Autonomous heliostat with automatic calibration via accelerometer and solar aiming

Abstract

The present invention relates to an autonomous heliostatic device designed to orient a reflective surface so as to reflect sunlight in a fixed direction throughout the day, independently of the sun's movement. The innovation lies in the combined use of an accelerometer and a solar aiming system to determine the position of the reflective surface on the one hand, and that of the heliostat's rotation axes in the local terrestrial frame of reference on the other. This combination allows the heliostat to position itself automatically, without user intervention, without risk of detachment, without casting a shadow, and regardless of its position, in order to operate completely autonomously. FIGURES BRIEF DESCRIPTION OF DRAWINGS [Fig. 1]: Illustration of an embodiment [Fig. 2]: Solar aiming system [Fig. 3]: Direct orthonormal frames of reference [Fig. 4]: Unit vectors [Fig.[5]: Determination of the z-axis of the heliostat in the accelerometer's frame of reference. [Fig. 6]: Possible positions of the sun in the accelerometer's frame of reference. [Fig. 7]: Detection of the center of the sun. [Fig. 8]: Construction of the x and y vectors of the Earth's frame of reference in that of the accelerometer. [Fig. 9]: Control of the mirror's position in the heliostat's frame of reference by the motorization system (2). [Fig. 10]: Calculation of the target's position given the initial positions of the mirror and the sun.

Description

Title of the invention: Autonomous heliostat with automatic calibration by accelerometer and solar aiming. FIELD OF THE INVENTION

[0001] This invention falls within the field of solar technologies, more specifically within that of heliostats. A heliostat is a device that reflects sunlight to a fixed point throughout the day, despite the sun's movement. This makes it possible to use solar energy where it would not naturally reach, or where it is poorly oriented. Heliostats are, of course, used in concentrated solar power systems, where they appear in large numbers to focus solar energy to a fixed point, but they also have applications in the home. They have great potential there, as they allow sunlight to be reflected to poorly oriented rooms, thus providing lighting and heating without the consumption of fossil fuels. Since the energy reflected by the reflective surfaces is commonly more than 90% of the incident flux, they also exhibit excellent efficiency. STATE OF THE ART

[0002] Heliostats have seen significant progress since the mid-18th century, primarily to adapt them for energy production by concentrating solar radiation. This explains why several techniques already exist for directing sunlight onto a given target, but these are generally reserved for this specific use. Since then, new, more general-purpose systems have emerged.

[0003] First, there are solutions based on fixed mirrors of various shapes, arranged so that at least one mirror always reflects the light in the correct direction at any given time. As the sun moves, another mirror reflects the rays back towards the target, and so on. For these systems to operate all day, many small mirrors with different orientations are required, each of which is actually used for only a minimal amount of time on a daily basis. It is therefore a simple system, but with poor efficiency.

[0004] To make the mirror profitable and use it 100% of the time, it must rotate, but calculating its position over time, which depends on the sun's movement, is difficult. The simplest system for adequately controlling a rotating mirror, currently found in domestic applications, is therefore the use of a feedback loop based on the reflected ray (example patent FR3113311). The orientation of the ray reflected by the mirror is then used to adjust its position, bypassing the calculation problem. of solar motion. This system eliminates the need for a microcontroller and avoids tedious calculations. Unfortunately, it has the disadvantage of having to block some of the reflected rays to control the feedback, thus generating a shadow on the returned image, which is unsightly and reduces efficiency. Furthermore, these systems often have the flaw of being able to "disengage." Indeed, if the sun does not illuminate the heliostat for a certain period (due to cloud cover or simply because of night), the reflected ray, when it reappears, may have moved far enough to no longer be detected by the sensors controlling the feedback. The system must then be reset, which requires human intervention and is undesirable for regular use.

