Method for controlling a photovoltaic plant and relative control system

Abstract

Described is a method for controlling a photovoltaic plant positioned in a geographical area, wherein the photovoltaic plant comprises at least one photovoltaic panel, which is configured to rotate about an axis of inclination according to a respective angular configuration and to provide a predetermined energy production value. The method includes a learning step which comprises processing production information acquired in a massive manner from the photovoltaic panel. The production information comprises a plurality of energy production data, associated with the energy production value, which are acquired within a specific time interval, and, for each energy production data, the production information additionally comprises the respective production parameters, that determined the energy production data. The production parameters comprise an angular configuration of the photovoltaic panel and environmental parameters acquired in the vicinity of the photovoltaic plant. The control method comprises creating, using machine learning, a model of the photovoltaic plant suitable for analysing the production information in order to identify a set of production parameters correlated with each other, and correlated with the respective energy production data, which have over time influenced the energy production values of the photovoltaic panel. The relative control system of the photovoltaic plant is also proposed.

Description

[0001] DESCRIPTION

[0002] Method for controlling a photovoltaic plant and relative control system

[0003] The present invention relates to a method for controlling a photovoltaic plant and the relative control system.

[0004] A photovoltaic plant comprises a plurality of photovoltaic panels. As is known, a photovoltaic panel is essentially an object comprising one or more photovoltaic modules, each provided with a respective capture surface intended to be irradiated by the solar radiation. In a manner known in the sector, each photovoltaic module has a flattened parallelepiped shape, defined by two main surfaces separated by a distance, which expresses a thickness of the module and is decidedly less than the sides of the main surfaces. Of the two main surfaces, one is the capture surface, which is opposite a rear surface of the module. As is known in the sector, each photovoltaic module is capable of converting the energy carried by the solar radiation into an electric potential difference, which is in turn converted into an alternating electric current by a field inverter associated with the photovoltaic panel.

[0005] A photovoltaic panel comprises a support frame for supporting the one or more photovoltaic modules, in such a way that the capture surfaces are substantially coplanar with each other. The capture surfaces of the photovoltaic modules together define a capture surface of the photovoltaic panel.

[0006] In order to ensure that, in ideal weather conditions without fog or clouds, the capture surface intercepts the maximum possible portion of solar radiation, that is, has an optimal exposure to it, each photovoltaic panel must be positioned so that it presents a theoretical angle of inclination at which, with respect to the axis of inclination, the projection of the capture surface onto a plane perpendicular to the direction of the solar rays is as great as possible.The angle of inclination of the photovoltaic panel refers to the slope relative to the ground on which each panel is installed and, in geometric terms, expresses the angle that the panel forms with the horizontal line of the ground.

[0007] The theoretical angle of inclination depends substantially on the time of day, that is, on an angle of elevation of the sun, and therefore changes progressively during the day. The theoretical angle of inclination also changes in relation to the latitude of the geographical area in which the photovoltaic panel is positioned and the time of year. In fact, the seasons influence the production of the photovoltaic panels both because of the position of the ground relative to the sun and because of temperature and other meteorological factors.

[0008] For this purpose, based on the angle of elevation of the sun at each time interval of interest, it is possible to define the respective theoretical angle of inclination, for each day of the year and based on the geographical area of installation of the panel itself. It should be noted that the theoretical angle of inclination can be provided in the form of predefined tabular data, or it can be calculated by means of GPS astronomical time.

[0009] In the highest-performance photovoltaic plants, each solar panel is configured to rotate about the axis of inclination for a predetermined angular range of inclination. The angular range of inclination depends on the angle of elevation of the sun from sunrise to sunset and the geographical area of installation of the photovoltaic panels. As is known in the sector, the axis of inclination is oriented North-South so that the photovoltaic panel can follow the trajectory of the sun from East to West during the daylight hours.

[0010] Within the predetermined angular range of inclination, each photovoltaic panel is configured to rotate at predefined time intervals in order to position the panel at its theoretical angle of inclination, so as to ensure optimal exposure and maximise electricity production during the daylight hours. For this purpose, the photovoltaic panel comprises a movement device configured to rotate the photovoltaic panel.Under optimal weather conditions, a nominal energy production of the photovoltaic panel can therefore be calculated, which corresponds to the energy production that the photovoltaic panel could produce under ideal conditions, rotated and positioned in a plurality of positions corresponding to the respective theoretical angles of inclination so as to guarantee ideal exposure throughout each day.

[0011] However, numerous other factors, in addition to optimal exposure, can influence the energy production of the photovoltaic panels. For example, atmospheric conditions with the presence of rain, fog or diffuse cloudiness, which are far from optimal atmospheric conditions, can influence the energy production of the photovoltaic panels to a much greater extent than optimal exposure, because both the power of the solar irradiation and the direction of origin of the solar irradiation are altered.

