Patent Application
20120174909
The present disclosure relates generally to controlling heliostats using cameras.
A concentrating solar power system generally includes a number of heliostats configured to reflect light into a central receiver. The resulting heat can then be converted into power. Use of heliostats as a source of solar energy often requires receiver temperatures of nearly 1000° C., which in turn requires sunlight to be reflected from the heliostats into the receiver at high concentrations.
In one aspect, a heliostat control system includes a heliostat having a reflective surface and at least one reflective element, the reflective surface having a different radius of curvature than the at least one reflective element, a receiver configured to receive sunlight reflected from the reflective surface, and a camera configured to receive sunlight reflected from the at least one reflective element and to generate an image including pixels having a brightness dependent on an orientation of the reflective surface.
Implementations can include one or more of the following. The at least one reflective element may have a greater radius of curvature than the reflective surface. The at least one reflective element may be circular in shape. The at least one reflective element may be convex.
The at least one reflective element may be on the reflective surface. The at least one reflective element may be located proximate to a center of the reflective surface. There may be a plurality of reflective elements. At least some of the reflective elements may be located proximate to corners of the reflective surface. The at least one reflective element may include a first portion and a second portion, the first portion having a lower reflectivity than the second portion. The first portion of the at least one reflective element may be closer to a center of the reflective surface than the second portion. The at least one reflective element may be generally tubular in shape. The at least one reflective element may be bowed outward from the reflective surface. There may be a plurality of heliostats, and the receiver may be configured to receive sunlight reflected from a reflective surface of each of the plurality of heliostats, and the image may include pixels having a brightness depending on an orientation of each of the reflective surfaces.
A controller may be configured to receive the image from the camera and calculate an error in the orientation based upon the image. The controller may be configured to associate pixels of the image with the at least one reflective element. There may be plurality of reflective elements, and the controller may be configured to calculate the error by determining whether a brightness of pixels associated with one reflective element differs substantially from a brightness of pixels associated with a different reflective element. The controller may be configured to send a signal to change the orientation of the mirror based upon the determined error. There may be a plurality of reflective elements, and the controller may be configured to repeat the steps of receiving an image, calculating an error, and sending a signal until a brightness of pixels associated with one reflective element is substantially equivalent to a brightness of pixels associated with each of the other reflective elements. The controller may be configured to associate a portion of the image with the heliostat.
A plurality of cameras may be positioned on different sides of the receiver, and a controller may be configured to receive images from the cameras. The controller may be configured to generate an error signal from a comparison of intensity values from the images from the cameras. The controller may be configured to cause the heliostat to point toward a location off-center of the receiver. The controller may be configured to subtract an intensity value from a first image from a first camera of the plurality of cameras from an intensity value from a second image from a second camera of the plurality of cameras to generate an error signal, and the controller may be configured to control orientation of the heliostat so that the error signal reaches a non-zero target value.
In another aspect, a method of heliostat control includes receiving sunlight in a receiver, the sunlight received in the receiver reflected from a reflective surface of a heliostat, receiving sunlight in a camera, the sunlight received in the camera reflected from a at least one reflective element of a heliostat, the at least one reflective element having a different radius of curvature than the reflective surface, generating an image from the sunlight reflected into the camera, and determining an error in an orientation of the reflective surface based upon the image.
Implementations can include one or more of the following. The sunlight may be reflected from a plurality of reflective surfaces, and each reflective surface may have a corresponding heliostat, and determining may include determining an error in an orientation of each of the reflective surfaces. Pixels of the image may be associated with the at least one reflective element. There may be plurality of reflective elements, and calculating the error may include determining whether a brightness of the pixels associated with one reflective element differs substantially from a brightness of pixels associated with a different reflective element. A signal may be sent to change the orientation of the reflective surface based upon the determined error. There may be a plurality of reflective elements, and the steps of generating an image, determining an error, and sending a signal may be repeated until a brightness of pixels associated with one reflective element is substantially equivalent to a brightness of pixels associated with each of the other reflective elements. A portion of the image may be associated with the heliostat. The camera may be cooled with a cooling system. Determining an error may include comparing images generated from a plurality of cameras. Determining an error may include comparing the image with an expected image. The heliostat can be controlled to point toward a off-center location of the receiver. Controlling the heliostat can include subtracting an intensity value from a first image from a first camera of a plurality of cameras from an intensity value from a second image from a second camera of the plurality of cameras to generate an error signal, and adjusting orientation of the heliostat so that the error signal reaches a non-zero target value.
