WO2012073161A1

WO2012073161A1 - Control of lighting systems

Info

Publication number: WO2012073161A1
Application number: PCT/IB2011/055273
Authority: WO
Inventors: Dave Alberto Tavares Cavalcanti
Original assignee: Koninklijke Philips Electronics NV
Current assignee: Koninklijke Philips NV
Priority date: 2010-11-29
Filing date: 2011-11-24
Publication date: 2012-06-07
Anticipated expiration: 2013-05-29

Abstract

A lighting system comprises controllable window light barriers for blocking exterior light in the event of glare. The system is controlled by identifying manual adjustments made to the window light barriers, and recording external light sensor measurements at the times of the adjustments as well as recording the times. A glare event area is determined from these manual adjustments, which comprises a range of external light levels and a range of time of day values at which manual adjustments had been made to reduce glare. During a subsequent period of time during which automated window light barrier control is implemented, an automated control strategy based on the determined glare event area, and control of the window light barriers is performed using the strategy.

Description

Control of lighting systems

FIELD OF THE INVENTION

This invention relates to the control of lighting systems, in particular to make optimum use of daylight in the lighting of a building and at the same time to avoid glare for occupants of the building.

BACKGROUND OF THE INVENTION

Lighting contributes to approximately 28 % of the overall energy consumption in buildings (in the US), therefore daylight control systems including artificial lights and window shades or blinds (or even curtains) and which are generically called "window light barriers" in the description below) can provide considerable energy savings.

In addition to energy savings, user comfort is a key aspect in the design of daylight control systems. The main challenge is to achieve a good balance between maximizing energy savings without annoying the user, and possibly improving user comfort.

One of the key open problems is to design control strategies that optimize the amount of daylight used while avoiding glare problems to the occupants. Glare detection and avoidance is a challenging problem due to the wide variety of possible environment conditions (e.g. window size, location, room orientation, whether, etc.) and the subjective perception of glare, which may change from one occupant to another.

In order to strike a good balance between energy savings and comfort, daylight controls systems have to take into account the needs and behaviours of the occupants.

Occupants play a key role in the overall energy performance of buildings, mainly related to the way they interact with control systems, including lights and windows controls. These issues are discussed in the paper by Reinhart, C. (2004). Lightswitch-2002: A model for manual and automated control of electric lights and blinds, Elsevier Solar Energy, 15-28.

Some control systems allow users to input their preferences, such as the desirable amount of lighting, and automatic control algorithms work towards meeting the desired user inputs, while trying to avoid glare during certain periods of the day. An example is in US-7111952. However, existing systems usually require pre-calibration and continuous user interaction to react properly to glare conditions. Moreover, existing systems do not take full advantage of daylight when applying glare control, and some systems are designed for the worst-case conditions because they cannot adapt to user needs.

One of the main problems with integrated daylight control systems is to decide when to close/open window blinds to control glare while still taking maximum advantage of daylight to dim artificial light while providing the required illumination. Existing approaches to this problem apply either open loop control of blinds based on time, location and sun position, or closed loop control based on internal and/or external light sensor measurements and set points (thresholds) that will determine when to react to potential glare.

The integrated daylight control system described in US-7111952 tries to take into account users' preferences to adjust artificial lights and windows to maximize energy savings and avoid glare. The users can provide input before the system starts operation in the form of set points for internal light sensors that will control the lights and windows in reaction to daylight entering the room. Furthermore, the system applies a correction factor to the sensor readings during pre-determined time zones (a few hours after sunrise and before sunset) to account for potential glare and react accordingly. The system requires calibration of the internal light sensor to determine the amount of external daylight entering the window and precise information on sun position in order to define the time periods when glare is expected. Also, the system is based on the assumption that glare occurs only during predetermined time zones. However, in reality, the occurrence of glare may change depending on the building/window orientation, season, time of the day, etc. Therefore, the system parameters would have to be customized for every location and even within a single building, different rooms could require different configurations for glare zones, sensor calibration and thresholds.