[0012] It is therefore known that the energy production of photovoltaic panels can also be very different from the nominal energy production which can be obtained under optimal conditions.

[0013] The current photovoltaic plants are not able to take into account the influence of atmospheric conditions on the energy production of the photovoltaic panels and it may therefore happen that the theoretical angle of inclination associated with a specific time interval is not the one that allows the greatest energy production.

[0014] It should be noted that the forecasts of solar energy production (for example, over an hour, a day, or a period of time) have become a service of fundamental importance for different parties present on the market, for various purposes.

[0015] For example, the production forecasts can be significant for solar energy producers in order to optimise the a configuration of the photovoltaic plant or to plan the management operations, such as maintenance or cleaning of the photovoltaic plant.

[0016] On the other hand, the energy service companies that sell energy to end consumers may need to know the production forecasts of a photovoltaicplant that supplies them with energy, in order to organise the management of the market, as well as the transmission and distribution of electricity. In fact, considering various local electricity networks and since each of them may involve generation by means of a traditional fossil-fuel energy production system or by means of electricity generation systems from renewable energy sources, the production of each local electricity network must be integrated with the entire national electricity network.

[0017] The aim of the present invention is to provide a method for controlling a photovoltaic plant and a related control system that are free of the drawbacks described above and that, at the same time, are easy and economical to implement.

[0018] Another aim of the present invention is to provide a method for controlling a photovoltaic plant and a related control system that guarantees the highest possible energy production and at the same time allows a reliable prediction of the energy production.

[0019] According to the present invention, a method for controlling a photovoltaic plant and a control system are provided according to the accompanying claims.

[0020] Features and advantages of the invention are more apparent from the detailed description which follows of an embodiment of the invention according to the invention, illustrated by way of a non-limiting example in the accompanying drawings in which:

[0021] - Figure 1 shows a front schematic perspective view of a photovoltaic plant, comprising a photovoltaic panel and an image acquisition device configured to frame the photovoltaic panel and / or the sky;

[0022] - Figure 2 shows the photovoltaic panel of figure 1 , seen from the back; - Figure 3 shows a side view of the photovoltaic panel of Figure 1 and of a radiation detector, both of which are connected to a position controller of a control unit of the photovoltaic plant;

[0023] - Figure 4 shows the photovoltaic panel of Figure 1 having a first end inclination of a predetermined angular range of inclination;- Figure 5 shows the photovoltaic panel of Figure 1 having a second end inclination of the angular range of inclination;

[0024] - Figure 6 shows the control system of the photovoltaic plant, which comprises a single image acquisition device;

[0025] - Figure 7 shows an enlarged variant of the control system of Figure 6, wherein each unit of four photovoltaic panels is associated with a respective image acquisition device.

[0026] With particular reference to the accompanying drawings, a control system 1 for a photovoltaic plant 2 is described.

[0027] It should be noted that control system 1 according to the present invention is particularly suitable for the control of photovoltaic plant 2, positioned in a geographical area, which comprises one or more photovoltaic panels 201. Reference will be made in the following description to a single photovoltaic panel 201 for greater simplicity, but what is said may be referred to any number of photovoltaic panels 201.

[0028] As shown in Figures 1 to 5, the term photovoltaic panel 201 is meant to be essentially an object comprising one or more photovoltaic modules 202. Each photovoltaic module 202 is provided with the capture surface 202a of the module intended for being irradiated by the solar radiation. As is known in the sector, each photovoltaic module 202 has a flattened parallelepiped shape, defined by two main surfaces separated by a distance, that is, the thickness of module 202, which is less than the sides of those main surfaces. Of the two main surfaces, one is the capture surface 202a of the module, which is opposite a rear surface 202b of the module 202 itself. In the preferred but not exclusive embodiment shown, the photovoltaic modules 202 of the panel 201 are mounted on a support frame 203 so that all the capture surfaces 202a of the module are substantially coplanar to each other and define overall a capture surface 204 of the photovoltaic panel 201.

[0029] Each photovoltaic panel 201 is configured to rotate about an axis of inclination X and to provide a predetermined value of energy production.In detail, each photovoltaic panel is configured to rotate about the axis of inclination X by an angle of inclination a, shown in Figure 3. For this purpose, the support frame 203 of each photovoltaic panel 201 is associated with a supporting structure 205, allowing rotation about the axis of inclination X and the photovoltaic panel 201 comprises a movement device configured to rotate the photovoltaic panel 201.

