earth and the Sun at the target time) * The ratio of the average distance r to the solar altitude α can be calculated by the following procedure.

[0120] First, calculate the angle θ0 that has moved due to revolution from New Year's Day to the actual day using the following formula: dn is the total number of days from New Year's Day, and can be calculated simply from the date (month a, day b). θ0=2π(dn-1) / 365 ……(II)

[0121] Using the angle θ0 calculated by the above formula (II), the solar declination δ and the geocentric solar distance r / r * and the equation of time Eq can be found. δ=0.006918-0.399912cos(θ0)+0.070257sin(θ0) -0.006758cos(2θ0) +0.000907sin(2θ0) -0.002697cos(3θ0) +0.001480sin(3θ0) ……(III) r / r * =1 / √{1.000110+0.034221cos(θ0) +0.001280sin(θo) +0.000719cos(2θ0)+0.000077sin(2θ0)} ……(IV) Eq=0.000075+0.001868cos(θ0)-0.032077sin(θ0)-90.014615cos(2θ0)-0.040849sin(2θ0) ……(V)

[0122] Furthermore, the solar hour angle h can be calculated from the equation of time Eq, the longitude difference from standard time, and the meridian using the following formula: h = (standard time - 12)π / 12 + longitude difference from the standard meridian + Eq ……(VI)

[0123] From the values ​​thus obtained, the solar azimuth ψ and solar altitude α at a certain time and location can be calculated using the following formulas. α=arcsin{sin(φ)sin(δ)+cos(φ)cos(δ)cos(h)} ……(VII) ψ=arctan[cos(φ)cos(δ)sin(h) / {sin(φ)sin(α)-sin(δ)}] ……(VIII)

[0124] From the above calculations, the solar altitude α and the geocentric solar distance r / r * has been calculated, and by substituting these values ​​into the above formula (I), the extra-atmospheric global solar radiation Q0 can be obtained. This extra-atmospheric global solar radiation Q0 is the amount of solar radiation (energy received per unit area and time) received by a virtual surface outside the atmosphere corresponding to the target point at the target time.

[0125] The value obtained by adding the attenuation due to passing through the atmospheric layer to this extra-atmospheric global solar radiation Q0 is treated as the solar radiation Q. The solar radiation Q is a function of the extra-atmospheric global solar radiation Q0, and specifically, can be calculated as an estimated value using the following formula. Q=kQ0……(IX)

[0126] That is, when it is desired to estimate the amount of solar radiation at a certain point at a certain time, the extra-atmospheric global solar radiation Q0 is first calculated based on the above formulas (I) to (VIII), and then this is multiplied by a separately determined coefficient k to obtain the estimated solar radiation value Q. In this way, the amount of solar radiation can be estimated simply and accurately.

[0127] Here, k is a coefficient that represents the attenuation due to passage through the atmospheric layer, and can be calculated, for example, by dividing the actual solar radiation measured at a certain point in time by the theoretical extra-atmospheric global solar radiation Q0 at the same point in time.

[0128] In the air-conditioning systems shown in FIGS. 1 to 3, when an estimated value based on the above formula (IX) is used as the amount of solar radiation, whether to use this estimated value during operation at a certain time and what value to use as the coefficient k can be determined as needed based on weather and other conditions. For example, when the weather is clear and no shadow is cast on the pyranometer 100 by obstructions O1 and O2, the actual measured value from the pyranometer 100 can be used as the amount of solar radiation. Furthermore, when the weather is cloudy or rainy, most of the solar radiation is scattered light, and the influence of direct solar radiation is not significant. Therefore, the influence of shadowing is relatively small. Therefore, depending on the application, it may not be necessary to use an estimated value. In such cases, the amount of solar radiation may be estimated under the condition that the weather is clear and the shadowing period is the shaded period, and the estimated value may be used as the amount of solar radiation only when the influence of shadowing on the measured amount of solar radiation is significant. Of course, even when the weather is cloudy or rainy, if the influence of shadowing is a concern, the estimated value based on the above formula (IX) may also be used.

[0129] Furthermore, even when using the above formula (IX), the value of the coefficient k can be changed depending on the conditions. For example, when the value estimated by the above formula (IX) is used as the amount of solar radiation when the weather is cloudy, the coefficient k should be a different value from the coefficient k when the weather is sunny.

[0130] Furthermore, even if the weather is the same, sunny, the calculation method of the coefficient k in the above formula (IX) or the value used as the coefficient k may be changed depending on the time of day. As described above, the value of the coefficient k can be calculated by dividing the actual measurement value at a certain time by the extra-atmospheric global solar radiation Q0. For example, if a shadow falls on the pyranometer 100 during the time period from t1 to t2 immediately after time t0 on a sunny day, the actual measurement value at time t0 can be divided by the extra-atmospheric global solar radiation Q0 at the same time t0, and the solar radiation Q from time t1 to t2 can be calculated as an estimated value based on the above formula (IX). However, according to the research of the present inventors, empirically, it has been found that the effectiveness of the estimated value obtained by such an operation varies depending on the time of day. Specifically, while the solar radiation can be accurately estimated using the above-described operation during the afternoon, when an estimated value obtained using the same method is used during the morning, the deviation of the estimated value from the actual value (the solar radiation measured separately at a position without shadow) increases over time.