The daylight control strategy of US-7111952 works as follows. The system will always try to open the window shades whenever the internal light sensor (or sensors) indicates there is enough light , i.e., the light level is within a "dead band" (between min and max set points); or when there is not enough light (the light level is below the min set point of the "dead band"). On the other hand, the system will try to close the window shades when the internal light level is above the maximum set point of the dead band.

Also, the default min and max set points that define the dead band are defined a priori during a calibration phase. Glare control is applied to the above strategy by adjusting the sensor thresholds (min and max set points) during the times of the day that falls within glare control zones. For any time outside the two glare control zones, there is no glare control, i.e., there is no correction factor to account for higher chance of glare. If there is too much external light outside the glare control zone, the user would have to manually override the system in order to close the blinds.

After the manual override is complete, since the default set points have not been changed, the system will again try to open the blinds to take advantage of daylight and dim the artificial lights. In order to keep the blinds closed, the user would have to issue another manual override. The system can change the default set points after several manual overrides. This is another problem which may distract and annoy users. Another problem happens once the max set point is changed to account for the manual override and keep the blinds closed. Since the sensor is located inside the room the system cannot detect when the external environment changes and re-open the window when there is no more glare. In other words, if the set points are changed from the pre-configured values to keep the blinds closed, the system will require another user intervention in order to open the blinds, and energy savings opportunities are missed in this case if the user does not react to open the blinds when there is no more glare.

Previous studies have shown that many occupants do not actively control their blinds, or only react to some extreme conditions, which means that in many cases blinds could still be closed even when there is no glare.

In summary, although the system described in US-7111952 is one of the most advanced daylight control systems presented, it has several limitations:

It is not a scalable solution as it cannot easily adapt to different locations and scenarios without requiring customization and calibration.

It can only react to user commands, and if the reaction is slow, it may require multiple sequential overrides, which may annoy the users. On the other hand, if the reaction is made faster, the system will reduce the amount of energy saving due to daylight.

The system does not respond well after a manual override to close the blinds, as it cannot take advantage of daylight without user intervention.

Another possible approach to daylight control is the application of an appropriate glare model to quantify discomfort as a result of glare. However, since the sensation of discomfort due to glare is subjective, the precise detection of glare would require complex models to account for the user perception.

Fig. 1 shows an approximation of the daylight light level (e.g. measured outside a window) as a function of the time during a typical glare situation, where the daylight light level increases as the direct sun light intensity increases and reduces after a certain time as the sun position changes in the sky.

The light level Soi at time of day T_Di is a glare event 10, and this would cause the user to close the blinds.

This approximation does not consider fast changes in the daylight level due to weather events, such as cloud movement, but it provides the overall trend of the light level during a glare event. Although Fig. 1 gives only an ideal approximation, real measurements obtained using an external daylight sensor are shown in Fig. 2, which confirm the glare event pattern. In Fig. 2, plot 20 is the external sensor signal, and plot 22 shows a low pass filtered version.

The fundamental automatic glare detection and avoidance problem is to determine the exact time to close the blinds for any given day (TDi). Most of the existing daylight control systems, such as the one described in US-7111952, use a threshold (SDi) for the light level, measured by a single sensor (internal or external) or a combination of sensors, to detect potential glare and close the blinds. Such a set point may be adjusted based on geographical, weather and sun position information to account for glare during predetermined time zones during the day. However, one of the problems with such solutions is the fact that they require pre-configuration of the system parameters and calibration for every new deployment. Furthermore, even for a single user, the reaction to control blinds due to glare may change from one day to another, as the sun light incidence pattern changes along the year. Hence, the glare event, such as the one depicted in Fig. 1, may change in time, and one day the occupant may want to close the blinds earlier (at light level and time SDI, TDI), while in another day, he may want to close the blinds later (at light level and time SD2, TD2).

