Estimation and mitigation of indoor heat island effect

Abstract

A computer-implemented indoor urban heat index (UHI) effect prediction method includes receiving weather information of an area and predicting a predicted solar light level in the area at least based on the weather information of the area. The method further includes receiving sensed indoor air quality data and predicting predicted indoor air quality data of an indoor space in the area at least based on the sensed indoor air quality data. The method also includes predicting indoor UHI effect at the indoor space at least based on the predicted solar light level and the predicted indoor air quality data.

Description

[0001] Estimation and mitigation of indoor heat island effect

[0002] FIELD OF THE INVENTION

[0003] The present disclosure relates generally to urban heat islands, and more particularly to artificial intelligence (Al) based estimation and mitigation of urban heat island effects in indoor spaces.

[0004] BACKGROUND OF THE INVENTION

[0005] An urban heating island (UHI) is an area that is a warmer than surrounding areas. UHI effect is a phenomenon where urban areas experience a significantly elevated temperature than surrounding areas due to urbanization. Elevated temperatures impact the environment and public health in multiple ways including increased energy consumption, air pollution, impaired water quality, and the well-being of urban residents. For example, UHI effect may result in increased building cooling energy consumption that can lead to increased pollution such as increased particulate matter PM 2.5 (PM2.5). Air pollution can impact the health and wellbeing of people exposed to the air pollutants. Air pollutants, such as PM2.5, may also result in increased electricity consumption equivalent to an increase in temperature. In some cases, estimating current or future UHI effects at indoor spaces may enable improved building management that has benefits such as reduced exposure to pollutants and reduced energy consumption. In some cases, precise weather and air quality information may be needed to reliably estimate indoor UHI effects (i.e., UHI effects in indoor spaces). Thus, a solution that enables reliably estimating indoor UHI effects may be desirable.

[0006] SUMMARY OF THE INVENTION

[0007] The present disclosure relates generally to urban heat islands, and more particularly to Al based estimation and mitigation of urban heat island effects in indoor spaces. In an example embodiment, a computer-implemented indoor urban heat index (UHI) effect prediction method includes receiving weather information of an area and predicting a predicted solar light level in the area at least based on the weather information of the area. The method further includes receiving sensed indoor air quality data and predicting predicted indoor air quality data of an indoor space in the area at least based on the sensed indoor air quality data. The method also includes predicting indoor UHI effect at the indoor space at least based on the predicted solar light level and the predicted indoor air quality data.

[0008] These and other aspects, objects, features, and embodiments will be apparent from the following description and the appended claims.

[0009] BRIEF DESCRIPTION OF THE DRAWINGS

[0010] Reference will now be made to the accompanying drawings, which are not necessarily drawn to scale, and wherein:

[0011] FIG. 1 illustrates a block diagram of a system for indoor UHI effect prediction according to an example embodiment;

[0012] FIG. 2 illustrates a light level prediction Al model of the system of FIG. 1 in training mode according to an example embodiment;

[0013] FIG. 3 illustrates sensors that provide indoor air quality data for use in the system of FIG. 1 according to an example embodiment;

[0014] FIG. 4 illustrates a system for implementing indoor UHI effect predictions according to an example embodiment; and

[0015] FIG. 5 illustrates a method of indoor UHI effect prediction according to an example embodiment.

[0016] The drawings illustrate only example embodiments and are therefore not to be considered limiting in scope. The elements and features shown in the drawings are not necessarily to scale, emphasis instead being placed upon clearly illustrating the principles of the example embodiments. Additionally, certain dimensions or placements may be exaggerated to help visually convey such principles. In the drawings, the same reference numerals used in different figures may designate like or corresponding but not necessarily identical elements.

[0017] DETAILED DESCRIPTION OF THE EMBODIMENTS

[0018] In the following paragraphs, example embodiments will be described in further detail with reference to the figures. In the description, well known components, methods, and / or processing techniques are omitted or briefly described. Furthermore, reference to various feature(s) of the embodiments is not to suggest that all embodiments must include the referenced feature(s).