[0131] This is thought to be due to the fact that in the morning, the sun is low in the sky at the start of shadow time, making it susceptible to the effects of absorption and scattering by the atmosphere. Morning shadow time begins, for example, around 8:00, at which point the sun is low in the sky and sunlight travels a long distance through the atmosphere before reaching the position of pyranometer 100. For this reason, if coefficient k is calculated using the above formula (IX) based on the actual measured value of solar radiation just before the start of shadow time, the value of coefficient k will be significantly lower than 1. However, as time passes after shadow time begins, the sun's position rises higher, and the ratio of actual solar radiation to extraatmospheric global solar radiation becomes larger (approaching 1) than coefficient k calculated just before the start of shadow time. Therefore, if the coefficient k calculated immediately before the start of the shadow period continues to be used throughout the shadow period thereafter, the estimated value of solar radiation will deviate from the actual value (dashed line), as shown by the dashed-dotted line in Figure 14, for example. (Note that Figure 14 is a graph showing an example of the relationship between the measured solar radiation value by the pyranometer 100, the actual solar radiation, and the estimated solar radiation value, where the measured solar radiation value is shown by the solid line and the actual solar radiation value is shown by the dashed line. The dashed-dotted line and the double-dashed line each represent estimated values ​​calculated using a different coefficient k.) This effect is particularly significant in seasons when shadow periods are long, when the sun's altitude is low throughout the entire time and sunrise is late.

[0132] Therefore, for morning shadow time, a preset value is used as coefficient k, rather than using a coefficient k calculated based on the actual measured value immediately before the start of the shadow time. This facilitates relatively accurate estimation of solar radiation, as shown by the two-dot chain line in FIG. 14 . When using a preset coefficient k, coefficient k can be calculated as the average value of the ratio of the actual measured value of solar radiation on a clear day outside the shadow time to the extra-atmospheric global solar radiation. When calculating coefficient k in this manner, this calculation may be performed monthly, for example, so that a different coefficient k value is used each month. The specific value of coefficient k may vary depending on the time of day, season, etc., but is generally around 0.65 to 0.75.

[0133] On the other hand, for the afternoon shadow time, the solar radiation is estimated using the coefficient k calculated based on the actual measured value at the time immediately before the start of the shadow time. This allows for accurate estimation of the solar radiation in accordance with the actual solar radiation conditions in the morning and afternoon.

[0134] Furthermore, it is thought that the same effect will be greater with the afternoon shadow time if the time is later (the ratio of actual solar radiation to extraterrestrial solar radiation will become smaller (moving away from 1) than the coefficient k calculated just before the start of the shadow time), so in such cases, a pre-set value can also be used as the coefficient k.

[0135] An example of a specific procedure for estimating the amount of solar radiation using the above formula (IX) will be described with reference to the flowchart in Fig. 15. Fig. 15 illustrates a procedure for determining the weather at the beginning of a daylight period (for example, about 1.5 hours after sunrise) and determining subsequent operations related to the amount of solar radiation based on the weather determination.

[0136] The solar radiation estimator 101 first acquires the current time (step S101) and determines whether the acquired current time is a weather determination time (step S102). Here, the "weather determination time" is, for example, a time that is set in advance as "a sufficiently early time during the time period when there is sunlight, a time when the sun is high enough in the sky to allow the weather to be determined from the amount of sunlight, and a time when no shadow is cast on the position of the pyranometer 100," and specifically, for example, a time 1.5 hours after sunrise. Here, the "sufficiently early time during the time period when there is sunlight" refers to "a time between sunrise and that time when the amount of sunlight is low even if the weather is clear, and when the amount of sunlight does not have much impact on the operation of the system even if the measured value of the amount of sunlight by the pyranometer 100 deviates from the actual amount of sunlight." A time period that meets the above condition of "a sufficiently early time during the time period when there is sunshine and when the sun is high enough in the sky to be able to judge the weather from the amount of solar radiation" is, for example, about one to two hours after sunrise, and a time within this time period at which it is known that no shadow will be cast at the position of the pyranometer 100 can be set as the weather judgment time. Note that "a time at which it is known that no shadow will be cast at the position of the pyranometer 100" can be determined empirically, for example, from the actual shadow duration around the target date and time. Alternatively, it can be calculated from the movement of the sun at the target date and time and the location of objects and structures around the pyranometer 100.

[0137] If it is determined in step S102 that the current time is the weather determination time, the weather at that time is determined as the weather for that day (step S103). The weather can be determined, for example, by the amount of solar radiation measured by the pyranometer 100. If the amount of solar radiation exceeds a certain threshold (for example, 200 W / m 2 ) or more, it can be determined that it is sunny, and if it is less than the threshold, it can be determined that it is cloudy or rainy. After determining the weather, the process returns to step S101.

[0138] If it is determined in step S102 that it is not time to determine the weather, the result of the weather determination is further determined (step S104). When step S104 is executed, if step S102 for that day has already been executed and the weather is fine, the process proceeds to step S106, but if step S102 for that day has not been executed, or if step S102 has been executed but the weather is not fine, the process proceeds to step S105.

[0139] In step S105, it is determined that the actual measurement value of the pyranometer 100 is used as the amount of solar radiation, and the process returns to step S101.

[0140] In step S106, it is determined whether the current time is in a shadow time. This determination is made by referring to the shadow time data input in advance to the solar radiation amount estimation unit 101 in the form shown in Table 1 below. [Table 1]

[0141] Note that while the shadow time can be accurately calculated, for example, from the movement of the sun at a given date and time and the location of objects and structures around the pyranometer 100, the setting of the shadow time does not need to be very accurate, at least when estimating solar radiation for the purpose of operating an air conditioning system. As shown in Table 1 above, it is sufficient to roughly grasp the time periods during which the pyranometer 100 is shaded in a certain season and set the shadow time broadly to cover that entire period. In this way, there will be time periods during which the pyranometer 100 is not actually shaded but which correspond to the shadow time in the setting. During such time periods, even though the pyranometer 100 is not shaded (and therefore, even if the actual measured value is used as the amount of solar radiation, there is no influence of the shadow), an estimated value will be used as the amount of solar radiation instead of an actual measured value. However, as long as the estimated value does not deviate significantly from the actual amount of solar radiation, no major problems will occur. Conversely, if the set shadow time is shorter than the actual shadow time, there will be times when the actual measured value is used instead of the estimated value for solar radiation, even though the shadow affects the measured value (actual shadow time), which may cause problems in understanding the amount of solar radiation.