Another problem with the existing systems is the fact that they do not take into account the individual needs of occupants on a case-by-case basis, and it is well known that the perception of glare is subjective. In fact, apart from some theoretical works, existing daylight control solutions do not make use of any real-time user information, except for reaction to user's override commands. Therefore, they cannot optimize their control strategies to meet different user needs.

There is therefore a need to achieve increased energy savings and better user comfort with a more advanced daylight control system that can adapt to the individual needs and environment conditions and does not require site-specific calibration. SUMMARY OF THE INVENTION

According to the invention, there is provided a method of controlling a lighting system which comprises controllable window light barriers for blocking exterior light in the event of glare, wherein the method comprises (i) during a calibration period of time, identifying manual adjustments made to the window light barriers, and recording external light sensor measurements at the times of the adjustments as well as recording the times, (ii) determining a glare event area, which comprises a range of external light levels and a range of time of day values at which manual adjustments had been made to reduce glare, and (iii) during a subsequent period of time during which automated window light barrier control is implemented, deriving an automated control strategy based on the determined glare event area, and performing control of the window light barriers using the strategy.

This invention overcomes the above limitations and provides a much simpler and cost effective way to control daylight while avoiding glare without pre-configured parameters, site-by-site calibration and complex glare detection models. The calibration period of time is automatic, and the user will not be aware that calibration is taking place. For the user, the blinds are simply being controlled manually, but they are in use, so that the calibration is at the same time as fully functional manual operation. The solution can adapt to fit all scenarios and users by taking advantage of real-time and past measurements of both external conditions and user interaction with the control system.

The proposed invention can detect glare events for individual users and react accordingly at different times of the day, and can dynamically adapt throughout the year, without requiring sun position. The system can also dynamically adapt to any location or building orientation without customization or calibration.

The system can incorporate data collection and storage features, that enables the use of past state information to anticipate user needs and adapt in real time as the user actions and preferences change. Instead of predicting glare with a single threshold or a dead band with minimal and maximum thresholds, the invention uses a two-dimensional glare event area, which includes a range of light levels and time intervals during which a glare event is expected to occur based on past measurements for every individual control system.

Preferably, the glare event area is processed to increase the range of the time of day values before deriving the automated control strategy. This extends the time window so that the automated glare control can commence slightly before the glare event is anticipated. The range of time of day values can for example be doubled before deriving the automated control strategy.

The control of the window light barriers can comprise performing a gradual closure during an initial period, maintaining closure during a middle time period, and performing a gradual opening during an end period. This provides an improved user experience.

The calibration period of time lasts at least until manual adjustments have been made giving a non-zero range of times of day and external light levels. The identified manual adjustments made to the window light barriers comprise adjustments to close the barriers. However, commands to open the barriers can also be taken into account - although these are less likely to correspond in time exactly to the end of the glare event, since a user is likely to delay opening the blinds, since there is no discomfort (such as glare) to prompt the user to do so.

The method can further comprise controlling internal lights of the lighting system in dependence on one or more of internal light sensor measurements, the time of day, and the setting of the window light barriers.

The control method can be implemented as an algorithm, so that the invention can be implemented by a computer program.

The invention also provides a lighting system comprising controllable window light barriers for blocking exterior light in the event of glare, an external light sensor, a clock, and a controller, wherein the controller is adapted to (i) during a calibration period of time, identify manual adjustments made to the window light barriers, and record external light sensor measurements at the times of the adjustments as well as recording the times, (ii) determine a glare event area, which comprises a range of external light levels and a range of time of day values at which manual adjustments had been made to reduce glare, and (iii) during a subsequent period of time during which automated window light barrier control is implemented, derive an automated control strategy based on the determined glare event area, and performing control of the window light barriers using the strategy.