[0030] The movement device may comprise an actuator 206’, installed on the support frame 203 and configured to perform the movement of the photovoltaic panel 201 , a field control device 206”, shown schematically in Figures 3 and 6 to command this movement, and an inclination sensor (not illustrated), also installed on the support frame 203 to ensure that the angle of inclination a, which is commanded to the actuator 206’, is effectively reached. The field control device 206” may, for example, be a PLC (programmable logic controller) that commands the actuator 206’ to rotate and receives from the inclination sensor the angle of inclination reached by the photovoltaic panel 201 at the end of the rotation. The field control device 206” may be positioned in the vicinity of the photovoltaic panel so as to allow a wired connection directly to the actuator 206’ and to the inclination sensor, even though this is not necessary.

[0031] As is known in the sector, each photovoltaic module 202 is capable of converting the energy carried by the solar radiation into an electric potential difference, which is in turn converted into a direct current and then supplied to the power grid as alternating current.

[0032] In fact, each photovoltaic panel 201 comprises a respective field inverter 207, shown in Figure 6, which is capable of converting the direct current generated by the photovoltaic panel 201.

[0033] The field inverter 207 is therefore configured to provide the determined energy production value.

[0034] It should be noted that, in relation to the configuration of the photovoltaic plant 2 and the number of photovoltaic panels 201 present, the field controldevice 206”, as well as the field inverter 207, can command and / or be associated with one or more photovoltaic panels 201 .

[0035] By way of example, Figures 1 and 2 show a photovoltaic plant 2 comprising a pair of photovoltaic panels 201 , each of which comprises a plurality of photovoltaic modules 202, which are supported by the same support frame 203 and by the same supporting structure 205 and are, therefore, simultaneously moved in rotation by the actuator 206’. This pair of photovoltaic panels 201 may have associated with it a single field inverter 207 shared between them, and therefore the two photovoltaic panels 201 may be considered as a single photovoltaic panel 201 with a double capture surface 204.

[0036] In Figure 6, in which the control system 1 of the photovoltaic plant 2 is shown, it should be noted that the photovoltaic plant 2 has been shown as comprising twelve photovoltaic panels 201, divided into three different groups each comprising four photovoltaic panels 201. The four photovoltaic panels of each group are associated with a single field control device 206” and with a single field inverter 207, but it should be noted that this is provided only byway of example, since the field control device 206” and the field inverter 207 can be associated, in general and without restriction, with one or with any number of photovoltaic panels 201 , depending on the dimensions and number of photovoltaic panels 201 present in the photovoltaic plant 2. The number of photovoltaic panels 201 in a group may vary, precisely because it can depend on the shape of the ground and the geographical area in which the photovoltaic plant 2 is installed.

[0037] Regardless of the configuration of the photovoltaic plant 2, the fact remains valid that each photovoltaic panel 201 comprises an actuating device, comprising the actuator 206’ and the field control device 206”, and a field inverter 207, which may be for exclusive use or shared, at least partially, with other panels.

[0038] The angle of inclination a of the photovoltaic panel 201 expresses the slope with respect to the ground on which each photovoltaic panel 201 is installedand, in geometric terms, expresses the angle that the photovoltaic panel 201 forms with respect to the horizontal line of the ground, as shown in Figures 1 to 5.

[0039] It should be noted that the photovoltaic panel 201 is configured to rotate about the axis of inclination for a predetermined angular range of inclination, which is between a first end inclination (31 (for example +55°), shown in Figure 4, and a second end inclination [32, shown in Figure 5, for example -55°.

[0040] The angular range of inclination between the first end inclination and the second end inclination, which in the example shown in Figures 4 and 5 is between +55° and -55°, depends on an angle of elevation of the sun from sunrise to sunset and on the geographical area where photovoltaic plant 2 is installed. The axis of inclination X is typically oriented North-South so that each photovoltaic panel 201 of the photovoltaic plant 2 can follow the trajectory of the sun from East to West during the daylight hours.

[0041] As already described above, in order for the capture surface 204 to intercept the maximum possible portion of solar radiation, in ideal weather conditions free from fog or clouds, the photovoltaic panel 201 can be positioned at a theoretical angle of inclination, based on the angle of elevation of the sun, which maximises, with respect to the axis of inclination, the projection of the capture surface 204 onto a plane perpendicular to the direction of the solar rays.

[0042] The theoretical angle of inclination depends substantially on the time of day, that is, the angle of elevation of the sun, changes progressively during the day and depends on the latitude of the geographical area in which photovoltaic panel 201 is located and on the time of year.

[0043] For this reason, under ideal weather conditions, the photovoltaic panel 201 is rotated at intervals of time and can be positioned in a plurality of positions corresponding to respective theoretical angles of inclination so as to ensure the ideal exposure during each day.A nominal theoretical energy production of the photovoltaic panel is defined as the energy production in ideal weather conditions.