[0142] It should be noted that the above Table 1 is merely an example, and it goes without saying that the number of times shaded in a day may be more or less than this depending on the conditions around the pyranometer, etc. Also, while the above example shows a table in which shaded hours are set by month, data in which shaded hours are set by week or day, for example, may also be used.

[0143] If the current time is not in the shadow period, the process proceeds to step S105, where it is determined to use the actual measurement value of the pyranometer 100 as the amount of solar radiation, and the process returns to step S101. If the current time is in the shadow period, the process proceeds to step S107.

[0144] In step S107, it is determined whether the current time is in the morning. If it is in the morning, the process proceeds to step S108, where it is determined that a predetermined value to be used as the coefficient for the morning shadow time is to be used as the coefficient k (see formula (IX) above). Next, the solar radiation estimation unit 101 multiplies the extra-atmospheric global solar radiation Q0 at that time by the value of this coefficient k (see formulas (I) to (VIII) above for the method of calculating the extra-atmospheric global solar radiation Q0) to calculate an estimated value Q of the solar radiation (step S109). Once the calculation is complete, the process returns to step S101.

[0145] If it is determined in step S107 that the current time is not morning, the process proceeds to step S110, where a coefficient k to be used to estimate the amount of solar radiation during the subsequent shadow period is calculated. The coefficient k can be calculated by dividing the actual measured value of solar radiation immediately before the start of the shadow period by the extra-atmospheric global solar radiation Q0. Note that the "actual measured value of solar radiation" used to calculate the coefficient k here should not be a measured value at a specific point in time, but rather an average value of measured values ​​over a certain period of time (for example, the average value of solar radiation measured in the 10 minutes immediately before the shadow period). This is to reduce the influence of noise on the estimated value of solar radiation. The calculated value of coefficient k is multiplied by the extra-atmospheric global solar radiation Q0 at that time to calculate the estimated value Q of solar radiation (step S111). After calculation, the process returns to step S101.

[0146] 15, the weather is determined only once a day at a predetermined early time (1.5 hours after sunrise). The weather determination can be performed based on the measurement value of the pyranometer 100. Therefore, it is not necessary for an operator to input the weather for that day into the solar radiation estimation unit 101 or to use information such as a weather forecast obtained from an external source, and it is possible to estimate the amount of solar radiation according to the weather with a minimum number of steps using only the information obtained by the pyranometer 100.

[0147] Another example of the procedure for estimating the amount of solar radiation using the above formula (IX) will be described with reference to the flowchart in Fig. 16. Fig. 16 illustrates the procedure for determining the weather during each shadow period and determining subsequent operations related to the amount of solar radiation based on the weather determination.

[0148] In this procedure, the solar radiation estimation unit 101 acquires the current time (step S121), and then determines whether the current time corresponds to a shadow time (step S122). The shadow time can be determined by referencing the current time to the data shown in Table 1 above, but may also be determined by other methods.

[0149] If the current time is not in the shadow period, the process proceeds to step S123, where it is determined to use the actual measurement value of the pyranometer 100 as the amount of solar radiation, and the process returns to step S121. If the current time is in the shadow period, the process proceeds to step S124.

[0150] In step S124, it is determined whether the current time is AM. If it is AM, the process proceeds to step S125, where the weather is determined. The weather determination here is performed by referring to weather forecast data, for example.

[0151] If the weather is not clear, the process proceeds to step S123, where it is determined to use the actual measurement value of the pyranometer 100 as the amount of solar radiation, and the process returns to step S121. If the weather is clear, the process proceeds to step S126.

[0152] In step S126, it is determined that a predetermined value used for the morning shadow time is used as the coefficient k (see formula (IX) above). Next, the solar radiation estimation unit 101 multiplies the extra-atmospheric global solar radiation Q0 at that time by the value of this coefficient k to calculate an estimated solar radiation value Q (step S127). Once the calculation is complete, the process returns to step S121.

[0153] If it is determined in step S124 that the current time is not in the morning, the process proceeds to step S128, where the weather is determined by referring to, for example, weather forecast data or data input by the worker as the current weather.

[0154] If the weather is not clear, the process proceeds to step S123, where it is determined to use the actual measurement value of the pyranometer 100 as the amount of solar radiation, and the process returns to step S121. If the weather is clear, the process proceeds to step S129.

[0155] In step S129, a coefficient k to be used to estimate the amount of solar radiation during the subsequent shadow period is calculated. The coefficient k can be calculated by dividing the actual measured value of solar radiation immediately before the start of the shadow period by the extra-atmospheric global solar radiation Q0. The calculated value of coefficient k is multiplied by the extra-atmospheric global solar radiation Q0 at that time to calculate an estimated value Q of solar radiation (step S130). Once the calculation is complete, the process returns to step S121.

[0156] In the procedure shown in Figure 16 above, the weather is judged when calculating the amount of solar radiation during shadow time, so compared to the procedure shown in Figure 15 above, in which the weather is judged only once at the beginning of the day, this allows for operation that is more in line with actual weather and is particularly effective during seasons when the weather is changeable.

[0157] Another example of the procedure for estimating the amount of solar radiation will be described with reference to the flowchart of Fig. 17. The procedure described in Fig. 17 is generally similar to the procedure in Fig. 15, but in the procedure in Fig. 17, whether to use an estimated value or an actually measured value as the amount of solar radiation for the morning shadow time is determined depending on the weather, while for the afternoon shadow time, an estimated value is used regardless of the weather determination.

[0158] The process from obtaining the current time to determining the weather is the same as the procedure shown in Fig. 15. First, the current time is obtained (step S101), and it is determined whether the obtained current time is the time to determine the weather (step S102). If it is the time to determine the weather, the weather at that time is determined (step S103). Once the weather has been determined, the process returns to step S101.