[0005] A device that does not present these disadvantages is the remote-controlled system. This is simply a mirror mounted on two axes, the position of which can be adjusted remotely by the user using a remote control. It is therefore the user who, manually and unfortunately frequently, must direct the sun's rays towards the desired location to compensate for the sun's movement. This solution obviously requires far too much human intervention.

[0006] To solve the problems of shadowing and slackness without human intervention, microcontroller-controlled systems have been developed, primarily in concentrated solar power plants. Thanks to precise calibration and meticulous adjustment by specialized technicians on site, these microcontroller-controlled systems can efficiently reflect sunlight to a fixed point. They contain the equations for determining the sun's position at any given time and calculate the orientation of the mirror so that the ray is reflected in the correct direction. Unfortunately, these systems require painstaking adjustment, as the positions of their rotation axes and their mirror must be precisely determined in the local Earth-based reference frame.These calibrations are too complicated for use by "non-specialists", and in practice it is too complicated to have a specialized technician come on site each time the heliostat is started up to perform them.

[0007] One solution is to ask the user to position the heliostat themselves in a reference position. Typically, they might be asked to place the heliostat perfectly horizontally using a spirit level and orient it towards a predefined azimuth. This setup phase is tedious, carries a risk of being performed incorrectly, and only allows for one configuration of use for the heliostat. For example, it is impossible in this case to fix it perpendicular to a tree trunk, on sloping ground, or upside down without complex modifications. While not an insurmountable obstacle, this limits the heliostat's usability while complicating the user experience.

[0008] To circumvent the previous problem, one solution is to leave the placement of the heliostat open, but to equip it with sensors that allow it to deduce its position independently. The user then attaches the device arbitrarily, and the microcontroller subsequently determines the heliostat's position in the Earth's frame of reference during a calibration phase. Generally, two reference points are sufficient for locating the device. An accelerometer can be used to locate the zenith, and an (electronic) compass to orient the device relative to the cardinal directions. With more complex calculations, this solution works, at least theoretically. However, searches for "heliostat compass," "heliostat compass," or "heliostat IMU" returned no results in the INPI database as of August 27, 2024, even though it is highly likely that systems based on this principle have already been tested.From our experience, the use of an electronic compass is a bad idea for three major reasons, which explain why no convincing system has been developed so far: . a. Because the Earth's magnetic field is inherently very weak, compasses are easily affected by surrounding magnetic and paramagnetic objects. The information they provide can be distorted by the mere presence of nearby electric motors. To try to eliminate these disturbances, a dual calibration (soft iron / hard iron) is necessary, ideally performed on-site. Unfortunately, even with this calibration, the results are not always sufficiently accurate. b. Another problem with compasses is that they indicate magnetic north, not true north, which is necessary for the calculations performed by the microcontroller. The difference in angle between these two "norths" is called magnetic declination, and it is not constant with respect to location or time. The magnetic masses moving within our planet affect the magnetic field and cause its orientation to vary, thus skewing the data used. While some models exist to correct this error, they must be implemented in the microcontroller, and even then, there remain local variations in the magnetic field that the models cannot predict. The resulting errors cannot be known in advance, making the commercialization of such systems highly uncertain. c. Compasses are sensitive to temperature. However, a heliostat installed outdoors can be exposed to temperatures ranging from -20 to +50°C. While prior calibration is possible to try to minimize this phenomenon, there remains a risk of inaccuracy in determining North. Furthermore, we We observed that temperature gradients seemed to pose a real problem of accuracy in the information transmitted by electronic compasses.

[0009] These three phenomena limit the accuracy of heliostats calibrated by electronic compass to 5–10° of angle. However, the expected accuracy of a good heliostat is on the order of 1° of angle. Currently, we have not succeeded, with a prototype operating on the basis of an electronic compass, in obtaining a system that meets the requirements of a large-scale commercialization project.