BRIEF DESCRIPTION OF THE DRAWINGS

Examples of the invention will now be described with reference to the accompanying drawings, in which:

Fig. 1 shows an approximate representation of external light level as a function of the time during a glare event; Fig. 2 is an example of a potential glare event as measured using an external light sensor;

Fig. 3 shows an example of the range of timings of manual commands to close blinds;

Fig. 4 shows a glare event area derived from the timings of Fig. 3 and used for automated control in accordance with the invention;

Fig. 5 shows a system of the invention;

Fig. 6 shows how blinds are controlled with ramped start and end control; Fig. 7 shows how user manual controls are used to trigger an update the models of glare event areas;

Fig. 8 shows how the models of glare event areas are updated; and Fig. 9 shows how the automated control functions.

DETAILED DESCRIPTION OF THE EMBODIMENTS

The invention provides a method of controlling a lighting system in which a set of light levels and time of day values are determined at which manual adjustments had been made to reduce glare, and this data is used to derive an automated control strategy for controlling window light barriers.

The invention is based on the concept of a glare event "area" (Fig. 4). By "area" is meant a region bordered by upper and lower threshold functions for two different parameters (so that if the two parameters are plotted against each other, an area on the graph is defined). These threshold functions are simply boundary values in the example below, but they may be more complex boundaries - for example a linear combination function of the two parameters. This glare event area is an expanded approximation based on a real glare detection area (Fig. 3) as defined by the actions of the user to control blinds and avoid glare. In other words, the glare event area is a prediction of the region (in time domain and light levels) during which a glare event is likely to occur for a given user.

A preferred implementation of the invention also provides a mechanism to detect and update multiple glare event areas for each user in real-time. Since the parameters that define a glare event area are computed based on real measurements for every user, it can accurately match the real conditions experienced by each user with respect to glare control.

A preferred implementation of the invention also provides a control strategy applied for each glare event area, which can detect both the start and end of a glare event based on the current system state information and act accordingly to enhance user satisfaction and maximize energy savings. Moreover, the system can take into account any user command during operation to update system configuration parameters, such as the limits of the glare detection areas.

Fig. 3 shows the area of occurrence of a glare detection event for a user. The glare event 10 explained with reference to Fig. 1 will always take place within a certain time of day - corresponding to the rising sun or the setting sun, depending on the direction faced by the window carrying the blinds. Thus, there is a time region Tmi_n to T_max within which all such glare events take place. Similarly, the light level is always within boundaries Smi_n and Smax. These two sets of boundaries define a glare detection area 30, which can be determined based on an analysis of the timing and light levels for blind control measures implemented manually by a user when there is no automatic system.

In one implementation of the invention, this measured glare detection area 30 is converted into a larger glare event area 40 as shown in Fig. 4, which is then used for automated control. This has the same maximum and minimum light levels, but an increased time window.

In the example of Fig. 4, the time window is twice as wide.

Fig. 5 shows a system of the invention;

As shown in Fig. 5, the system includes at least one internal light sensor 50, and one external light sensor 52, motorized blinds 54 over the window 55 and controllable internal light sources 56 (e.g. dimmable). A clock 57 records the time of day. The system also has occupancy sensors 58. All devices within a certain area/zone are under the control of single master controller 60, named here as zone controller (ZC). All devices within the zone can communicate, preferably via a wireless medium, with the ZC and/or other devices within the same zone. The ZC is connected through a network interface (wireless or wired) to a database 62, which can store all the data collected during the system operation.

The light sensors 50,52 periodically transmit their measurements according to a constant sampling rate to the ZC, which is responsible for processing the data and executing an integrated control algorithm for controlling lights and blinds, including the glare control strategy of the invention. The integrated control of the lights and blinds may use both internal and external sensor measurements. The glare control strategy uses primarily measurements from the external sensor, which should be placed on the window facing 55 outside. The internal light sensor is used for internal lighting control. This internal lighting control can be entirely conventional and is not the subject of this invention, although of course the internal lighting requirements will depend on the blind settings. In order to avoid unnecessary reaction to fast fluctuations in the external light level reported by the sensor, for instance, due to cloud movements, the ZC applies a low pass filter to the measurements.