[0019] FIG. 1 illustrates a block diagram of a system 100 for indoor UHI effect prediction according to an example embodiment. In some example embodiments, the system 100 includes a light level prediction Al model 102, an air quality prediction Al model 104, and a UHI effect prediction Al model 106. A local server or a remote server (e.g., a cloud server) may execute the light level prediction Al model 102, air quality prediction Al model 104, and the UHI effect prediction Al model 106. The UHI effect prediction Al model 106 may predict indoor UHI effect 112 in one or more indoor spaces in an area based on a predicted solar light level 116 and predicted indoor air quality data 118 that includes indoor air pollution level. In general, the indoor UHI effect 112 may include or may refer to an amount of increase in temperature of an indoor space in an area, where the amount of increase in temperature is caused by solar light in the area and / or indoor air quality elements such as indoor air pollution level, relative humidity, etc. The UHI effect prediction Al model 106 may predict the amount of increase in temperature of the indoor space (302) caused by the predicted indoor air quality data only, i.e., without the contribution of the predicted solar light level. In addition to the indoor UHI effect 112, the UHI effect prediction Al model 106 may provide UHI effect time information 114. For example, the UHI effect time information may indicate time duration of the indoor UHI effect 112. As an illustrative example, the indoor UHI effect 112 may be 3 degrees Celsius, and the UHI effect time information 114 may be 2 hours, which indicates the duration of time that the indoor UHI effect 112 is 3 degrees Celsius. Alternatively, the indoor UHI effect 112 and / or the UHI effect time information 114 may be expressed differently without departing from the scope of this disclosure.

[0020] In some example embodiments, the UHI effect prediction Al model 106 may receive the predicted solar light level 116 from the light level prediction Al model 102. The predicted solar light level 116 may include estimated current outdoor solar light level and / or forecasted solar light level that is, for example, for a time period from current time.

[0021] To illustrate, in some example embodiments, the light level prediction Al model 102 may receive weather information 108 (i.e., outdoor weather information) and predict / estimate the predicted solar light level 116 in an area. The area may be, for example, a city, a county, a zip code area, a neighborhood, or another region (e.g., 5.5 km radius area). A local server or a remote server (e.g., a cloud server) may execute the light level prediction Al model 102 to predict / estimate the predicted solar light level in the area using the weather information 108 as input. The weather information 108 may include meteorological data and solar radiance data (e.g., Global Horizontal Irradiation (GHI)) for the area. The meteorological data may include cloud cover information, precipitation (e.g., rain, snow, etc.), temperature, humidity, wind, and / or other similar information. In general, the solar radiance data does not account for weather conditions such as cloud cover, wind, precipitation, etc.

[0022] In some example embodiments, the weather information 108 may include current weather information (i.e., for current time) and / or forecasted weather information (i.e., for a later time). For example, the forecasted weather information may be for one hour, two hours, six hours, etc. from the current time. As another example, the forecasted weather information may be a seasonal weather forecast information that is, for example, for an upcoming season.

[0023] In some example embodiments, the light level prediction Al model 102 may receive / obtain the weather information 108, such as meteorological and solar radiance data, from a weather information source such as, for example, weatherbit.io. For example, a weather application programming interface (API) may be used to receive / obtain the meteorological and solar radiance data from a weather information source. Because the solar radiance data obtained from a weather information source generally does not account for weather conditions such as, for example, cloud cover, precipitation, and / or wind, the solar radiance data may not accurately and / or precisely indicate the solar light level in the area. The solar radiance data obtained from a weather information source may also be for a larger area than the particular area of interest. In contrast, the predicted solar light level 116 that is predicted / estimated by light level prediction Al model 102 based on the weather information 108 accounts for weather conditions and can be specific to the area of interest.

[0024] To illustrate, FIG. 2 illustrates the light level prediction Al model 102 of the system 100 of FIG. 1 in training mode according to an example embodiment. In some example embodiments, the light level prediction Al model 102 may be trained using historical meteorological and solar radiance data 202 (i.e., historical weather information). For example, the light level prediction Al model 102 may receive or otherwise obtain the historical meteorological and solar radiance data 202 from the national solar radiation database (NSRDL) or from another source. The historical meteorological data may include cloud cover information, precipitation (e.g., rain, snow, etc.), temperature, humidity, wind, and / or other similar information. The historical solar radiance data may include GHI data. The historical meteorological and solar radiance data may include weather data for a period of time such as, for example, 3 years. Alternatively, the historical meteorological and solar radiance data may be for a shorter or longer time period.