[0159] If it is determined in step S102 that it is not the time to determine the weather, the procedure first determines whether the current time is in shadow (step S134). If the current time is not in shadow, it is determined that the actual measured value of the pyranometer 100 is used as the amount of solar radiation (step S105), and the process returns to step S101.

[0160] If the current time is in the shadow time, it is then determined whether the current time is in the morning (step S136). If the current time is in the morning, the current time is in the shadow time and is also in the morning. In this procedure, the weather is determined at this stage (step S137).

[0161] If the weather is determined to be clear in step S137, the process proceeds to step S138, where a predetermined value is used as the coefficient for the morning shadow time as the coefficient k. The solar radiation estimation unit 101 multiplies the extra-atmospheric global solar radiation Q0 at that time by the value of this coefficient k to calculate an estimated solar radiation value Q (step S139). After the calculation, the process returns to step S101.

[0162] If it is determined in step S137 that the weather is not clear, the process proceeds to step S105, where it is determined to use the actual measurement value of the pyranometer 100 as the amount of solar radiation, and the process returns to step S101.

[0163] If it is determined in step S136 that it is not morning, the current time is in the shadow and afternoon, in which case the process proceeds to step S140.

[0164] In step S140, a coefficient k to be used to estimate the amount of solar radiation during the subsequent shadow period is calculated. Here, coefficient k is calculated by dividing the actual measured value of solar radiation immediately before the start of the shadow period by the extra-atmospheric global solar radiation Q0. Here, the "actual measured value of solar radiation" used to calculate coefficient k may be the average value of measurements over a certain period of time. The calculated value of coefficient k is multiplied by the extra-atmospheric global solar radiation Q0 at that time to calculate an estimated value Q of solar radiation (step S141). Once calculation is complete, the process returns to step S101.

[0165] In this procedure, estimated values ​​are used as insolation during morning shadow time only if the weather is determined to be clear, and estimated values ​​are used as insolation during afternoon shadow time regardless of the weather determination. Because weather changes constantly, if the weather determination for that day is performed only once in the early morning, the reliability of the weather determination result decreases over time. Therefore, in the procedure shown in Figure 17, the weather determination result is referenced for morning shadow time and the operation related to insolation is changed depending on the weather, but the weather determination result is not referenced for afternoon shadow time and an estimated value is used as insolation regardless of whether the weather is clear or not. Note that even if the weather is not clear, the coefficient k can be calculated from the extraatmospheric global solar radiation Q0 and the actual measured value of insolation, and it is possible to use this to calculate an estimated value of insolation during shadow time (although the value of coefficient k used in this case will be smaller than the value of coefficient k for clear weather). When it is cloudy or rainy, the amount of solar radiation fluctuates greatly because the solar radiation is blocked by the clouds, but the influence of direct solar radiation on the amount of solar radiation is not zero. Therefore, as explained above, by calculating the coefficient k based on the actual measured value of the amount of solar radiation obtained at that time each time and then calculating an estimated value of the amount of solar radiation based on this, it is possible to estimate the amount of solar radiation with a certain degree of accuracy even when it is cloudy or rainy.

[0166] The three procedures described herein are merely examples. When implementing the solar radiation estimation method of the present invention, appropriate modifications may be made, such as changing the content and order of each step, omitting some steps, or adding other steps. For example, the above description describes a case in which solar radiation is not estimated unless the weather is determined to be clear, and the actual measured value is used as the solar radiation. However, if it is desired to minimize the influence of shadows even on cloudy or rainy days, solar radiation may be estimated during shadow periods using a coefficient k corresponding to the weather, even when the weather is cloudy or rainy. Furthermore, depending on the start and end times of each shadow period, coefficient k may be calculated from the actual measured value of solar radiation during morning shadow periods, or a predetermined coefficient k may be used during afternoon shadow periods. Furthermore, while the above description describes a procedure for estimating solar radiation in real time, it is possible to envision applications in which retrospective estimation of solar radiation is sufficient. In such cases, the order of each step may be significantly changed.

[0167] According to this method for estimating the amount of solar radiation, it is possible to obtain an amount of solar radiation close to the actual value by eliminating the influence of the shadow and using an estimated value instead of an actually measured value during the shadow time when the sun casts a shadow on the pyranometer 100. The estimated amount of solar radiation can be calculated by a simple calculation based on the amount of extra-atmospheric global solar radiation, and is very simple because no special equipment other than the pyranometer 100 is required.

[0168] To eliminate the influence of shadows on pyranometer measurements, it is of course possible to install the pyranometer in a location where shadows are not generated, but this may be difficult. For example, various devices are often installed on the rooftops of office buildings, and there is often no space available to install a pyranometer in a location where shadows are not generated. In such cases, it is theoretically possible to provide a structure such as a platform for installing the pyranometer, but this would naturally incur additional costs. Another possible solution would be to install multiple pyranometers in different locations and use the actual measurements of one pyranometer during times when a shadow is generated on another pyranometer. However, installing multiple pyranometers naturally incurs considerable costs. According to the method or system of the present embodiment, it is possible to easily estimate solar radiation and eliminate the influence of shadows on the pyranometer with a minimal configuration, while minimizing the need for additional structures, devices, or complex calculations.

[0169] By following the above procedure, the influence of shading can be minimized and an amount of solar radiation close to the correct value can be obtained by appropriately using an estimated value as the amount of solar radiation at the installation position of the pyranometer 100 during shadow hours. Next, a specific method for applying the amount of solar radiation thus obtained (actually measured value or estimated value) to air conditioning control at each position in the perimeter zone P on each floor will be described.

[0170] When using the rooftop solar radiation to generate the temperature estimation model M by the model generation unit 14 or to estimate the temperature near the floor at each location using the temperature estimation model M, it is advisable to take into account the shadow time at each location and use the rooftop solar radiation accordingly.