[0044] However, as will be described in detail below, in the control system 1 of the present invention, for the photovoltaic plant 2 in turn comprising the photovoltaic panel 201 having a certain energy production value, the photovoltaic panel 201 may be configured to rotate about the axis of inclination X according to an angular configuration, which may comprise not only the theoretical angle of inclination.

[0045] The control system 1 comprises a storage platform 3 configured for storing production information 301 acquired in a massive manner from the photovoltaic panel 201.

[0046] The production information 301 comprises energy production data, which are associated with the energy production value and are acquired within a specific time interval.

[0047] For each energy production data, the production information 301 also includes the respective production parameters that have determined the energy production data in the predetermined time interval.

[0048] The production parameters comprise the angular configuration of the photovoltaic panel 201 and environmental parameters acquired in the vicinity of the photovoltaic plant 2.

[0049] In other words, the production information 301 stored in the storage platform 3 is indicative not only of the production value acquired by the photovoltaic panel 201 in the predetermined time interval, but also of the additional energy production data associated with that production value and of the environmental parameters that have determined that production value. It should be noted that in a photovoltaic plant 2 with a plurality of photovoltaic panels, for each photovoltaic panel 201 the respective production information 301 is acquired. In fact, although in theory each photovoltaic panel 201 forming part of the photovoltaic plant 2 should have, within the same time interval, the same angular configuration and the same energy production data, this may not always be the case, depending on theamplitude of the geographical area and the different morphology of the ground across which the plurality of photovoltaic panels 201 are installed. On the contrary, it is assumed that the geographical area in which the photovoltaic panels 201 of the plant are positioned is limited and that therefore all the photovoltaic panels 201 have the same environmental parameters as they are all subject to the same environmental conditions. However, as described in detail below, this may not occur, as it also depends on the size and extension of the photovoltaic plant 2.

[0050] The control system 1 comprises a control unit 4 which comprises:

[0051] - an inverter controller 401 , connected to the field inverter 207 of the photovoltaic panel 201 and configured for acquiring the plurality of energy production data of the photovoltaic panel 201 , in the predetermined time interval;

[0052] - an environmental meteorological unit 402, positioned in the vicinity of the photovoltaic plant 2 and configured for acquiring, for each energy production data, the relative environmental parameters of the production parameters; - a position controller 403, connected to the movement device of the photovoltaic panel 201 and configured to determine the angular configuration of the photovoltaic panel 201 in the predetermined time interval.

[0053] The environmental meteorological unit 402 and the position controller 403 are, therefore, configured to acquire the environmental parameters associated with each production information 301.

[0054] The control system 1 comprises a management server 5 which comprises a processing component 501 , which is configured, during a learning step, to process said production information 301 stored in the storage platform 3 and to create, by machine learning, a model 502 of the photovoltaic plant 2 suitable for analysing the production information 301 in order to identify a set of production parameters correlated with each other, and correlated with the respective energy production data, which have over time influenced the energy production values of the photovoltaic plant 2.Thanks to the fact that the management server 5 is able to create in the learning step the model 502 of the photovoltaic plant 2 by machine learning, since the processing component 501 is able to analyse the production information 301 stored in a massive manner in the storage platform 3, it is possible to determine how the energy production data are correlated with the production parameters and how they have been influenced by these over time. In fact, the large quantity of acquired production parameters, which comprise not only the angular configuration of the photovoltaic panel 201 but also the relative environmental parameters, allows long-term considerations to be made to ensure that the model 502 can take into account all the environmental conditions to which the photovoltaic panel 201 is subject over time.

[0055] The processing component 501 is configured to process by machine learning the production information 301 stored in the storage platform 3, (the energy production data of the photovoltaic panel 201 and the respective production parameters, which include the angular configuration of the photovoltaic panel 201 and the environmental parameters acquired in the vicinity of the photovoltaic plant 2) and to consider the production value associated with each production information 301 stored as a target result to be evaluated.

[0056] In fact, in more detail, the processing component 501 comprises a plurality of modules, not illustrated, configured to create by machine learning the model 502 of the photovoltaic plant 2.

[0057] The processing component 501 may optionally comprise a normalisation module, configured to check any spurious and / or incomplete and / or missing and / or incorrect production information 301 and normalise it, so as to make all production information 301 mutually congruent and normalised before further processing. For example, the normalisation module may be configured to carry out linearisations and interpolations if the production information 301 records are absent, levelling of maximum and minimumvalues, transformations of peaks, and classification of off-scale values, for each production information 301.