[0025] Referring to FIGS. 1 and 2, in some example embodiments, ground truth light level data 204 used in the training of the light level prediction Al model 102 may be obtained from one or more sensors located in the area. For example, the sensors may be outdoor solar light sensors that are integrated in outdoor luminaires, and solar light level data from the sensors may be stored over time in a solar light level database and used as the ground truth light level data 204. Generally, the ground truth light level data 204 may be collected from sensors (e.g., outdoor sensors 322, 324 shown in FIG. 3) that are located in the area for which the predicted solar light level 116 is predicted / estimated by the light level prediction Al model 102. For example, the historical meteorological and solar radiance data 202 and the ground truth light level data 204 may be specific to the area for which the light level prediction Al model 102 is used to provide the predicted solar light level 116. Alternatively, the historical meteorological and solar radiance data 202 and the ground truth light level data 204 may be obtained for another area (e.g., outdoor solar light sensors in another area) but may be deemed adequate to train the light level prediction Al model 102 for use in the particular area of interest. The ground truth light level data 204 used in training the light level prediction Al model 102 may be for the same duration of time as the historical meteorological and solar radiance data 202.

[0026] In some example embodiments, after the light level prediction Al model 102 is trained using historical meteorological and solar radiance data 202 (i.e., historical weather information) and ground truth light level data collected from solar light sensors, the light level prediction Al model 102 may use the weather information 108, for example, from a weather information source (e.g., weatherbit.io) and predict / estimate the solar light level in the area of interest. As described above, the weather information 108 may include meteorological data and solar radiance data for the area. Because the solar light level predicted / estimated by the light level prediction Al model 102 based on the weather information 108 accounts for weather conditions indicated by the weather information 108, the solar light level predicted / estimated by the light level prediction Al model 102 may be more accurate than the solar radiance information included the weather information 108. In addition, if the ground truth solar light level data 204 used in training the light level prediction Al model 102 is from sensors in the area of use of the light level prediction Al model 102, the light level prediction Al model 102 may account for additional factors particular to the area and thus resulting in even higher accuracy compared to the solar radiance data included in the weather information 108.

[0027] In some example embodiments, the UHI effect prediction Al model 106 may receive predicted indoor air quality data 118 from the air quality prediction Al model 104. For example, the predicted indoor air quality data 118 may include predicted indoor temperature, predicted indoor relative humidity, and predicted indoor air pollution level that are predicted / estimated by the air quality prediction Al model 104. The predicted indoor air pollution level included in the predicted indoor air quality data 118 may be or may include predicted PM2.5 level and / or predicted carbon dioxide level.

[0028] In some example embodiments, the air quality prediction Al model 104 may predict / estimate the predicted indoor air quality data 118 based on the sensed indoor air quality data 110. A local server or a remote server (e.g., a cloud server) may execute the air quality prediction Al model 104 to predict / estimate the predicted indoor air quality data 118 using the sensed indoor air quality data 110 as input. The sensed indoor air quality data 110 may include sensed indoor temperature, sensed indoor humidity, sensed indoor air pressure, sensed indoor air pollution level, and / or other sensed air quality information. The sensed indoor air pollution level may include sensed PM2.5 level and / or sensed carbon dioxide level. The air quality prediction Al model 104 may receive the sensed indoor air quality data 110 from one or more indoor air quality sensors located in one or more indoor spaces that are in the area for which the solar light level is predicted / estimated using the light level prediction Al model 102 as described above. For example, the indoor spaces may be rooms in a building that is in the area.

[0029] To illustrate, FIG. 3 illustrates air quality sensors 310, 312, 314, 316 that provide the sensed indoor air quality data 110 for use in the system 100 of FIG. 1 according to an example embodiment. Referring to FIGS. 1 and 3, in some example embodiments, the sensors 310-316 may be located inside a building 318 located in an area 320. The area 320 may be a city, a county, a zip code area, a neighborhood, or another region (e.g., 5.5 km radius area). For example, the weather information 108 used as input to the light level prediction Al model 102 shown in FIG. 1 may be obtained for the area 320. That is, information such as temperature, humidity, cloud cover, precipitation, etc. included in the weather information 108 used by the light level prediction Al model 102 is for the area 320 where the building 318 is located.

[0030] In some example embodiments, the building 318 may include indoor spaces / rooms 302, 304, 306, 308. The sensor 310 may be located in the room 302 to sense air quality in the room 302. The sensor 312 may be located in the room 304 to sense air quality in the room 304. The sensor 314 may be located in the room 306 to sense air quality in the room 306. The sensor 316 may be located in the room 308 to sense air quality in the room 308. Sensor data from each sensor of the sensors 310-316 may be distinctly included in the sensed indoor air quality data 110 and provided to the air quality prediction Al model 104. For example, the sensed indoor air quality data 110 may include sets of sensor data from the sensors 310-316, where the sets of data are delineated from each other based on the respective source (i.e., the sensor 310, 312, 314, or 316) of each set of sensor data. In some cases, identification information of the sensors 310-316 may be provided to the air quality prediction Al model 104 along with the sensed indoor air quality data 110. In some cases, location information (e.g., room number) of the sensors 310-316 may be provided to the air quality prediction Al model 104 along with the sensed indoor air quality data 110.