[0171] The effect of solar radiation on a certain location in a building varies greatly depending on whether or not there is solar radiation at that location, i.e., whether the location is in the sun or shade. The presence or absence of solar radiation also varies depending on the direction of the location, the number of floors in the building, the location on the floor, and the season and time of day. Therefore, when estimating some condition related to solar radiation at a certain location in a space to be air-conditioned, a highly accurate estimation of the effect of solar radiation can be made by taking into account the presence or absence of solar radiation at the target location in addition to the amount of solar radiation on the rooftop determined using the above method.

[0172] In the air conditioning system of this embodiment described above, if it is desired to estimate the near-floor temperature at each location at a certain time while taking into account the amount of solar radiation, for example, the shadow time at each location can be determined, and based on this, the presence or absence of solar radiation at each location at each time can be compiled as data such as that shown in Table 2 below. That is, in addition to the amount of solar radiation determined using the above method, the presence or absence of solar radiation at the target location can be determined using shadow time data such as that shown in Table 2 below, and this can be used as an explanatory variable in the temperature estimation model M. In Table 2 below, the names listed in the top row (5F 1-1-1, 36F 4-2-3, etc.) correspond to each location on each floor, and "RF" refers to the rooftop. The left column shows each time of year in 5-minute increments. At a certain location (for example, 1-1-1 on the 6th floor) at a certain time (for example, 10:00 on April 1st or 9:00 on June 12th), if there is sunlight (the location is in the sun), the value is defined as "1," and if there is no sunlight (the location is in the shade), the value is defined as "0." [Table 2]

[0173] In this way, by referring to Table 2 above, the shadow duration at a target location on a target floor in a building to be air-conditioned can be determined, and the presence or absence of solar radiation at a certain time can be input into the temperature estimation model M as a variable of "1" or "0." That is, in the temperature estimation model M, for example, if there is solar radiation at a target location at a certain time, "1" is input from the data in Table 2 above as an explanatory variable indicating the presence or absence of solar radiation. Based on this, the temperature estimation model M calculates the near-floor temperature by performing some kind of processing, such as multiplying an explanatory variable such as the amount of solar radiation on the rooftop (the amount of solar radiation actually measured or estimated as the amount of solar radiation measured by the pyranometer 100) by an appropriate value. Also, if there is no solar radiation at the target location, "0" is input as the explanatory variable indicating the presence or absence of solar radiation. Based on this, the temperature estimation model M calculates the near-floor temperature by performing some kind of processing (calculation processing different from that when the explanatory variable indicating the presence or absence of solar radiation has a value of 1) on the explanatory variable such as the amount of solar radiation on the rooftop. Note that this explanation is a hypothetical example, and the actual processing performed based on each setting variable within the temperature estimation model M generated by machine learning will vary depending on the machine learning process and the specifications of the temperature estimation model M.

[0174] Furthermore, when using data relating to the presence or absence of solar radiation such as that in Table 2 above in the temperature estimation model M, rather than inputting the presence or absence of solar radiation directly into the temperature estimation model M in the form described above (either "1" or "0"), it is also possible to input a value obtained by performing some kind of calculation processing on the rooftop solar radiation in advance depending on the presence or absence of solar radiation into the temperature estimation model M. In addition, when using the rooftop solar radiation or the presence or absence of solar radiation at each position in the temperature estimation model M, any appropriate method can be used as a specific method.

[0175] Data on the presence or absence of solar radiation at each location, such as that shown in Table 2 above, can be created, for example, based on a sky factor map for each location. A sky factor map (or sky map) is a diagram that shows the portion of the sky that an object, such as a building, occupies as seen from a certain measurement point. By creating a sky factor map, it is possible to determine the duration of shadowing for each day of the year based on the longitude and latitude of the measurement point and the position and shape of the object. Therefore, by creating a sky factor map for each target location, it is possible to create data on the duration of shadowing and the presence or absence of solar radiation, such as that shown in Table 2 above.

[0176] However, in a high-rise building, for example, it would be extremely time-consuming to create sky exposure factor maps for all locations expected to be subject to air conditioning control and to ascertain shadow time. Therefore, for example, some floors among the air-conditioned locations could be designated as representative floors, and the shadow time calculated at the location of the representative floor based on the sky exposure factor map. For the shadow time of floors other than the representative floor, the shadow time of the representative floor could be used as is, or the shadow time of the representative floor could be adjusted in some way. This is because, even if the floors or locations are different, as long as the locations are close to each other, the shadow time is generally considered to be roughly the same.

[0177] For example, if the 6th to 30th floors of a building are to be subject to air conditioning control, the 8th, 13th, 18th, 23rd, and 27th floors are defined as representative floors, and the shadow time data of the 8th floor is used for the 6th to 10th floors, the shadow time data of the 13th floor for the 11th to 15th floors, the shadow time data of the 18th floor for the 16th to 20th floors, the shadow time data of the 23rd floor for the 21st to 25th floors, and the shadow time data of the 27th floor for the 26th to 30th floors. Alternatively, for example, the 9th to 12th floors may be interpolated based on the data of the 8th and 13th floors. For example, if the start time of morning shadow time at a certain location on the 8th floor on a certain day is 9:00, and the start time of morning shadow time at a corresponding location on the 13th floor on the same day is 9:30, the start times of morning shadow time on floors 9 to 12 may be interpolated proportionally between 9:00 and 9:30, and determined to be 9:06, 9:12, 9:18, 9:24, etc., respectively.

[0178] In addition, when there are a large number of locations on one floor that are expected to be subject to air conditioning control, some locations can be designated as representative locations to calculate the shadow time, and for other locations, the shadow time at a nearby representative location can be used as is, or the shadow time can be corrected by interpolating based on the shadow time at multiple representative locations.