[0058] The processing component 501 comprises, in addition:

[0059] - a training module, not illustrated, configured to apply a supervised training on a first unit of normalised production information (equal to approximately 80% of the total production information 301 stored), using the energy production value associated with each production information 301 as the target comparison; and

[0060] - a validation module, not illustrated, configured to select a model from the plurality of possible models, already known, that could theoretically be suitable for the processing production information 301 ; and

[0061] - a test module, not illustrated, configured to apply a second group of normalised information to the model selected by the validation module and to compare, for each normalised information, the energy production value obtained from that model selected by the validation model with the actual energy production value acquired and contained in the corresponding production information 301.

[0062] It should be noted that the target output data of the selected model can be expressed as an absolute value (and therefore directly indicate the energy production value) or as a percentage value (and therefore indicate in a derived manner the energy production value, if the percentage value indicates an efficiency coefficient) with respect to a nominal energy production of the photovoltaic panel 201 under optimal atmospheric conditions.

[0063] The training module is configured to apply an algorithm with regression curves, which is assessed by means of one or more of the following reliability modules, not illustrated, which are configured to assess the reliability of the model selected by the validation module in order to avoid Overfitting or Underfitting of the production information 301.A first reliability module can be of "rSquared (r2)" type. This module is configured to determine the production parameters that are least relevant to a variation in the value of energy production, considered as the target result. A second reliability module can be of the "Adjusted rSquared" type. This module is configured to evaluate how the production parameters, considered together, can explain a variation in the energy production value, considered as the target result.

[0064] Having several production parameters that may affect a selected model 502, but not other possible models, the evaluation by means of the Adjusted rSquared reliability module determines which production parameter is useful to the selected model 502 for the purpose of a predetermined prediction.

[0065] A third reliability module can be of “RMSE (Root-Mean-Square Error in machine learning) type. This reliability module is configured to evaluate whether model 502, selected by the validation module, is adequate with respect to the production information 301 used.

[0066] The RMSE reliability module measures the difference between the energy production values calculated by the model and the corresponding measured values. RMSE identifies the largest differences in the first group of normalised production information, used by the training module, and is able to evaluate the quality of said first group of production information as a whole.

[0067] The training module, the validation module, the test module and the reliability module of the processing component 501 are configured to be put into execution several times, after the execution of the normalisation module, if present, precisely in order to identify, iteratively and among the various models theoretically available to the validation module, the most reliable model that can be selected as model 502 created for the photovoltaic plant 2 by the processing component 501.

[0068] In this way, in an operational production step subsequent to the learning step, the processing component 501 is configured to receive from theenvironmental meteorological unit 402 the current environmental parameters and to use the model 502, created in the learning step, in order to obtain an optimal inclination configuration to be provided to the position controller 403 of the photovoltaic panel 201 so as to maximise the current energy production based on the current environmental parameters.

[0069] The expression “current environmental parameters” means the same environmental parameters described above and used in the learning step to create by means of machine learning the model 502, but which are acquired in the operational production step.

[0070] Thanks to the model 502, the control system 1 can be configured to control the photovoltaic panel 201 so as to arrange it in an optimal configuration in order to optimise production.

[0071] The control unit 4 may optionally comprise a remote access device 404 configured to access, for each energy data, respective weather maps relating to the geographical area in which the photovoltaic plant 2 is located. In this case, in fact, the production parameters may advantageously comprise the weather maps relating to the geographical area in which the photovoltaic plant 2 is positioned and, therefore, the processing component 501 is configured to process the weather maps.

[0072] The weather maps can, for example, be provided by government agencies and can contain both the same environmental parameters that can be acquired by means of the environmental meteorological unit 402 but also further environmental parameters, more closely related to the physics of the Earth’s atmosphere, such as, for example, atmospheric pressure, temperature, cloud cover and rainfall.

[0073] The processing component 501 can be configured to extract from these weather maps the environmental parameters contained therein and compare these environmental parameters, which are "official" (since they are officially provided by government agencies), with those acquired in the vicinity of the photovoltaic plant 2 in the same time interval for the same geographical area. In fact, the “official” environmental parameters could bedifferent from those effectively acquired in the vicinity of the photovoltaic plant.

[0074] This enables the reliability efficiency of the model 502 of the photovoltaic plant 2 to be increased by enabling an assessment of whether and by how much the official environmental data deviate from the actual environmental data, assuming all other conditions are equal.