[0031] In some example embodiments, the sensed indoor air quality data 110 may include sensor data from one of the sensors 310-316 at a time. That is, in some cases, the air quality prediction Al model 104 may be executed using sensor data from one of the sensors 310-316 at a time. For example, sensor data from the sensors 310-316 may be provided to the air quality prediction Al model 104 serially on a per sensor basis as part of the sensed indoor air quality data 110.

[0032] As described above, the air quality prediction Al model 104 may predict / estimate the predicted indoor air quality data 118 based on the sensed indoor air quality data 110. Because the sensed indoor air quality data 110 may include delineated sets of sensor data associated with respective ones of the sensor 310-316, the predicted indoor air quality data 118 may include sets of predicted air quality data associated with the respective sensors.

[0033] In some example embodiments, the predicted indoor air quality data 118 from the air quality prediction Al model 104 may be for current time or for a later / future time. To illustrate, the sensed indoor air quality data 110 may include real-time air quality data as sensed by one or more of the sensors 310-316. In such cases, the predicted indoor air quality data 118 may effectively match the real-time air quality data from one or more of the sensors 310-316. For example, an input (e.g., time information) indicating whether the air quality prediction Al model 104 is to predict / estimate the predicted indoor air quality data 118 for the current time or for a later / future time may be provided to the air quality prediction Al model 104. In general, the later / future time for the predicted indoor air quality data 118 predicted / estimated matches the later / future time for which the weather information 108 is forecasted. For example, the later / future time for both may be 1 hour, 2 hours, 3 hours, 8 hours, etc.

[0034] In some example embodiments, the air quality prediction Al model 104 may be trained using time series data from the sensors 310-316 and / or other indoor air quality sensors. For example, using indoor air quality data from the sensors 310-316 sensed at a first time, the air quality prediction Al model 104 may be trained to predict / estimate the predicted indoor air quality data 118 using, as ground truth data, indoor air quality data from the sensors 310-316 sensed at a second / later time. The air quality prediction Al model 104 may be trained to predict / estimate predicted indoor air quality data 118 for multiple later / future times.

[0035] In some example embodiments, the UHI effect prediction Al model 106 may predict / estimate the indoor UHI effect 112 in the one or more of the indoor spaces / rooms 302, 304, 306, 308 that are in the building 318 located in the area 320. In general, the UHI effect prediction Al model 106 is an Al model that is trained to predict / estimate the indoor UHI effect 112. The UHI effect prediction Al model 106 may predict / estimate the indoor UHI effect 112 based on the predicted solar light level 116 and / or the predicted indoor air quality data 118 that may include indoor air pollution level such as PM2.5 level and / or carbon dioxide level. The predicted solar light level 116 may include predicted outdoor solar light level at the current time and / or at a later time. The predicted indoor air quality data 118 may include predicted indoor air quality data at the current time and / or at a later time that is, for example, the same time as the later time with respect to the predicted solar light level 116. For example, the predicted solar light level 116 and the predicted indoor air quality data 118 may be both for the current time or both for the same later time (e.g., 1 hour, 2 hours, or another time from current time). The UHI effect prediction Al model 106 may predict / estimate the indoor UHI effect 112 based on a time series data of the predicted solar light level 116 and / or the predicted indoor air quality data 118 that correspond, for example, to the current time and multiple later / future times or strictly to multiple later / future times. As such, the indoor UHI effect 112 predicted / estimated by the UHI effect prediction Al model 106 may correspond to the current time and / or and multiple later / future times.