The light estimate can comprise:

This combines a sensor reading with the previous filtered sensor reading, to provide a low pass averaging function, where β is an adjustable configuration parameter which adjusts the weighting of the combination.

The glare event area (GEA) 40 illustrated in Fig. 4 can be calculated based on measurements of user interaction with the blinds (manual commands) and light sensor measurements taken over a period of time after the system is installed, but before the automatic control strategy is applied. The challenge is to use values that best satisfy the user's expectations and maximize energy savings.

Multiple glare event areas can be identified by monitoring the user commands and the specific time of the day each command is issued. Typically, users only react to glare during certain time intervals, thus only a few glare events would occur during the day due to the sun movement. The specific time for each glare event may change due to several factors, such as location, window orientation, time of the day and season. Therefore commands separated in time for more than a certain threshold, say AG (e.g. 6 hours, early morning and afternoon), should be taken into account for the definition of distinct glare areas. Thus, there may be two different GEAs, for example for a window exposed to the rising and setting sun.

In one embodiment, after identifying distinct GEAs and determining the corresponding light thresholds (Smin, Tmin) and (Smax, Tmax) for each GEA, the system defines a time window A=Tmax-Tmin and central time value T_TRG= Tmin + Δ/2.

As shown in Fig. 4, a 2Δ time duration is used for the GEA. The idea is that whenever the current time is within this 2Δ window, an automatic control algorithm is executed in order to avoid glare.

This enlarged time window enables the blinds to be controlled gradually at the beginning and end of the time window, as shown in Fig. 6, which shows the blind position versus time. At the beginning 60 of the blind control period, the blinds are closed gradually, and they then remain closed for a period. At the end 62 of the blind control period, the blinds are opened gradually.

The GEA generally determines the start time for the closing of the blinds. As discussed further below, the re-opening of the blinds is based on analysis of the external light level and the trend (increasing or decreasing) of that light level. However, time duration can also be used to control the timing when the blinds are re-opened.

Multiple glare event areas for a given user can be automatically detected and updated with the system in normal operation.

Fig. 7 shows how the data is collected during the manual period of operation.

After the system is installed, without any calibration or site-specific customization, it starts monitoring all user commands, and whenever the user closes the blinds, the command and corresponding time instant is recorded.

Step 70 comprises monitoring user commands to close the blinds. The number N GEA of glare event areas detected is monitored. The first glare event area (until

N_GEA=0 as detected in step 72) is used to initialize a new GEA in step 74. If there are already glare event areas, a search is made for the closest in time (step 76). If a match is found, the data for the GEA is updated in step 78. Thus, the system checks whether any GEAs have already been detected. If no GEA has been detected, a new GEA will be initialized. Otherwise, the system searches the database for a match (based on the time of the day) of the current command with existing GEAs. If a match is found, the system updates the information about the GEA. If no match is found, e.g. if the command time is considerably different from any GEA data, a new GEA is initialized when the update algorithm is executed.

The process of Fig. 7 is carried out both during the initial calibration stage, but also during the automated control, so that any manual operations by the user (which can indicate that the automated control is not perfectly matching the user's desires) are also captured so that the GEA data is updated.

For a user issuing the first manual command to close the blinds after the system has been installed, since no GEA is defined, a new GEA will be initialized, and the system will first record the Smin and Tmin values for the GEA. No control reaction is taken at this stage, since the system is still learning the right configuration to use for this user. Suppose in a second day the user issues again another command to close the blinds at a time instant close to the same instant in the previous day. In this case, an existing GEA will be detected that matches the command, and since this is the only second command, the system will update the GEA information by filling up the Smax and Tmax variables, as well as the Δ and T_TRG values, which will receive values different from zero.