[0075] The remote access device 404 may optionally be configured to receive “weather alerts” from the appropriate government agencies. The expression “weather alert” means an automatic warning system managed by a government agency (for example the National Civil Protection service) that signals the risk of intense weather phenomena (storms, strong wind, snow, etc.) to warn of possible dangers to people and objects, when there are variations in local weather forecasts from an ordinary criticality to a high criticality. The remote access device 404 can be configured to access the weather maps in the vicinity of the photovoltaic plant more frequently in the presence of a "weather alert", so that the processing component 501 can extract the "official" environmental parameters from them and compare them with those acquired in the vicinity of the photovoltaic plant 2 in order to determine the start of any critical conditions and move accordingly, as we will see, the photovoltaic panel 201.

[0076] The processing component 501 may also be configured to extract further environmental parameters related to the physics of the Earth’s atmosphere from the weather maps.

[0077] The processing component 501 can be configured, during the production step, to receive from the remote access device 404 the current weather maps and to use the model 502 so as to calculate a theoretical energy production for the same photovoltaic plant 2, in a different time interval, or calculate a theoretical energy production for a different photovoltaic plant 2, located in a different geographical position during that different time interval, based on knowledge of the historical weather maps for the specified different time interval and / or geographical position.It should be noted that the historical weather maps must be of good quality, with sufficient depth of the data, that is to say, be available at hourly / daily / monthly / yearly intervals, and report at least the same environmental parameters obtainable by means of environmental meteorological unit 402.

[0078] If the weather maps of interest are accessed by means of the remote access device, the model 502 can calculate a theoretical energy production forecast, using, for example, the historical and forecast weather maps of that geographical area where the photovoltaic plant 2 is located, in order to forecast the production of that same plant 2.

[0079] Similarly, it is possible to carry out a forecast of theoretical energy production for a different photovoltaic plant in a different geographical area, knowing the historical and forecast weather maps in the different geographical area.

[0080] An example application of the model 502 is the ability to predict a daily energy production value, taking into consideration a theoretical photovoltaic plant installed in a specific geographical area, identifiable by its latitude and / or longitude, with a predetermined number of photovoltaic panels of a certain size, assuming predetermined environmental parameters, which can be derived from the relative historical and forecast weather maps.

[0081] It should be noted how the creation of the model 502 in the learning step with machine learning allows both an improvement in the current production during the operational step and a reliable forecast of energy production in the same or different geographical areas.

[0082] The control unit 4 comprises an acquisition component 405 that is configured to acquire, both in the learning step and in the operational production step, in a massive manner and at predetermined time intervals, the production information 301 comprising the plurality of energy production data of the photovoltaic plant 2, from the inverter controller 401 , and the relative production parameters, from the environmental meteorological unit 402 and from the position controller 403.If the remote access device 404 is present, the acquisition component 405 is also configured to access, for all the energy data, the respective weather maps as described above and also any “weather alerts”.

[0083] The control system 1 may also optionally comprise, alternatively, or in addition to the remote access device 404, at least one image acquisition device 210, positioned in the vicinity of the photovoltaic plant 2, and configured to acquire images of the photovoltaic panel 201 and / or of the sky, in order to obtain any visible weather events of the earth's atmosphere or of the environment surrounding the photovoltaic plant 2 itself, such as, for example, cloud cover and / or precipitation (such as rain, hail, snow) and / or strong wind, which occur in real time, referred to herein as weather events in real time.

[0084] It should be noted that, although in the following description we will refer generally for the sake of simplicity to an image acquisition device 210, the latter can be configured for acquiring single images or a stream of images to define a video. In other words, the image acquisition device 210 may be a camera or a video camera for acquiring images and / or videos.

[0085] The production parameters may advantageously also comprise such weather events in real time, which are therefore related to the specific geographical area in which the photovoltaic plant 2 is positioned and may be stored in the storage platform 3 and acquired by the acquisition component 405, in a manner similar to the other production parameters acquired by the environmental meteorological unit 402, by the position controller 403 and optionally by the remote access device 404.

[0086] The image acquisition device 210 may be mounted on a support 211 , which may be a pole, shown in Figure 1 , spaced from the photovoltaic panel 201 , or alternatively it may be an arm (not illustrated) connected or connectable to the supporting structure 205 of the photovoltaic panel 201.

[0087] The image acquisition device 210 may be oriented and inclined with respect to the support 211 , for example around a hinge, in order to be able to frame alternatively the photovoltaic panel 201, and / or the sky, and / or theenvironment surrounding the photovoltaic panel 201 , and thus acquire images of the photovoltaic panel 201 and / or the sky and / or the surrounding environment.

[0088] Alternatively, if the image acquisition device 210 can neither be inclined or oriented, a pair of image acquisition devices 210 may be present, wherein one is configured to frame the sky and the other is configured to frame the photovoltaic panel 201 and / or the surrounding environment of the photovoltaic panel 201 itself.