[0036] To illustrate, as described above, in addition to the indoor UHI effect 112, the UHI effect prediction Al model 106 may provide UHI effect time information 114. For example, the UHI effect time information 114 may indicate time duration corresponding to the indoor UHI effect 112. To illustrate, if the indoor UHI effect 112 remains generally the same value (e.g., 2 degrees Celsius) from the current time period until a first time period (e.g., 2 hours), the UHI effect time information 114 may indicate the first time period (e.g., 2 hours). If the indoor UHI effect 112 remains generally the same (e.g., 3 degrees Celsius) from the first time period until a second time period (e.g., 6 hours), the UHI effect time information 114 may indicate the second time period (e.g., 6 hours). Alternatively, the UHI effect time information 114 may indicate the current time and / or the future times corresponding to the predicted solar light level 116 and / or the predicted indoor air quality data 118 used by the UHI effect prediction Al model 106 to predict / estimate the indoor UHI effect 112.

[0037] In some example embodiments, as described above, because the sensed indoor air quality data 110 may include per-sensor delineated sets of sensor data, the predicted indoor air quality data 118 may include per-sensor delineated sets of predicted air quality data. As such, the UHI effect 112 generated based on the predicted indoor air quality data 118 may include UHI values associated with the respective ones of the sensors 310-316. As an illustrative example, the UHI effect 112 may include 1.5 degree Celsius associated with the sensor 310, 1.8 degree Celsius associated with the sensor 310, 2.0 degree Celsius associated with the sensor 312, and 2.1 degree Celsius associated with the sensor 314. Because the locations of the sensor 310-316 are known (e.g., room numbers of the rooms 302-308), the indoor UHI effect 112 effectively provides spatial UHI effect information.

[0038] In some example embodiments, the UHI effect prediction Al model 106 may predict / estimate the indoor UHI effect 112 based on the predicted solar light level 116 and excluding one or more elements of the predicted indoor air quality data 118. For example, the UHI effect prediction Al model 106 may predict / estimate the indoor UHI effect 112 based on the predicted solar light level 116 and without the PM2.5 level (e.g., with the PM2.5 level set to zero). By comparing values of the indoor UHI effect 112 predicted / estimated with and without the PM2.5 level, the amount of contribution, if any, of the PM2.5 to the overall UHI effect can be determined. For example, if the value of the indoor UHI effect 112 predicted / estimated with the PM2.5 level is the same or approximately the same (e.g., within a threshold) as the value of the indoor UHI effect 112 predicted / estimated without the PM2.5 level, a determination may be made that PM2.5 does not meaningfully contribute to the UHI effect in a particular space / room (e.g., one or more of the rooms 302-308). In some alternative embodiments, the UHI effect prediction Al model 106 may predict / estimate the indoor UHI effect 112 with and without one or more other elements (e.g., carbon dioxide and / or relative humidity) of the predicted indoor air quality data 118 to determine the contribution of the one or more other elements to the indoor UHI effect in a particular space / room.

[0039] In some example embodiments, the system 100 may provide (e.g., display and / or send / transmit) space management information to mitigate the indoor UHI effect 112 said space management information based on the predicted UHI effect 112. For example, the system 100 may include a server that provides the space management information, for example, to a property manager of the building 318. To illustrate, if the predicted indoor UHI effect 112 is high in the rooms 302-308 of the building 318 mainly due to air pollution (e.g., PM2.5), the system 100 may provide such information and recommend remote work for a certain period of time (e.g., afternoon or all day based on the UHI effect time information 114) based on the to reduce health risks and HVAC energy consumption. As another example, if the predicted indoor UHI effect 112 is high in the rooms 302-308 of the building 318 mainly due to solar energy, the system 100 may provide such information and recommend that window curtains and / or a roof cover be applied. As another example, if the indoor UHI effect 112 is high in some rooms (e.g., the room 302) but not in other rooms (e.g., rooms 304-308), the system 100 may provide such information and recommend an action to be taken with respect to the particular room (e.g., the room 302) where the predicted indoor UHI effect 112 is high. As yet another example, the space management information may include information that is used to redistribute workers among different locations (e.g., the rooms 302-304) of a building based on the respective indoor UHI effect at the different locations.

[0040] In some embodiments, the system 100 may provide the predicted indoor UHI effect 112 to a second actuation system, wherein the control output of the second actuation system depends on the indoor UHI effect (e.g., to a HVAC control system, to a blinds system, etc.)

[0041] By using the predicted solar light level 116 that accounts for weather conditions such as cloud cover, the system 100 can predict / estimate the indoor UHI effect 112 with relatively high accuracy. By using the distributed indoor sensors such as the sensors 310-316, the system 100 can provide area-specific indoor UHI effect information. By determining the indoor UHI effect 112 in a structure such as the building 318, the system 100 can provide information that can result in improved property management. For example, more precise property management, such as energy saving, mitigating health risks, etc., can be achieved.