[0010] The invention presented in this patent retains the advantages of a microcontroller-based system while overcoming the difficulties inherent in using a compass, by employing a new calibration method that allows the heliostat to determine its orientation on the Earth's surface without human intervention and with the necessary precision. It therefore operates accurately regardless of its position, without user intervention, without risk of loss of alignment, and without casting a shadow. DETAILED DESCRIPTION

[0011] The present invention relates to a self-contained heliostatic device with automatic calibration by accelerometer and solar aiming. This device is designed to orient a reflective surface so as to reflect sunlight in a fixed direction throughout the day, independently of the sun's movement.

[0012] Main innovation:

[0013] The innovation lies in the combined use of an accelerometer and a solar targeting system to determine the positions of the reflective surface and the rotation axes of the heliostat in the local Earth reference frame. This combination allows it to locate itself and operate automatically regardless of its orientation, without user intervention, without risk of stalling, and without casting a shadow.

[0014] Description of components:

[0015] The device comprises: a. a reflective surface (1), mounted on two axes of rotation, b. a motorization system (2), allowing adjustment of the orientation of the surface (1), c. an accelerometer (3), which provides the local upward vertical, d. a solar sighting system (4), which detects the position of the sun, e. a microcontroller (5), which receives information from the sensors and provides the positioning orders, f. a power supply (6), which powers the various components, g. and a system for determining the theoretical position of the sun at any point on the globe and at any time in the local terrestrial reference frame (7) Details of how it works:

[0016] Initial calibration phase: The microcontroller (5), powered by the power supply (6), uses the data provided by: a. the accelerometer (3), b. solar sighting (4), c. and the system for determining the position of the sun (7), in order to determine, in the local terrestrial frame of reference: a. the orientation of the rotation axes of the device, b. the orientation of the reflective surface (1), c. and the orientation of the target towards which it must direct the sun's rays.

[0017] Real-time tracking

[0018] This allows it, in a second step, to give the motorization system (2) the necessary instructions to position the reflective surface (1) so as to keep the reflected solar ray fixed in the Earth's frame of reference. This operation is repeated indefinitely at regular intervals.

[0019] Alternative embodiments:

[0020] In certain embodiments, a. The reflective surface (1) can be curved, concave, convex, and made of mirror-polished metal or glass, b. The motorization system (2) may consist of stepper or continuous motors, with or without an encoder, oriented at 90° to each other or at a different angle, c. The accelerometer (3) can be replaced by an inertial mass system, or a system that actually detects the nadir, d. The solar aiming system (4) may consist of a matte tube containing a photodiode, or of solar panels whose received intensity indicates the orientation of the sun, e. The power supply (6) can be a battery or an external power source, f. The system for determining the theoretical position of the sun in the Earth's reference frame (7) may rely on a real-time clock (RTC) and a GPS chip, or use the time transmitted by certain radio frequencies, or use manually entered data.

[0021] These alternative embodiments are not exhaustive and are simply given as an example.

[0022] PRESENTATION OF A METHOD OF IMPLEMENTATION

[0023] Mechanical part

[0024] A diagram illustrates the device [Fig. 1]. A standard mirror coated with a varnish to protect it from moisture is used as the reflective surface (1). The drive system (2) consists of two stepper motors with 200 steps per revolution, mounted at 90° to each other, each coupled to a 1:50 reduction gear and controlled via a dedicated electronic board. The motor directly connected to the base, at the bottom of [Fig. 1], is the main motor, which controls the rotation around the main axis. The other motor, to the right of [Fig. 1], is the secondary motor, which rotates the mirror around the secondary axis. The accelerometer (3) is provided via a standard IMU chip. The sun sighting system (4) is a matte tube containing, at one end, a photodiode facing inwards towards the tube, connected in series with a resistor [Fig. 2].When this tube is pointed towards the sun, the photodiode conducts current, increasing the voltage drop across the resistor, which can be detected by the microcontroller (5). The microcontroller (5) can be any commercially available chip; here, an Arduino Uno module was chosen. Power is supplied (6) via a battery, photovoltaic panels, and a dedicated electronic board managing charging and discharging. The system for determining the sun's position relative to the Earth's frame of reference (7) consists of a real-time clock (RTC) coupled to a GPS chip and a set of equations from astrophysics that predict the sun's theoretical position for the coming years (these equations are pre-programmed into the microcontroller (5)).