This process is repeated and the GEA information will be updated daily whenever the user reacts. The updating process enables a more accurate representation of the GEA as well as changes in the GEA with the time.

This process 78 of updating the GEA information is shown in Fig. 8. The process operates when manual commands ("use blinds cmd") have been received. This can be during the calibration period (which the user is not aware of) or during the subsequent automated period.

As shown, the system distinguishes between open and close commands.

When a close command is detected (step 80), if there is already a GEA defined, as determined in step 82, then the GEA parameters are updated in step 84, by redefining the maximum or minimum values of time and light sensor reading, if any of these have been exceeded (i.e. above the max level or below the min level) by the new data. The T_TRG and Δ values are redefined in step 86. If there is no GEA defined, it is determined if there is any light level recorded in step 87. If so, this means there has been one data recording so far only - since as soon as there are two data recordings at the same general GEA time, then a GEA is defined. If there is one data recording so far, step 84 is used to complete the maximum and minimum values based on the two recordings that now have been made. If there is no recording yet (so that this is the first data input), then the recorded light level and time is used for both the maximum and minimum value definitions in step 88.

Thus, the first reading defines the maximum and minimum values, and the second reading gives a proper set of different maximum and minimum values (assuming they are different) so that a GEA is defined.

When an open command is detected (step 90), it is determined if S>Smax

(step 91).

If the user opened the blinds when S>Smax, it indicates there is no glare and any previous GEA that was recorded may not be valid, since the user is satisfied with the blinds opened when S>Smax.

Therefore, in this case step 88 is repeated which resets the GEA. In this way it will take another close command to start recording another Smin. This should not happen if the assumption that the user closed the blinds due to glare generally when S>Smin is correct. There may be a specific event which means that the user wants to open the blind. If the user opens the blind when S<Smin (determined in step 92), no change to the GEA model is made, since this is the expected behaviour, e.g., user opens blind when he goes home at night or early in the morning.

For the case when S>Smin and S< Smax, this might be at the end of the GEA.

In the example shown, the rate with which the blind is opened can be increased in step 93 so that the automatic performance for the GEA is updated. Alternatively, the GEA could be reset in step 88 to reflect that the GEA has not accurately represented the user desired.

In a simpler implementation, there can be a single test of S>Smin?, and if yes, the GEA is reset (step 88), otherwise no action is taken. This test S>Smin indicates that the lower light level stored for the GEA is too low, since the user is happy to open the blinds even when the light level is higher.

Thus, manual user commands to open the blinds can be used to adapt the automatic control strategy, either by resetting the GEA information (as shown in Fig. 8) or adapting the setting to change the way the blinds open - for example by changing the rate at which the blinds open (which is the example shown in Fig. 8), by changing the timing at which they are opened, or by changing the GEA threshold values.

After at least one GEA is defined (e.g. T_TRG≠0 and Δ≠0), an automatic control algorithm described in Fig. 9 can start execution for every update from the external light sensor. Thus may be every few minutes, or every 10 seconds for example. For every new sensor data received (as determined in step 100), the light sensor reading is low pass filtered (step 102) before the system checks whether the current time (T) falls within the 2Δ range for a GEA, in step 104.

If that is the case, the system will check whether the current light level estimate Si (after the low pass filter is applied) is below or above the minimum level threshold that defines the GEA (Smin), in step 106. If the light level is still below Smin, no action will be taken and the system will wait for the next event.

If, on the other hand, Si>Smin and the trend of the latest measurements indicates the light level is increasing (e.g. the value AS = Si - Si_i could be used to identify the trend as shown in step 107), the system will start closing the blinds by D degrees (to allow gradual closing of the blinds) after setting a glare flag to 1. The glare flag setting is shown in steps 108 and 110, and the blind closing is step 112. If the glare flag is already 1 , the blind closing continues if the light level is below the maximum level (i.e. still within the GEA). If the light level is greater than the maximum level of the SEA, then the blinds are closed completely in step 1 14.