[0179] Furthermore, while Table 2 above schedules the presence or absence of sunlight at each location in 5-minute increments, it is also possible to calculate the sunlight exposure time for only some days and use this to determine the sunlight exposure time for other days, even when seasonal variations in sunlight exposure time occur. While it is theoretically possible to accurately determine the sunlight exposure schedule for each location based on the sky exposure factor map, performing similar calculations for all 365 days of the year would require a significant amount of calculation effort. Therefore, it is possible to calculate the sunlight exposure time for a limited number of days, such as the first day of each month, as a representative value, and then interpolate the calculated sunlight exposure time for other days to determine the sunlight exposure time. For example, the sunlight exposure time for the first day of each month can be calculated based on the sky exposure factor map, and the sunlight exposure time for April 2 to April 30 can be calculated by interpolating between the sunlight exposure time schedule for April 1 and the sunlight exposure time schedule for May 1.

[0180] Furthermore, if you try to calculate the amount of solar radiation for each day of the year, as in a general solar radiation simulation, you cannot use the data for a representative floor or representative day as explained above as the basis and apply it to other locations or days, and you have to perform all simulations individually every five minutes.As explained above, by setting the presence or absence of solar radiation as a variable separate from the amount of solar radiation and understanding this as a schedule, you can significantly reduce the amount of work required for calculations. Table 3 below shows the results of examining the estimation accuracy of near-floor temperatures for several types of temperature estimation models constructed to use parameters selected from the parameters listed above as explanatory variables. [Table 3] Model 1 is a temperature estimation model that uses the current outdoor air temperature, current solar radiation, the operating status of the perimeter air conditioner, and the current ceiling temperature (current measurements from ceiling temperature sensors installed in eight locations) as explanatory variables. Outdoor air temperature and solar radiation are major factors in the exterior load of the perimeter zone. The operating status of the perimeter air conditioner (the air conditioner that supplies intake air to the perimeter zone) naturally directly affects the air conditioning status in the perimeter zone. As mentioned above, the ceiling temperature is the direct control target when controlling the near-floor temperature in the air conditioning system of this embodiment, and is correlated with the near-floor temperature. Model 1, which uses these values ​​as explanatory variables, had an RMSE value of 0.587, which indicates the estimation accuracy of the near-floor temperature, and was able to estimate the near-floor temperature with sufficient accuracy for practical purposes. In addition to the four explanatory variables used in Model 1, Model 2 also uses past ceiling temperature values ​​(up to 30 minutes in 5-minute increments) as an explanatory variable. As mentioned above, control of the ceiling temperature affects the temperature near the floor with a time lag, so it is thought that past ceiling temperature values ​​have a particularly high correlation with the temperature near the floor. Model 2, which uses past ceiling temperature values, has an RMSE value of 0.564, which is even lower than Model 1, and is even more accurate than Model 1. Model 3 uses moving average values ​​of outdoor temperature and solar radiation (i.e., the average of the current value and multiple past values ​​close to the present) as explanatory variables instead of the current values ​​used in Model 1. Outdoor temperature and solar radiation have a significant impact on the load on the perimeter zone exterior, but there is a time lag between their changes and the resulting change in perimeter zone temperature. For this reason, past values ​​are often more strongly correlated with the perimeter zone's near-floor temperature than current values, and moving average values ​​that also reflect past values ​​are thought to have a stronger correlation than current values. Furthermore, values ​​that tend to fluctuate significantly over a short period of time, such as solar radiation, are difficult to calculate using the gradient method used to determine appropriate weighting when using machine learning with neural networks, and average values ​​are easier to learn. Model 3, which uses these explanatory variables, had an RMSE of 0.576, again demonstrating higher accuracy than Model 1. In addition to the four explanatory variables used in Model 1, Model 4 also uses the measured supply air temperatures of the perimeter and interior air conditioners as explanatory variables. While it goes without saying that the supply air temperature of the perimeter zone (the supply air temperature of the perimeter air conditioner) directly affects the temperature in the perimeter zone, the supply air temperature of the interior zone (the supply air temperature of the interior air conditioner) also has a significant impact on the temperature in the perimeter zone. For example, in the interior zone of an office building, where people are likely to be active and many devices are operating, heating may be used in the morning, shortly after people begin to be active and devices are operating, and cooling may be used in the afternoon once the indoor air has warmed. Furthermore, cooling and heating in the interior zone have significantly different effects on the air conditioning status in the perimeter zone. Therefore, in addition to the supply air temperature of the perimeter air conditioner, the supply air temperature of the interior zone air conditioner is also useful as an explanatory variable when estimating the temperature near the floor in the perimeter zone. Model 5, which used these explanatory variables, had an RMSE value of 0.573, which was also more accurate than Model 1. In addition to the four explanatory variables used in Model 1, Model 5 also uses the amount of heat flowing outdoors as an explanatory variable. The heat balance in air conditioning is related to the movement and generation of various types of heat, but among these values, the amount of heat flowing outdoors has a particularly large impact on the indoor temperature. Model 5, which uses this amount of heat flowing outdoors as an explanatory variable, had an RMSE value of 0.564, which also showed higher accuracy than Model 1. Models 6 and 7 use moving average values ​​for outdoor temperature and solar radiation, just like Model 2, and also use the operating status of the perimeter air conditioners and interior air conditioners, the actual measured intake air temperature, and the current values ​​of the outflow heat and ceiling temperature as explanatory variables. Models 6 and 7 also use the operating status of the variable air volume devices facing the perimeter zone as an explanatory variable. The operating status of the variable air volume devices facing the perimeter zone naturally has a direct impact on the temperature in the perimeter zone. Furthermore, Models 6 and 7 also use the presence or absence of solar radiation at multiple locations (Points 1 to 6) as an explanatory variable, and as a result, the RMSE was 0.548 or 0.551, demonstrating particularly high accuracy compared to the other Models 1 to 5. Furthermore, Model 6 in particular also uses past ceiling temperature values ​​as an explanatory variable, but the RMSE value is even lower than Model 7, which does not use past ceiling temperature values ​​(0.551 for Model 7, compared to 0.548 for Model 6). As mentioned above, changes in ceiling temperature affect the near-floor temperature with a time lag, so it is thought that by using past ceiling temperature values ​​as an explanatory variable in this way, it is possible to estimate the near-floor temperature with even greater accuracy.