[0025] The mechanical structure can be largely manufactured using 3D printing. The main axis is guided by a metal tube rotating within another tube; conventional ball bearing guidance can also be used. The secondary axis is guided by a metal tube sliding within a 3D-printed bearing. A housing containing the accelerometer (3) is attached behind the mirror (1). The solar sighting system (4), consisting of a 3D-printed tube, is also attached behind the mirror (1).

[0026] Reference frames and vectors used

[0027] The device manages to automatically locate itself on Earth by using, for orientation, several direct orthonormal reference frames and specific normalized vectors [Fig.3][Fig.4]: a. The local terrestrial reference frame (R?), whose three axes are: i. xT, directed towards the South at the location of heliostat use, ii. y directed towards the East at the location of heliostat use, iii. zT, directed towards the zenith instead of using the heliostat. b. The Accelerometer reference frame (RAh) whose position of the three axes xA, VA ct Z a is not imposed a priori and is determined when fixing the accelerometer on the mirror. c. The fixed CR-Hb Heliostat reference frame in Ry, whose three axes are: i. XH, perpendicular to z^ belonging to the plane xTlzT and such that the scalar product of x? and x^ is 0, ii. y, direct with and iii. Zjp oriented along the main axis of the heliostat, from its base towards the mirror. d. g, oriented towards the center of the sun from the center of the heliostat. e. oriented towards the zenith of the location where the heliostat is situated. Do not to be confused with the "z" axes of the reference frames, which are written in lowercase. f. ~q, which defines the desired orientation of the reflected ray, from the center of the heliostat towards the target. g. y, which defines the orientation of the solar sighting system, h. which defines the orientation of the mirror. This vector is perpendicular to the surface, coming out of the reflective side. i. 3, which defines the orientation of the secondary axis, from the secondary motor outwards from the heliostat.

[0028] When a vector is described in a given reference frame, it will be noted between parentheses, with the subscript being the letter specific to the reference frame in which it is expressed. For example, the "y" axis of the accelerometer base, when expressed in the terrestrial reference frame, will be noted

[0029] Factory calibration

[0030] For the calculations to proceed correctly, the vectors y and 3 must be known beforehand in the accelerometer's frame of reference. Since these vectors are physically attached to the accelerometer, they are fixed in the RA basis and can be easily determined. This is done during construction, either by imposing these values ​​and fixing the accelerometer to the mirror accordingly, or by fixing the accelerometer and measuring these values ​​afterward. The method presented here is the second, which is easier to implement manually. To calibrate y, the sun sight is oriented vertically with a plumb line, and y is assigned the value read by the accelerometer at that instant. For , a beam of light is shone vertically and downwards onto the mirror, and the value read by the accelerometer when the mirror is oriented so that the reflected ray returns to its source. To calibrate 5, the procedure is the same as in step

[0034] .b below, but with the secondary motor running instead of the main motor. The resulting vectors are then normalized to make them unitary. These manual calibrations can be easily automated within a high-volume manufacturing process.

[0031] Use / Internal calculations

[0032] The use of the heliostat then takes place in 3 steps for the user and 10 steps for the microcontroller. The only constraints that apply at the time of calibration are the following: the sun must be clearly visible (sunlight must be able to reach the heliostat, in particular the sun sight), and the sun must not be perfectly at the zenith (which can only occur between the tropics, at a very specific time of day).

[0033] User: a. The user attaches the heliostat to any support, b. He positions the mirror so that the reflected ray is in the correct direction, by manually manipulating it (the proposed motor-reducer pair offers sufficient braking to prevent unwanted movements due to wind, while still allowing manual manipulation). c. He presses the ON button.