Thus, if this is the first time a GEA is entered, the glare flag is set to 1. Such flag is useful to differentiate whether the system is at the glare detection area (i.e. beginning of the glare event) or at the end of the glare event, when the blinds should start opening. However, if no increasing trend in the light level is determined, the system will not react and will go back to the waiting state. Whenever the system is in the GEA (i.e. glare flag=l) and the light level crosses the upper bound of the GEA (Smax), the system will close the blinds if they are not already closed.

A value D can be defined as the minimum number of degrees per step to close the blinds within the 2Δ duration of the glare detection window. This is for a set of blinds with bars rotated into their light blocking position. Assuming blinds commands can occur at the sensor data reporting rate (sampling rate), the number of steps required to fully close the blinds would be 2A/sampling_rate, and assuming R is the remaining degrees to cover in order to close the blinds after entering the GEA for the first time in a given day, D could be defined as D=R*sampling_rate/2A.

When the system receives sensor data and it has already passed the glare detection area, i.e., T > T_TRG+Δ as detected in step 120, it will check whether the glare flag is still set, in step 122.

If the light level is below the lower GEA threshold Smin as detected in step 124, the blinds are opened and the glare flag set to zero - steps 126, 128. This indicates the end of a GEA.

If the glare flag is 1 and the light level is above the lower threshold Smin, the system will start a reaction to open the blinds whenever the light level decreases below the upper limit of the GEA (Smax) as detected in step 130 and shows a decreasing trend (e.g. ASO in step 132).

No action is taken as long as the light level is still above the level Smax, so that the closure period for the blinds is extended. Even when S< Smax the system will start opening the blinds by only K degrees at a time in step 134 and this can be configured to ensure a slow reaction that reduces the chances of causing glare to the user.

Thus, the glare flag is maintained after the Tmax time value, if the light is still above Smax or if it is between Smin and Smax but is increasing. Thus, the blinds only open after Tmax and only once the light level is between below Smin or is in the Smin-Smax range but is decreasing.

An integrated daylight and artificial light control strategy can be applied to dim the lights as much as possible while opening the blinds to get daylight in, for any time before a GEA, i.e., T< T_TRG-Δ and after the GEA has ended T> T_TRG+Δ. During a GEA, the lights would react accordingly to compensate the blind control adjustments to avoid glare.

Existing daylight and glare control solutions cannot easily adapt to different locations/scenarios without requiring site-specific customization, calibration, and/or complex glare detection models. Although some solutions can react to user commands, such reaction may be too slow to meet the user needs or not optimized for energy savings. Also, existing solutions do not respond well after a glare event is detected and cannot take advantage of daylight without user intervention. This invention provides simple and cost effective way to control daylight while avoiding glare without pre-configured parameters, site-specific calibration and complex glare detection models.

The invention can be used in an integrated daylight control system that employs internal and external light sensors to control both internal lights and blinds. By applying low pass filters to the sensor readings, unnecessary reaction to fast fluctuations in measurements is avoided. User interactions with the system (e.g. commands) along with corresponding system state (e.g. sensor measurements) and time information are used to control the system. The collected data can determine potential glare detection areas throughout the day (one or more different areas are possible).

As outlined above, a glare detection area is defined as a two dimensional region in a light level vs. time plot represented by (Smin, Smax) and (Tmin, Tmax), where Smin and Smax are the minimum and maximum external light threshold for which the user has closed the blinds within a certain glare event area, and Tmin and Tmax are the time instant for the earliest and latest user command issued to close the blinds within the same glare event area.

These values are then used to control the automated control. In the example above, an expanded glare event area is defined during which an automatic control strategy will be applied to provide efficient glare control predicting and matching the actual user's needs. However, the glare event area does not need to be expanded, for example if enough user data is obtained.