[0042] In some example embodiments, the system 100 may include other models and / or blocks than shown in FIG. 1 without departing from the scope of this disclosure. In some alternative embodiments, the building 318 in FIG. 3 may include more or fewer rooms than shown without departing from the scope of this disclosure. In some alternative embodiments, the system 100 may be used with other structures than the building 318 without departing from the scope of this disclosure. In some alternative embodiments, the system 100 may include more or fewer outdoor sensors than shown without departing from the scope of this disclosure. In general, the indoor UHI effect 112 and the UHI effect time information 114 may be expressed in different manner than described above without departing from the scope of this disclosure. In some example embodiments, the indoor UHI effect 112 and the UHI effect time information 114 may be determined with respect to a single indoor space / room (e.g., the room 302, 304, 306, or 308). In some alternative embodiments, the historical meteorological and solar radiance data 202 and the ground truth light level data 204 used in training the light level prediction Al model 102 as shown in FIG. 2 may be for an area other than the area (e.g., the area 320) for which the light level prediction Al model 102 predicts / estimates the predicted solar light level 116 without departing from the scope of this disclosure.

[0043] FIG. 4 illustrates a system 400 for implementing indoor UHI effect predictions according to an example embodiment. Referring to FIGS. 1-4, in some example embodiments, the system 100 of FIG. 1 may include a device / server 402 (e.g., a computer) that includes, for example, a processor 408 (e.g., one or more microprocessors), one or more memory devices 410 (e.g., one or more flash memory units), communication interfaces 412 (e.g., a wired interface such as an Ethernet interface unit and / or wireless interfaces such as Wi-Fi interface units), etc. The server 402 may receive the weather information 108 and the sensed indoor air quality data 110 and execute the light level prediction Al model 102, the air quality prediction Al model 104, the UHI effect prediction Al model 106, and other software code to predict / estimate the indoor UHI effect 112 and the UHI effect time information 114 and to provide building management information based on the indoor UHI effect 112, and / or the UHI effect time information 114. For example, the light level prediction Al model 102, the air quality prediction Al model 104, the UHI effect prediction Al model 106 and other software code and data may be stored in the one or more memory devices 410 and accessed and executed or otherwise used by the processor 408 to perform operations described herein with respect to the server 402.

[0044] In some example embodiments, the server 402 may receive the weather information 108, for example, from a weather information server 404. For example, the server 402 may execute a weather API to obtain the weather information 108 from the weather information server 404. The server 402 may receive the sensed indoor air quality data 110 from distributed indoor air quality sensors 406 that may include, for example, the air quality sensors 310-316.

[0045] In some example embodiments, the system 400 may include other components than shown in FIG. 4 without departing from the scope of this disclosure. In some example embodiments, the server 402 may be a local server that is located, for example, in the building 318. Alternatively, the server 402 may be a remote server such as a cloud server without departing from the scope of this disclosure. In general, the system 400 may be used to execute operations described herein with respect to FIGS. 1-5. Alternatively, a different system may be used to execute the operations without departing from the scope of this disclosure.

[0046] FIG. 5 illustrates a method 500 of indoor UHI effect prediction according to an example embodiment. Referring to FIGS. 1-5, in some example embodiments, at step 502, the method 500 includes receiving weather information of an area. For example, the server 402 may receive the weather information 108 of the area 320 or another area. At step 504, the method 500 may include predicting / estimating predicted solar light level at least based on the weather information. For example, the server 402 may execute the light level prediction Al model 102 to predict / estimate the predicted solar light level 116 in the area 320 at least based on the weather information 108 of the area 320.

[0047] In some example embodiments, at step 506, the method 500 includes receiving sensed indoor air quality data. For example, the server 402 may receive the sensed indoor air quality data 110 sensed by, for example, the air quality sensor 310. The sensed indoor air quality data 110 may be real-time air quality data from, for example, the air quality sensor 310. Alternatively, the sensed indoor air quality data 110 may be air quality data sensed, for example, by the air quality sensor 310 but not necessarily provided in real time. As described above, the sensed indoor air quality data 110 may include sensed indoor temperature, sensed indoor humidity, sensed indoor air pressure, sensed indoor air pollution level, and / or other air quality information sensed or otherwise measured, for example, by the air quality sensor 310. The sensed indoor air pollution level may include a PM2.5 level and / or a carbon dioxide level.