[0089] The image acquisition device 210 may also have functions which can be set up remotely, for example it may have an adjustable magnification factor, to magnify or shrink the framed scene, or it may be adjustable to execute a panoramic view (pan) from left to right, or vice versa, or from bottom to top, or vice versa, without changing the magnification.

[0090] The images acquired framing the photovoltaic panel 201 , suitably processed, may indicate particular operating conditions of the photovoltaic panel 201 , for example malfunctions of photovoltaic plant 2.

[0091] From the same images, or from the images acquired that also frame the sky and / or the surrounding environment, suitably processed, it is possible to obtain said weather events in real time.

[0092] For this purpose, the control unit 4 may comprise an image acquisition manager 410, connected to the image acquisition device 210.

[0093] The image acquisition manager 410 is configured to set up the orientation of the acquisition device 210 (if the image acquisition device can be inclined and oriented) and the type of magnification and / or the function required. In other words, the image acquisition manager 410 is configured to set up the configuration of the image acquisition device 210, for example the type of image and what the image acquisition device 210 should frame. The set up of the configuration is not performed for each acquired image, but may be necessary during an initial configuration step of the image acquisition device 210 itself, for particular weather events, or at predetermined time intervals.The image acquisition manager 410 is also configured for acquiring images from the image acquisition device 210 and for processing said images to obtain the weather events in real time.

[0094] It is understood that the image acquisition manager 410 may also be configured for acquiring the video stream of images in succession from the image acquisition device 210 and for processing said video stream to obtain the weather events in real time.

[0095] The assembly comprising the image acquisition device 210 and the image acquisition manager 410 for the images themselves defines a unit for the detection of weather events in real time in the geographical area where photovoltaic plant 2 is located.

[0096] The images (and / or the videos) of the photovoltaic panel 201 and / or the sky may be acquired by the image acquisition manager 410 at predetermined time intervals and processed substantially in real time, or may be temporarily stored and processed at time intervals greater than those of the acquisition performed by the acquisition component 405, to obtain the weather events in real time.

[0097] For example, the image acquisition manager 410 may be configured to acquire from the acquisition device 210, at predetermined time intervals during daylight hours, images of the sky and determine the presence of weather events, such as cloud cover, any fog, or haze, wind or heavy rain, hail, heavy snow, detecting them directly in real time with visual inspection, by processing the acquired images.

[0098] However, advantageously, the image acquisition manager 410 may be configured to perform a comparison between a plurality of images of the sky acquired with the same angle and magnification, but in a time sequence, in order to determine a speed of movement of the clouds and the type of clouds, to determine a time of arrival of the cloud cover above the photovoltaic plant 2 and the possibility that such cloud cover may bring rainfall and / or snow and / or hail.Having determined the presence of cloud cover in the vicinity of the geographical area where photovoltaic plant 2 is positioned, the image acquisition manager 410 may, for example, further decide to check whether only clouds are present, or whether there is precipitation such as rain, snow, or hail.

[0099] In the case of precipitation, the image acquisition manager 410 can also determine which photovoltaic panel 201 , among all the photovoltaic panels of the plant 2, is most exposed to it, by specifically framing that photovoltaic panel 201.

[0100] The image acquisition device 210 may be unique in the photovoltaic plant 2 and may be arranged to frame a single photovoltaic panel 201 , if all the photovoltaic panels 201 of the photovoltaic plant 2 are subject to the same environmental and meteorological conditions; or an image acquisition device 210 may be positioned, and suitably configured, to frame a plurality of photovoltaic panels 201. However, there may be several image acquisition devices 210, each associated with a respective group of photovoltaic panels 201 , wherein the number of photovoltaic panels 201 in each group may also change, if, due to the particular shape of the terrain and geographical area, the operating conditions of the photovoltaic panels 201 , or the cloud cover, might vary for each group of photovoltaic panels 201.

[0101] By way of example, it should be noted that the photovoltaic plant 2 has been shown in Figure 6 as comprising a single image acquisition device 210 while, in the variant of the photovoltaic plant 2 shown in Figure 7, the photovoltaic plant 2 has been shown as comprising a plurality of groups of four photovoltaic panels 201 , each group being associated with a respective image acquisition device 210.

[0102] For this reason, the acquisition component 405 is also configured to also receive the weather events in real time, for each energy data, if the image acquisition device 210 and the image acquisition manager 410 of control unit 4 are present, since the weather events in real time, as describedabove, can form part of the production parameters alongside the environmental parameters previously described and can be associated with the production value provided by each photovoltaic panel 201.