In another aspect, a method of heliostat control may include receiving sunlight in a receiver, the sunlight received in the receiver reflected from a reflective surface on a heliostat, receiving sunlight in a camera, the sunlight received in the camera reflected from at least one reflective element of a heliostat, oscillating the reflective element at a frequency, generating first image from the sunlight reflected into the camera, and assigning a portion of the image to the heliostat by detecting the frequency of oscillation in the first image.
Implementations can include one or more of the following. A second image may be generated from the camera, the assigned portion may be located in the second image, and an error in an orientation of the reflective surface may be determined based upon the assigned portion. A signal may be sent to change the orientation of the reflective surface. Sunlight may be received in the receiver from reflective surfaces of a plurality of heliostats. The sunlight may be received in the camera reflected from reflective elements of a plurality of heliostats. Each of the plurality of heliostats may be oscillated at different frequencies, and different portions of the image may be assigning to each of the plurality of heliostats by identifying the different frequencies of oscillation in the first image. Certain implementations may have one or more of the following advantages. Using cameras to detect the positioning of a heliostat can provide more precise positioning, be less expensive, require fewer components, and require less frequent maintenance and calibration than other tracking mechanisms, such as placing a sensor on each heliostat. Detecting the positioning of the heliostats allows the heliostats to be adjusted to provided higher concentration of the sunlight in the receiver. Higher concentration of sunlight in the receiver provides a higher quality or temperature of heat for the production of solar power.
By including reflective elements having a different radius of curvature than the heliostat mirror, a more accurate determination of an error in the orientation of the heliostat can be made. More accurately determining an orientation error allows the heliostat orientation to be more precisely adjusted. Adjusting the heliostat orientation precisely allows the receiver to have a higher concentration of sunlight for a greater fraction of time.
Rotating a feature on a heliostat at a known frequency provides an accurate determination of which portions of an image of the heliostat field correspond to a particular heliostat. Accurate assignment of portions of the heliostat field image allows for the adjustment of the proper heliostat. Adjusting the proper heliostat ensure that the receiver will have a higher concentration of sunlight for a greater fraction of the time.
The details of one or more embodiments are set forth in the accompanying drawings and the description below. Other features, aspects, and advantages will become apparent from the description, the drawings, and the claims.
Like reference numbers and designations in the various drawings indicate like elements.
When heliostats are used as a source of heat (which can in turn be used as a source of power), the concentration of sunlight reflected into the heat-collecting receiver can be lower than the theoretical ideal due to misalignment of the heliostat mirrors caused by errors such as difficulties detecting the orientation of the mirrors or relative position of the receiver, deformation of the mirror, or movement of the heliostat or receiver by natural causes. Using a camera to generate an image including pixels having a brightness dependent on the orientation of the heliostat mirrors allows sunlight to be more accurately reflected into the receiver. Moreover, having reflective elements on the mirror of the heliostat that have a different radius of curvature from the heliostat allows for more precise identification of errors in the heliostat orientation.
Referring to
An actuation system 132 is configured to move the heliostat mirror 160. The actuation system 132 can include multiple motors, such as a motor 120 to move the heliostat 100 in the azimuth direction using a motor shaft 130, and a motor 125 to adjust the altitude, i.e., angle of elevation, of the heliostat using a motor shaft 135. The actuation system 132 further includes control circuitry, e.g., a programmed microprocessor 172 and a transceiver 190 to receive commands directing the movement of the mirror 160. Wires 195 can electrically connect the transceiver 190 with the motors 120, 125, and the microprocessor 172 can convert the commands received by transceiver 190 into voltage signals on the wires 195 to control the motors 120, 125 and thus to control the orientation of the mirror. In other implementations, the actuation system can include hydraulic, pneumatic, cable and pulley, ballasted, or ball and socket mechanisms to move the heliostat mirror in the azimuth direction and/or to adjust the altitude.