[0048] In some example embodiments, at step 508, the method 500 includes predicting predicted indoor air quality data that includes predicted indoor air pollution level, predicted indoor temperature, and / or predicted indoor relative humidity. For example, the server 402 may execute the air quality prediction Al model 104 to predict / estimate the predicted indoor air quality data 118 of an indoor space, such as the room 302, at least based on the sensed indoor air quality data 110. As described above, the predicted indoor air quality data 118 includes predicted indoor air pollution level, predicted indoor temperature, and / or predicted indoor relative humidity, and the the predicted indoor air pollution level may include a PM2.5 level and / or carbon dioxide level. In some example embodiments, at step 510, the method 500 includes predicting indoor UHI effect based on the predicted solar light level and the predicted indoor air pollution level. For example, the server 402 may execute the UHI effect prediction Al model 106 to predict / estimate the indoor UHI effect 112 at the indoor space (e.g., the room 302) at least based on the predicted solar light level 116 and the predicted indoor air pollution level included in the predicted indoor air quality data 118.

[0049] In some example embodiments, at step 512, the method 500 may include predicting a second indoor UHI effect (i.e., a second value of the indoor UHI effect 112) at the indoor space (e.g., the room 302) at least based on the predicted solar light level 116 and without the predicted indoor air pollution level. At step 514, the method 500 may include comparing the indoor UHI effect 112 from step 510 (e.g., the value of the indoor UHI effect 112 predicted / estimated at step 510) to the second indoor UHI effect (e.g., the value of the indoor UHI effect 112 predicted / estimated at step 512) to determine a contribution of the predicted indoor air pollution level to the indoor UHI effect 112 at the indoor space (e.g., the room 302). As described above, the predicted indoor air pollution level may be a PM2.5 level.

[0050] In some example embodiments, at step 516, the method 500 includes providing building management information, for example, to a building management entity (e.g., a building manager). For example, the building management information may include indoor UHI effect 112 and / or one or more recommendations and / or instructions to mitigate the indoor UHI effect 112. The server 402 or another component of the system 100 may provide the building management information.

[0051] In some example embodiments, the method 500 includes indicating a time duration of the indoor UHI effect 112. For example, the server 402 may execute the UHI effect prediction Al model 106 and / or another executable software code to provide the UHI effect time information 114 that indicates a time period associated with a value of the indoor UHI effect 112. To illustrate, the UHI effect time information 114 that indicates the time duration of a value of the indoor UHI effect 112.

[0052] In some example embodiments, the method 500 includes predicting / estimating the indoor UHI effect 112 with respect to the same space / room (e.g., the room 302) as in step 510 but for a different time. To illustrate, the method 500 may include predicting / estimating a second indoor UHI effect (i.e., a second value of the indoor UHI effect 112) at the indoor space (e.g., the room 302) in the area 320 at least based on a second predicted solar light level (e.g., a second value of the predicted solar light level 116) predicted / estimated using the light level prediction Al model 102 and a second predicted indoor air pollution level in the indoor space (e.g., the room 302) predicted / estimated using the air quality prediction Al model 104. The predicted solar light level (e.g., a value of the predicted solar light level 116 predicted / estimated at step 504) and the second predicted solar light level (i.e., the second value of the predicted solar light level 116) correspond to different times of a day.

[0053] In some example embodiments, the method 500 includes predicting / estimating the indoor UHI effect 112 with respect to a different space / room from the space / room for which the indoor UHI effect 112 is predicted / estimated in step 510. To illustrate, the method 500 may include predicting / estimating a second indoor UHI effect (i.e., a second value of the indoor UHI effect 112) at a second indoor space (e.g., the room 306) in the area 320 at least based on the predicted solar light level (i.e., a value of the predicted solar light level 116 predicted / estimated at step 504) and a second predicted indoor air pollution level (i.e., a second value of the predicted solar light level 116) in the second indoor space (e.g., the room 306) predicted / estimated using the air quality prediction Al model 104. The indoor space (e.g., the room 302) and the second indoor space (e.g., the room 304) may be areas in a building structure (e.g., the building 318). In some example embodiments, the indoor UHI effect 112 may include values with respect to multiple spaces / rooms at the same time or sequentially.

[0054] In general, the steps of the method 500 may be executed by the server 402 of FIG. 4 and / or one or more other devices such as other computer(s), etc. In some alternative embodiments, the method 500 may include more or fewer steps than shown in and / or described with respect to FIG. 5 without departing from the scope of this disclosure. For example, some of the steps of the method 500 may be omitted without departing from the scope of this disclosure. In some alternative embodiments, some of the steps of the method 500 may be performed in a different order than shown without departing from the scope of this disclosure.