[0181] As described above, the air conditioner comprises an air conditioner 1 that sends out supply air A1, an air supply duct 2 that guides the supply air A1 from the air conditioner 1 to the target space S, a plurality of air outlets 4 that blow out the supply air A1 into the target space S, an air volume variable device 5 that adjusts the air volume of the supply air A1 blown out from one of the air outlets 4 that supplies the air to the perimeter zone P, a perimeter fan 6 that is provided in the perimeter zone P and sends indoor air A2 upward, an air intake 16 that is provided above the perimeter fan 6 and draws in at least a portion of the return air to the air conditioner, and a ceiling temperature sensor 9 that measures the ceiling temperature in the perimeter zone P, and is configured to estimate the near-floor temperature at least at each position in different directions in the perimeter zone P based on a parameter related to the near-floor temperature in the perimeter zone P, and to adjust the blowing air volume of the air volume variable device 5 facing the perimeter zone P based on the estimated near-floor temperature, and the estimation of the near-floor temperature is configured to be performed using at least the ceiling temperature sensed by the ceiling temperature sensor 9 as a parameter. In this way, the amount of air sent from the air outlet 4 to the perimeter zone P is appropriately increased based on the near-floor temperature, and the supplied warm air reaches the perimeter fan 6 and is sent to the upper air inlet 16, thereby blocking the cold air generated at the outermost perimeter of the perimeter zone P and preventing cold drafts from entering inside. In this way, the temperature distribution near the floor of the target space S can be corrected. In this case, by using the ceiling temperature, which is the direct control target when controlling the near-floor temperature, as a parameter, the near-floor temperature can be accurately estimated.

[0182] Furthermore, the air conditioning system of this embodiment is configured to estimate the near-floor temperature using as a parameter at least the past value of the ceiling temperature detected by the ceiling temperature sensor 9. In this way, by using as a parameter the past value of the ceiling temperature, which has a large influence on the current near-floor temperature, the near-floor temperature can be estimated with even greater accuracy.

[0183] Furthermore, the air conditioning system of this embodiment is configured to estimate the near-floor temperature further using the amount of solar radiation as a parameter, and the amount of solar radiation is grasped as an estimated value calculated based on the amount of extra-atmospheric global solar radiation during the shadow time set as the time when a shadow is cast on the pyranometer 100, while at other times it is grasped as an actual measured value by the pyranometer 100. In this way, when estimating the near-floor temperature using the amount of solar radiation as a parameter, by using the estimated value instead of the actual measured value during the shadow time, it is possible to eliminate the influence of shadow and obtain an amount of solar radiation close to the actual value.

[0184] Furthermore, in the air conditioning system of this embodiment, the estimated value of the amount of solar radiation during the shadow time can be calculated by multiplying the extra-atmospheric global solar radiation during the shadow time by a preset coefficient, and in this way, the amount of solar radiation can be estimated easily and accurately.

[0185] Furthermore, in the air conditioning system of this embodiment, the estimated value of the amount of solar radiation during the shadow time can be calculated by multiplying the amount of extraterrestrial global solar radiation during the shadow time by a coefficient calculated by dividing the actual measured value of solar radiation before the shadow time by the amount of extraterrestrial global solar radiation.In this way, the amount of solar radiation can be estimated easily and accurately.

[0186] In addition, in the air conditioning system of this embodiment, the estimated value of solar radiation during the morning shadow time is calculated by multiplying the extra-atmospheric global solar radiation during the shadow time by a preset coefficient, and the estimated value of solar radiation during the afternoon shadow time is calculated by multiplying the extra-atmospheric global solar radiation during the shadow time by a coefficient calculated by dividing the actual measured value of solar radiation before the shadow time by the extra-atmospheric global solar radiation. In this way, the solar radiation can be accurately estimated in accordance with the actual solar radiation conditions in the morning and afternoon time periods.

[0187] Furthermore, in the air conditioning system of this embodiment, the amount of solar radiation can be estimated under the conditions that it is a shadow time and the weather is determined to be clear. In this way, the estimated value can be used as the amount of solar radiation only when the influence of the shadow on the amount of solar radiation is large.

[0188] Furthermore, in the air conditioning system of this embodiment, the estimation of the near-floor temperature can be performed using the presence or absence of solar radiation at the target position as an additional parameter. In this way, when estimating the near-floor temperature at the target position using the amount of solar radiation, it is possible to perform a highly accurate estimation of the influence of the amount of solar radiation.

[0189] Furthermore, in the air conditioning system of this embodiment, the presence or absence of solar radiation can be determined based on the shadow time at the target position.

[0190] Furthermore, in the air conditioning system of this embodiment, the shadow time at a target location can be determined based on the sky factor chart.

[0191] Furthermore, the air conditioning system of this embodiment is configured to estimate the near-floor temperature of the perimeter zone P using a temperature estimation model M generated by machine learning. In this way, the near-floor temperature of the perimeter zone P can be estimated with high accuracy.

[0192] Furthermore, in the air conditioning system of this embodiment, the temperature estimation model M can be configured to estimate the near-floor temperature by further using some or all of the parameters selected from the following parameters as explanatory variables. Outside temperature Operating status of air conditioner 1 ·Outflow heat amount ·Wind direction ·wind speed ·Rainfall amount Measurement value of floor temperature sensor 11 Perimeter fan 6 operating status Operating state of variable air volume device 5 facing perimeter zone P ·date ·time ·day of week Furthermore, the air conditioning system of this embodiment can be configured so that the amount of solar radiation and the presence or absence of solar radiation at the target position are input to the temperature estimation model M as mutually separate explanatory variables.