A gradual glare avoidance control strategy is preferably applied when the system state is within the glare event area, to limit the user interaction. The system preferably can detect both the beginning and end of the glare event, therefore achieving maximum energy savings and anticipating user needs. The collected data is used to adapt the potential glare control zones on the time and/or light level domains as the daylight pattern and/or the user needs change with time.

The invention is applicable to daylight control systems and energy management systems for different types of buildings. The invention is applicable to integrated control in office spaces including private offices and open plan offices where users have individual control capabilities. The invention is applicable to daylight and artificial light integration systems. The invention is also applicable to lighting controls products for buildings and homes.

Other variations to the disclosed embodiments can be understood and effected by those skilled in the art in practicing the claimed invention, from a study of the drawings, the disclosure, and the appended claims. In the claims, the word "comprising" does not exclude other elements or steps, and the indefinite article "a" or "an" does not exclude a plurality. A single processor or other unit may fulfil the functions of several items recited in the claims. The mere fact that certain measures are recited in mutually different dependent claims does not indicate that a combination of these measured cannot be used to advantage. A computer program may be stored/distributed on a suitable medium, such as an optical storage medium or a solid-state medium supplied together with or as part of other hardware, but may also be distributed in other forms, such as via the Internet or other wired or wireless telecommunication systems. Any reference signs in the claims should not be construed as limiting the scope.

Claims

CLAIMS:

1. A method of controlling a lighting system which comprises controllable window light barriers for blocking exterior light in the event of glare, wherein the method comprises:

during a calibration period of time, identifying manual adjustments (80, 90) made to the window light barriers, and recording external light sensor measurements at the times of the adjustments as well as recording the times,

determining a glare event area (30), which comprises a range of external light levels (Smin-Smax) and a range of time of day values (Tmin-Tmax) at which manual adjustments had been made to reduce glare, and

during a subsequent period of time during which automated window light barrier control is implemented, deriving an automated control strategy based on the determined glare event area, and performing control of the window light barriers using the strategy.

2. A method as claimed in claim 1, further comprising:

processing the glare event area (30) to increase the range of the time of day values (Tmin-Tmax) before deriving the automated control strategy.

3. A method as claimed in claim 2, wherein the range of time of day values (Tmin-Tmax) is doubled before deriving the automated control strategy.

4. A method as claimed in claim 1, wherein the control of the window light barriers comprises:

performing a gradual closure during an initial period (60),

maintaining closure during a middle time period, and

performing a gradual opening during an end period (62).

5. A method as claimed in claim 1, wherein the calibration period of time lasts at least until manual adjustments have been made giving a non-zero range of times of day and external light levels.

6. A method as claimed in claim 1, further comprising:

identifying manual adjustments made to the window light barriers during the automated window light barrier control.

7. A method as claimed in claim 1, wherein the identified manual adjustments made to the window light barriers comprise adjustments (80) to close the barriers.

8. A method as claimed in claim 7, wherein the identified manual adjustments made to the window light barriers also comprise adjustments (90) to open the barriers, wherein the automated control strategy is updated if the manual adjustments to open the barriers conflict with the automated control strategy.

9. A method as claimed in claim 1, further comprising:

controlling internal lights of the lighting system in dependence on one or more of internal light sensor measurements, the time of day, and the setting of the window light barriers.

10. A computer program comprising code means adapted to perform the steps of claim 1 when said program is run on a computer.

11. A computer readable medium comprising a computer program as claimed in claim 10.

12. A lighting system comprising:

controllable window light barriers (54) for blocking exterior light in the event of glare,

an external light sensor (52),

a clock (57); and

a controller (60, ZC), wherein the controller is adapted to perform the method of any of claims 1 to

8.

13. A lighting system as claimed in claim 12, further comprising:

- an internal light sensor (50), and

- internal lights (56),

wherein the controller (602, ZC) is further adapted to control the internal lights (56) as well as the window light barrier (54).

14. A lighting system as claimed in claim 12, wherein the window light barrier (54) comprises blinds.