[0055] Although particular embodiments have been described herein in detail, the descriptions are by way of example. The features of the example embodiments described herein are representative and, in alternative embodiments, certain features, elements, and / or steps may be added or omitted. Additionally, modifications to aspects of the example embodiments described herein may be made by those skilled in the art without departing from the scope of the following claims, the scope of which are to be accorded the broadest interpretation so as to encompass modifications and equivalent structures.

Claims

CLAIMS1. A computer-implemented indoor urban heat index (UHI) effect prediction method (500), comprising: receiving (502) weather information (108) of an area (320); predicting (504) a predicted solar light level (116) in the area at least based on the weather information of the area; receiving (506) sensed indoor air quality data (110); predicting (508) predicted indoor air quality data (118) of an indoor space(302) in the area at least based on the sensed indoor air quality data (110); and predicting (510) indoor UHI effect (112) at the indoor space at least based on the predicted solar light level (116) and the predicted indoor air quality data (118), wherein the indoor UHI effect (112) includes an amount of increase in temperature of the indoor space (302) caused by the predicted solar light level and the predicted indoor air quality data providing building management information based on the indoor UHI effect (H2).

2. The method of Claim 1, wherein the predicted indoor air quality data (118) includes a predicted indoor air pollution level that includes a level of an air pollutant.

3. The method of Claim 2, further comprising: predicting (512) a second indoor UHI effect at the indoor space (302) based on the predicted solar light level (116) and without the predicted indoor air pollution level; and comparing (514) the indoor UHI effect (112) to the second indoor UHI effect to determine a contribution of the predicted indoor air pollution level to the indoor UHI effect (112) at the indoor space (302).

4. The method of Claim 1, further comprising providing (516) space management information to mitigate the indoor UHI effect (112) at the indoor space (302).

5. The method of Claim 1, further comprising indicating a time duration (114) of the indoor UHI effect (112).

6. The method of Claim 1, wherein the weather information (110) includes forecasted weather information.

7. The method of Claim 6, wherein the predicted indoor air quality data (118) is predicted for same time as the forecasted weather information.

8. The method of Claim 1, wherein the predicted indoor air quality data (118) includes a predicted relative humidity level.

9. The method of Claim 1, wherein predicting the predicted solar light level (116) in the area (320) is performed by executing a first artificial intelligence (Al) model (102) that is trained using ground truth data (204) from one or more solar light sensors (322, 324) in the area (320), wherein predicting the indoor air quality data (118) in the indoor space is performed by executing a second Al model (104), and wherein predicting indoor UHI effect (112) is performed by executing a third Al model (106).

10. The method of Claim 9, further comprising predicting second indoor UHI effect at the indoor space (302) in the area (320) at least based on a second predicted solar light level predicted by the first Al model (106) and a second indoor air quality data in the indoor space (302) predicted by the second Al model (104), wherein the predicted solar light level and the second predicted solar light level correspond to different times of a day.

11. The method of Claim 1, further comprising predicting a second indoor UHI effect at a second indoor space (306) in the area (320) at least based on the predicted solar light level (116) and a second indoor air quality data in the second indoor space (306), wherein the indoor space (302) and the second indoor space (306) are areas in a building structure (318).

12. The method of Claim 1, wherein the weather information (108) includes outdoor solar radiance level, outdoor temperature, outdoor wind speed, and outdoor humidity.

13. The method of Claim 1, wherein the sensed indoor air quality data (110) is obtained from an indoor air quality sensor (310) in the indoor space (302).

14. A device (402) for predicting indoor urban heat index (UHI) effect (112), the device comprising: a processor (408); a memory device (410); and a communication interface (412), wherein the processor is configured to execute a software code stored in the memory device to: receive, via the communication interface, weather information (108) of an area (320); predict a predicted solar light level (116) in the area at least based on the weather information of the area; receive, via the communication interface, sensed indoor air quality data (110); predict predicted indoor air quality data (118) of an indoor space (302) in the area at least based on the sensed indoor air quality data (110); predict indoor UHI effect (112) at the indoor space at least based on the predicted solar light level (116) and the predicted indoor air quality data (118); and provide space management information to mitigate the indoor UHI effect(112) at the indoor space (302).

Images