The heliostat control system 500 further includes a tower system 200. The tower system includes a receiver 230 to receive sunlight and a camera 250, which can optionally include a filtering element 255 to reduce the intensity of the sunlight, and an optical element 259 (see
The heliostat control system 500 further includes a programmed microprocessor or computer 290 to receive image data from the cameras 250, to compute the movement of any heliostat mirrors 160 necessary to keep the heliostat oriented to reflect light to the receiver 230, and to send commands to the transceivers 190 of the heliostats. The computer 290 can be part of the tower system 200, as shown in
In operation, sunlight rays 320, 310 from the sun 300 can strike the reflective surface 165 of the heliostat mirrors 160. The reflective surface 165 can then reflect rays 321, 311 towards the receiver 230. The reflected rays 321, 311, in addition to rays reflected from other heliostats in the field, can heat the receiver 230 to temperatures of between 900° C. and 1200° C., such as between 950° C. and 1150° C. The heat can be used to drive various heat engines to produce power. For example, the heat can be used to warm cold air, which can then be expanded through a turbine engine which turns a generator shaft, which creates power. The more concentrated the sunlight is in the receiver 230, the higher the temperature of the receiver 230, and the more efficient the power generation of the system 500 can be.
In order to maximize the concentration of rays on the receiver 230, the normal vector of the reflective surface 165 must bisect the angle between the rays 310, 320 from the sun and the rays reflected towards the center of the receiver 230. Thus, as the sun 300 moves across the sky, the orientation of the reflective surface 165 of the mirrors 160 must be adjusted to ensure that the reflected rays are hitting the receiver 230 without causing too much spillage, i.e., causing too many rays to be reflected outside of the receiver 230.
The camera 255 mounted on the receiver 230 can be used to determine whether a particular mirror 160 is oriented to reflect substantially the maximum amount of light into the receiver, i.e., to orient the reflective surface 165 such that its normal vector substantially bisects the angle between the sun and the receiver. When rays from the heliostats 100 are reflected into the receiver 230, and correspondingly to the camera 255, the camera 255 observes and produces an image. The image produced by the camera 250 can include different portions, e.g., groups of pixels, having a brightness dependent on the orientation of the different heliostats. As a result, as discussed below, the image can be used to determine an error in the orientation of the mirrors.
Initially, the computer 290 will establish communication with the heliostats 100 in the field. In brief, the computer 290 includes a transceiver and maintains a communication network, e.g., a wired or wireless LAN. When the heliostat is installed and connected to the communication network, it can be assigned a network address, e.g., an Internet Protocol (IP) address, in a conventional manner. Thereafter the computer 290 can have a database including a unique identifier for each heliostat in the field and information regarding how to communicate, e.g., the network address, with each heliostat.
Since the position of the heliostat, e.g., the position in the image produced by the camera, is not necessarily known by the computer 290, a calibration step can be performed prior to determining the actual error in orientation of a heliostat. During calibration, the assignment of a particular heliostat to a set of pixels in the camera's imaging array, can be determined. For example, during the calibration step, the camera 250 can observe and produce an image of the heliostat field. Portions of the image, or groups of pixels, can be associated with a particular heliostat.
The calibration step can be automated. In some implementations, the computer 290 can send a signal to the microprocessor 172 of a newly installed heliostat, e.g., a heliostat which has established communication on the network but has not been calibrated yet, to cause the heliostat to oscillate at a predefined frequency, i.e., a frequency stored in the computer 290. The computer 290 can then find, e.g., using conventional video processing techniques, that frequency of blinking in the image in order to identify the portion of the pixel array of the image that corresponds to the heliostat. The portion of the image corresponding to the heliostat is then stored by the computer 290. In another implementation, a spinning or rotating feature can be placed on or proximate to the heliostat. The RPM of the spin can match the flicker frequency of the image such that the computer 290 can identify the heliostat in the pixel array of the image. The rotation or spinning can be performed, for example, with a motor or with wind. If rotated or spun by the wind, an optical encoding method can be used to identify the rate of spin, and the computer 290 can then match the frequency to a particular heliostat. Proceeding through each of the heliostats sequentially, each heliostat in an entire field of heliostats may be identified. Such a mechanism is important for unique identification of each heliostat. Alternatively, the computer 290 can maintain a database that associates different frequencies different heliostats, cause the multiple heliostats to oscillate simultaneously at the different frequencies, and associate different portions of the image with the different heliostats essentially simultaneously by identifying the different frequencies of oscillation.