[0034] Microcontroller: a. It calculates the position of the sun in the Earth's reference frame using pre-recorded astrophysics equations, the GPS chip (6) and the RTC (7), and stores this value. b. It commands a full rotation of the main motor by taking measurements of the zenith vector at regular intervals. These vectors form a cone around the main axis of rotation of the heliostat [Fig. 5]. The axis of the cone gives the direction of 1, and the order of acquisition of the points gives its direction. (If the vectors are too collinear and do not define a sharp cone, they can simply be added together.) From this point on, hja is known for all t instant in RA, because we can maintain its position by rotating it around 3 when the secondary motor rotates by + OC. c. It updates the position of the sun in the Earth's frame of reference, and in particular determines the angle between the sun and the zenith at that instant (the complement of the elevation). Knowing the position of the zenith in the accelerometer's frame of reference, this allows it to construct a circle representing all the points where the sun can be located [Fig. 6]. d. f. g- Knowing both the axes of rotation of the motors and Çgj, he does point the sight towards each of the points constituting this circle, until The photodiode detects an edge of the sun. During each movement of the secondary axis, it does not forget to update the position of [7), as described in the article on the photodiode. point b. He then applied several movements to the solar sighting system, having the aim is to locate the center of the sun precisely (the detection threshold used here must be slightly more sensitive than in the previous step, to avoid "missing" the sun) [Fig.7]: i. It commands a rotation of the main motor in an arbitrary direction until the photodiode no longer detects the sun. ii. It goes back and passes through the sun, measuring its entry angular position and its exit angular position. iii. The main engine is then positioned to aim between these two points. iv. The same operation is performed for the secondary motor. When this is done, the sight then points towards the center of the sun. Known by the accelerometer, pjy-j by factory calibration, yl T using the equations of astrophysics, and Pæ' definition, he then calculates the position of the Earth's frame of reference in the accelerometer's frame of reference for this particular position, knowing that pyj and point in the same direction [Fig. 8]. We therefore find , and by transposition (R^ for this particular position, (y ) being known from point

[0034] .b v hj HAS Since cl(R^ has just been determined, we find . By definition of R then we can build easily and by transposed, we find. The relations R^Rj are therefore now known. Since the factory setting is known, and knowing the position of the vector in this position, we can find ct, therefore j. We can then calculate the angles that this vector makes with the axes of the heliostat's reference frame [Fig. 9]. These angles can now be directly controlled by the motors in order to place the mirror in a given position, whether that position is specified in RT or R^. h. Having recorded all the movements made by the motors since the ON button was pressed, we can return to the initial position of the mirror in R^ when this button was activated, and transpose it into RT. Having recorded the position of the sun in RT at that instant, we find, by Snell's law applied to reflection, the position of the target vector (c) 1¾1°]. i. Now that all the data necessary for the calibration phase has been obtained, the microcontroller can regularly update the sun's position, as well as the orientation of the mirror to ensure the reflected beam remains on its target. This position is then transferred to the heliostat's frame of reference, where it is converted into rotation commands for each motor. By repeating this operation at regular intervals, the reflected beam is maintained at the desired position. j. A night mode puts the Heliostat in a neutral position when the sun has set. This resting position places the photovoltaic panels in the direction of sunrise, in order to guarantee a rapid energy supply if the battery were to discharge during the night.

[0035] The device described here was successfully tested at different angles in August 2024, and gave excellent results for optimal ease of use.

Claims

Demands

1. A heliostatic device characterized in that it comprises a reflective surface (1), a motorization system (2), an accelerometer (3), a sighting system enabling it to locate the sun (4), a microcontroller (5), a power supply (6), and a system enabling it to determine the position of the sun in the Earth's frame of reference (7). The sighting system (4) consists of a matte hollow tube at the bottom of which is a photodiode.

Images