[0193] Therefore, according to the present embodiment, the air temperature near the floor of the perimeter zone can be suitably estimated.

[0194] The air conditioning system of the present invention is not limited to the above-described embodiment, and various modifications can be made without departing from the spirit of the present invention. [Explanation of symbols]

[0195] 1 Air conditioner 2 Air intake duct 4 Air outlet 5 Variable air volume device 6 Perimeter Fans 9 Ceiling temperature sensor 11 Near-floor temperature sensor 16 Intake port 100 pyranometer A1 Air supply A2 Indoor air M Temperature estimation model P Perimeter Zone S target space

Claims

1. An air conditioner that sends out supply air, an air supply duct that guides air from the air conditioner to a target space; A plurality of air outlets that blow supply air into the target space; an air volume variable device for adjusting the volume of supply air blown out from one of the air outlets that supplies air to the perimeter zone; a perimeter fan provided in the perimeter zone for blowing indoor air upward; an intake port provided above the perimeter fan for drawing in at least a portion of return air to the air conditioner; a ceiling temperature sensor for measuring a ceiling temperature in the perimeter zone, and estimating the temperature near the floor at each position in at least different directions in the perimeter zone based on a parameter related to the temperature near the floor in the perimeter zone; The airflow rate of the variable air volume device facing the perimeter zone is adjusted based on the estimated near-floor temperature, the estimation of the near-floor temperature is configured to be performed using at least the ceiling temperature sensed by the ceiling temperature sensor as a parameter, and is further configured to be performed using the amount of solar radiation as a parameter; The amount of solar radiation is grasped as an estimated value calculated based on the amount of extra-atmospheric global solar radiation during a shadow time set as a time when a shadow is cast on the pyranometer, while At other times, the system is configured to grasp the actual measured value of the pyranometer. An air conditioning system characterized by:

2. An air conditioner that sends out supply air, an air supply duct that guides air from the air conditioner to a target space; A plurality of air outlets that blow supply air into the target space; an air volume variable device for adjusting the volume of supply air blown out from one of the air outlets that supplies air to the perimeter zone; a perimeter fan provided in the perimeter zone for blowing indoor air upward; an intake port provided above the perimeter fan for drawing in at least a portion of return air to the air conditioner; a ceiling temperature sensor for measuring the ceiling temperature in the perimeter zone; The system is configured to estimate near-floor temperatures at least at positions in different directions in the perimeter zone based on parameters related to near-floor temperatures in the perimeter zone, and adjust the blow-out air volume of the variable air volume device facing the perimeter zone based on the estimated near-floor temperatures; the estimation of the near-floor temperature is configured to be performed using at least the ceiling temperature detected by the ceiling temperature sensor as a parameter, and is configured to be performed using the presence or absence of solar radiation at the target position as a further parameter; The presence or absence of solar radiation is determined based on the shadow time at the target location. An air conditioning system characterized by:

3. The estimation of the temperature near the floor is performed using at least a past value of the ceiling temperature detected by the ceiling temperature sensor as a parameter.

3. The air conditioning system according to claim 1 or 2,

4. The estimation of the near-floor temperature is further configured to be performed using an amount of solar radiation as a parameter, The amount of solar radiation is grasped as an estimated value calculated based on the amount of extra-atmospheric global solar radiation during a shadow time set as a time when a shadow is cast on the pyranometer, while At other times, the system is configured to grasp the actual measured value of the pyranometer.

3. The air conditioning system according to claim 2, wherein:

5. The estimated value of solar radiation during the shadow time is calculated by multiplying the extraterrestrial solar radiation during the shadow time by a preset coefficient.

5. The air conditioning system according to claim 1 or 4,

6. The estimated value of solar radiation during the shadow period is calculated by multiplying the coefficient obtained by dividing the actual measured value of solar radiation before the shadow period by the extraterrestrial global solar radiation during the shadow period. The air conditioning system according to any one of claims 1, 4 and 5, characterized in that

7. The estimated value of solar radiation during the morning shadow time is calculated by multiplying the extraterrestrial global solar radiation during the shadow time by a preset coefficient, The estimated value of solar radiation during the afternoon shadow period is calculated by multiplying the extraterrestrial global solar radiation during the shadow period by a coefficient calculated by dividing the actual solar radiation measured before the shadow period by the extraterrestrial global solar radiation.

5. The air conditioning system according to claim 1 or 4,

8. Estimate the amount of solar radiation under the condition that it is a shadow time and the weather is judged to be sunny. The air conditioning system according to any one of claims 1 and 4 to 7,

9. The estimation of the temperature near the floor is configured to be performed using the presence or absence of solar radiation at the target position as a further parameter. The air conditioning system according to any one of claims 1 and 3 to 8,

10. The presence or absence of solar radiation is determined based on the shadow time at the target location.

10. The air conditioning system according to claim 9,

11. The shadow duration at the target location should be determined based on the sky factor chart.

11. The air conditioning system according to claim 10,

12. The air conditioning system according to any one of claims 1 to 11, characterized in that it is configured to estimate the temperature near the floor of the perimeter zone using a temperature estimation model generated by machine learning.

13. 13. The air conditioning system according to claim 12, wherein the temperature estimation model is configured to estimate the near-floor temperature by further using some or all of the following parameters as explanatory variables: - Outside temperature -Operating state of the air conditioner ・Outflow heat amount ・Wind direction ·wind speed ・Rainfall amount - Measurement value of the floor-near temperature sensor that measures the temperature near the floor in the perimeter zone -Operating state of the perimeter fan -Operating state of the variable air volume device facing the perimeter zone ·date ·time ·day of week

14. The amount of solar radiation and the presence or absence of solar radiation at the target location are input to the temperature estimation model as separate explanatory variables.

14. The air conditioning system according to claim 12 or 13,

Images