In another implementation, two or more spinning or rotating features can be placed on or proximate to the heliostat. Including two or more spinning or rotating features on a particular heliostat helps ensure that the heliostat will have a corresponding oscillation frequency that is unique from other heliostats in the field. For example, if there are five unique frequencies of oscillation, and there are two spinning or rotating mechanisms on each heliostat, then there can be ten unique combinations of identification signals. Further, if the position of the spinning or rotating mechanism can be resolved in the image, then there are five to the power of two, or twenty-five, unique ordered combinations of unique signals. In other implementations, a rotating or spinning feature can be oscillated at increasing or decreasing frequencies or can be oscillated for a unique duration period to generate a unique signal from each heliostat. Using features that can produce unique signal can be useful for initial calibration of the heliostats. For example, each heliostat can be identified sequentially, as in the implementation described above. Alternatively, because each heliostat has a unique oscillation frequency, multiple heliostats can be identified simultaneously from a single set of images. Further, having unique identification signals for each heliostat can be useful for identification of heliostats during use, e.g., to identify a particular heliostat that may need to be repaired or adjusted. In general, if a heliostat that is flickering at a specific frequency that no other heliostat flickers at, then the heliostat has a unique flicker frequency and can be identified with video processing.
In some implementations, the physical position and approximate size of the heliostat relative to the receiver is stored by the computer 290, e.g., manually entered by an operator into the computer 290 and/or into the microprocessor 172 and transmitted to the computer 290. Assuming that the orientation of the camera 250 is stored by the computer 290, the computer 290 can be configured to calculate the portion of the image corresponding to the heliostat from this positional information.
In some implementations, calibration step can be performed manually. For example, the image from the camera 250 can be displayed to a user, and the user can use an input device to indicate the portions of the image associated with the heliostat, e.g., by clicking with a mouse on the location in the image.
In any event, once the calibration is performed, the database in the computer 290 can associate each of the heliostats in the field with the portions of the image corresponding to each of those heliostats.
Because there are generally multiple heliostats in a field, it may be advantageous to be able to detect when a heliostat is physically moved, or when a particular heliostat is added or removed from the field. Thus, in some implementations, the computer periodically compares new images with the original calibration images to look for changes. New bright spots that were not present in a previous image may indicate new heliostats, or that a heliostat has been moved. The system can automatically detect the positions of all of the heliostats in the field and recalibrate as necessary.
After the computer 290 has assigned a portion of an image to each heliostat in a field, the system 500 can determine an error in an orientation of the mirrors (i.e. to determine whether the mirror is oriented to reflect a maximum amount of light into the receiver 230) and subsequently to change the orientation of the mirrors 160 such that they reflect substantially the maximum amount of light into the receiver 230. In one implementation, the computer 290 can try to maximize the sunlight seen by the camera 250 from each heliostat. If a portion of the image assigned to a particular heliostat does not include a bright spot, or includes a spot that is not as bright as expected, the computer 290 can determine that the mirror 160 for that heliostat is not oriented accurately. For example, in
The computer 290 can send a command to motors on a particular heliostat, such as heliostat 100b in
For example, the motor 120 might first be commanded by the computer 290 to move a mirror 160 of a heliostat in a particular direction along the azimuth. If the brightness for the portion of the image assigned to that particular heliostat increases, then the computer 290 can command the motor 120 to continue to move the mirror 160 in that direction. In contrast, if the brightness decreases, then the computer 290 can command the motor 120 to move the mirror 160 in the opposite direction. These adjustment steps can then be repeated for elevating the mirror with motor 125.
The system 500 can also use additional factors to optimize the movement of the heliostats. For example, the computer 290 can take into account celestial data and data of previous days' heliostat orientation paths. Further, the computer 290 can observe the image of the sun and use its distinct features, such as solar flares, to determine which direction to move.
As the receiver 220 is a hot, harsh environment, the camera 250 may need to be protected from the heat. In some implementations, a cooling system can be used to protect the camera. For example, a jacket of coolant surrounding the camera can be connected to an external liquid coolant circulation system. Filters in front of the camera can allow only light at particular wavelengths to enter the camera while keeping the majority of the thermal energy away from the camera.
In some implementations, shown in
In operation, sunlight 351 reflected from heliostats 100 (not shown in
In other implementations, shown in
In operation, each camera 250 can produce a separate image of the same portion of the heliostat field. As discussed above, the computer 290 can determine whether a particular heliostat is misaligned by analyzing the image data for that particular heliostat. In this implementation, however, because the cameras 250 are not centered in the receiver 230, the amount of sunlight seen by the camera from each heliostat 100 or reflective element (as discussed with respect to
Additionally, when multiple images are produced from different cameras, the resulting images can be compared to one another. If the brightness is different from one camera to another in the portion of the image that corresponds to a particular heliostat or reflective elements of a heliostat, then the computer 290 can determine that the mirror of that particular heliostat is misaligned. For example, as shown in
The computer 290 can then command the motors of a misaligned heliostat to adjust the mirror 160. As in the implementation described above, the motors 120, 125 can first be commanded to move a particular heliostat in a particular direction. In this implementation, the mirror 160 can be moved by the motors until the mirrors see approximately equal brightnesses from the heliostat. For example, in
When cameras 250 are placed outside of the receiver 230, as shown in
In some implementations, examples of which are shown in
In one implementation, shown in
Each reflective element 402 can have a masked region 404 having a lower reflectivity than the unmasked portion. The masked region 404 can be oriented towards a center of the mirror 160, and can cover a portion of the reflective element, for example ⅛ to ⅓, such as ¼, of the surface area of the reflective element. For example, on the hemispherical reflective element, the masked region can be a wedge-shape. In an alternate implementation (not shown), the masked portions 404 can be oriented away from a center of the mirror 160 and can cover a larger portion of the reflective element, for example ⅔ to ⅞, such as ¾, of the surface of the reflective element. In this implementation, on the hemispherical reflective element, the unmasked region can be a wedge-shape.
In operation, using system 500 or 700, the computer 290 can associate the pixel or groups of pixels in the camera's imaging array with each reflective element of a particular heliostat. Because sunlight is reflected from the mirror 160 and reflective elements 402 differently, the reflective elements will show up as spots of higher or lower intensity (compared to the mirror 160) within or adjacent to the portion of the image that corresponds to the mirror 160 of the heliostat. The computer 290 can thus pick out the portions of the image with differing intensities relative to the rest of the mirror and associate those portions with the reflective elements.
The reflective elements 402 can then be used to determine an error in orientation of the mirror 160. Referring to
The reflective elements 402 can be configured such that each reflective element 402 reflects substantially an equal amount of sunlight when the mirror 160 is oriented correctly, i.e., with a normal bisecting the angle between the sun and the receiver (step 512). For example, in the configuration of
Furthermore, if there were x reflective elements equally spaced around the mirror 160 with (x−1)/x of the element masked off and oriented away from a center of the mirror, the computer 290 could determine that the mirror 160 was oriented correctly if all of the reflective elements generated light at nearly equal intensity. In such a setup, if the pixels associated with each reflective element show similar brightness (i.e. if all of them reflect light), the computer 290 can determine that the heliostat mirror 160 is oriented correctly (step 508).
On the other hand, if the brightness of a portion of the image assigned to one reflective element differs substantially from a portion of the image assigned to another reflective element, then the computer 290 can determine that there is an error in the orientation of the mirror 160 (step 508). For example, as shown in
On the other hand, for the situation where a camera, such as cameras 250a or 250b, is observing the heliostat field from outside of the receiver, it is possible that the reflective elements 402 will bounce more light than their adjacent ones on the same heliostat when the heliostat is properly oriented.
To address this situation, multiple cameras, such as cameras 250a and 250b, can be used to observe the heliostat field from outside of the receiver. In this case, the heliostat can be considered correctly oriented if the different cameras receive similar intensities from different groups of reflective elements. More particularly, each camera can be associated with a group of reflective elements that is on the same side of the mirror as the camera is on the receiver. If the first group of reflective elements associated with a first camera and the second group of reflective elements associated with a second camera have similar intensities, then the heliostat can be considered correctly oriented.
For example, referring to
Moreover, if there is an error in orientation, the computer 290 can use the reflective elements to determine a direction in which to reorient the mirror 160. In the implementation in which there are x reflective elements equally spaced around the mirror 160 with a portion of the element oriented towards the center of the mirror being masked off (as shown in
On the other hand, if there were x reflective elements equally spaced around the mirror 160 with a portion of the element oriented away from a center of the mirror masked off, then a reflective element having a different (higher) intensity than the rest would need to be rotated about the same axis perpendicular to the axis bisecting the reflective element, but would need to be rotated away from the receiver. Conversely, if the reflective element 402 shows a lower intensity than the others, then the mirror needs to be rotated about the axis 436 to move the reflective element 420 toward the receiver.
More generally, the computer 290 can use the relative intensities of the reflective elements to determine a direction in which to rotate the mirror to reduce angular difference between the normal vector of the reflective surface and the vector bisecting the vectors from the reflective surface to the sun and the center of the receiver 230. Alternative implementations of the reflective elements are possible, provided that a radius of curvature of the reflective elements is different from the radius of curvature of the mirror. For example, in another implementation, shown in
The sunlight reflected from the reflective elements can be treated as binary information. For example, if sun is reflected from a reflective element, it can be considered “on” or one, while if no sun is reflected from a reflective element, it can be considered “off” or zero. Alternatively, the sunlight reflected from the reflective elements can be treated as a quantified relative value.
In some implementations, multiple cameras can be placed in the receiver. In some implementations, multiple view ports can be placed in the receiver, each view port connected to a different camera. In some implementations, multiple cameras can be mounted at a distance from the receiver, and optical paths can be used that project the image to the cameras.
In some implementations the projected image from the field can be incident on the front side of a semitransparent screen. The cameras can observe the backside of this screen, where the intensity is drastically reduced. The screen could be made of glass, fabric, or thin ceramic.
In some implementations, the reflective elements can be faceted to more directly concentrate the sunlight on the receiver. In other implementations, the reflective elements can have a pyramidal or three-dimensional polygonal shape.
In some implementations, where the reflective elements are approximately tubular (as shown in
In controlling the speed of the heliostat, if the error is large, the heliostat can be moved at a high velocity. As the error gets small and the heliostat approaches the desired position, the speed can be decreased. By shaping the reflective bar differently, the error signal automatically changes more significantly when the error is large, but slower when the geometries are just about right. So given a constant velocity motor or mechanism, the error signal derived from the reflection, changes more or less rapidly.
In some implementations, an individual heliostat can be controlled to direct reflected light slightly off the center of the receiver, and a group of heliostats can be controlled to direct reflected light slightly at different locations (at least some of which are at different positions off the center of the receiver). By directing the reflected light at slightly different locations on the receiver, flux from the heliostats can be smoothed out on the receiver.
Referring to
Using the above system, on-center control can be accomplished by adjusting the orientation of the heliostat mirror 160 so that the error signal goes to zero. In addition, a particular heliostat can be pointed slightly off center using this system by adjusting the orientation of the heliostat mirror so that the error signal goes to a small but non-zero target value. This non-zero value can be, e.g., up to 5% of the maximum value 802 or minimum value 806. The value of 5% above is merely an example; given a maximum desired spillage, e.g., a beam that is off-center by 10% of the width of the receiver, corresponding maximum and minimum target values can be determined empirically. By selecting different target values for at least some of the heliostats, these heliostats can be pointed at different locations of the receiver. While the discussion above and
In some implementations, a reflective element is located proximate to the center of the heliostat. The reflective element can be configured to always reflect the sun when the mirror is aligned in the general direction of the sun, i.e. when the mirror is within 270° of the sun. For example, the reflective element can be generally hemispherical and have no masked sections. Therefore, provided that the mirror wasn't within the 90° window directly opposite the sun, the reflective element will always reflect light into the reflector and/or camera. The reflective element can therefore always allow the controller to find the center of the heliostat.
In some implementations, the reflective elements can include cross-hatches that appear different to the camera depending upon the angle.
Having individually identifiable reflective elements on or near the surface of the mirror can allow the controller to identify a particular orientation of a heliostat. Determining a particular orientation of the heliostat can allow for rotation of the heliostats to provide a closed-loop heliostat control system that ensures that sunlight is reflected from each heliostat into the desired receiving location. Given the available speed of image processing, errors in the heliostat reflection can be controlled on a real-time, or near-real time basis. Such a system allows concentrated sunlight to enter the receivers for a large fraction of the day in order to provide sufficiently high temperatures for the creation of solar power.
Particular embodiments have been described. Other embodiments are within the scope of the following claims.