[0103] It follows that if the production parameters also include the weather events in real time, the processing component 501 can be configured, during the learning step, to create by machine learning a model 502 of the photovoltaic plant 2, during the learning step, which is able to process the production information 301 stored in the storage platform 3 (the energy production data of the photovoltaic panel 201 and the respective production parameters, which include the angular configuration of the photovoltaic panel 201 and the environmental parameters acquired in the vicinity of the photovoltaic plant 2 and the real time weather events) which can identify the set of production parameters that are correlated with each other and correlated with their respective energy production data, which over time have influenced the energy production values of the photovoltaic plant much more precisely, as the actual weather conditions of the sky in the geographical area of the photovoltaic plant are known, without these having been obtained from weather maps acquired for the same area and for the same time period.

[0104] However, the processing component 501 may also be configured to compare the weather events in real time with said weather maps, if any. The real time weather event detection unit, comprising the image acquisition device 210 and the image acquisition manager 410, may therefore allow accurate validation of any weather maps related to cloud cover which can be obtained by the remote access device 404, if the remote access device 404 is present.

[0105] All the more so, this can be particularly advantageous in the event of a “weather alert” as weather events in real time can provide very precise indications and confirm whether an intense weather phenomenon covered by the weather alert is really underway, or when the intense weatherphenomenon is expected, if the latter has not yet begun but is expected soon.

[0106] However, the presence of the image acquisition device 210 and the image acquisition manager 410 allows one to disregard the presence of the remote access device 404 and the weather maps provided by the government agencies, and guarantees the ability to accurately correlate the actual production of the photovoltaic plant 2 with the cloud cover and / or any fog or haze, etc. detected in real time, which have a significant impact on the energy production of the photovoltaic panel 201.

[0107] The control unit 4 comprises, in addition, a storage component 406 which is configured for storing in the storage platform 3 the production information 301 , acquired by the acquisition component 405, both in the acquisition step and in the operational production step.

[0108] In the operational production step, the processing component 501 is configured to process the current production information 301 , thus acquired by the acquisition component 405 and stored by the storage component 406, to update the model 502 of the photovoltaic plant 2 by machine learning, by analysing the current production information 301.

[0109] The term "current production information" means the same production information described above, but which in this case is acquired during the operational production step.

[0110] In detail, the processing component 501 is configured to process the current production information 301 with the same modules previously described (training module, validation module, test module and reliability module) to update the model 502 created previously. It should be noted that the current production information 301 can be preliminarily processed by means of the normalisation module.

[0111] Thanks to the acquisition component 405 and the storage component 406, which continuously make the production information 301 available in the storage platform 3 at predetermined intervals and thanks to the continuous updating also during the operational production step of the model 502,based on the current production information 301 by the processing component 5, it is possible to further improve the model 502 in order to optimise the photovoltaic plant 2, and / or make the production forecasts increasingly reliable.

[0112] The increasing availability of the production information 301 in the storage platform 3, as the photovoltaic plant 2 is used over time, allows the processing component 501 to have a large number of possible combinations between angular configuration and environmental parameters, to which the weather maps and / or weather events in real time can optionally be added.

[0113] Regarding the energy production data, the inverter controller 401 is configured to acquire from the field inverter 207 the energy production value, the relative consumption necessary for the movement of the photovoltaic panel 201 and the efficiency of the photovoltaic panel 201 itself. In fact, although it is the energy production value that defines the photovoltaic panel 201 and is directly obtained from the field inverter 207, the optimisation of the energy efficiency of the photovoltaic panel 201 , which can be obtained by means of the consumption necessary for the movement of photovoltaic panel 201 and the efficiency of the photovoltaic panel 201 , is important for an accurate energy balance.

[0114] As mentioned above, the position controller 403 is configured to identify the angular configuration of the photovoltaic panel 201. Such angular configuration comprises, for each time interval, the theoretical angle of inclination of the photovoltaic panel 201 based on the angle of elevation of the sun, as described above, which changes progressively during the day at every interval of time.

[0115] The theoretical angle of inclination can be calculated by the position controller 403 for each time interval, having direct access to the respective angle of elevation of the sun stored in the form of tabular data. Optionally, the position controller may be provided with a GPS antenna through which an astronomical time and the relative position of the sun can be obtained,and thus the theoretical angle of inclination can be calculated on the basis of an angle of elevation of the sun derived from the GPS antenna.

[0116] In addition, the angular configuration may comprise an optimal angle of inclination of the photovoltaic panel 201 , which corresponds to the angle of inclination which is capable of delivering the maximum energy production value in the same time interval. The position controller 403 is connected to the movement device to provide the movement device with the optimal angle of inclination.

[0117] In detail, the position controller 403 is connected to the field control device 206” and is configured for transmitting to the latter the optimal angle of inclination. In turn, the field control device 206” is configured to apply the optimal angle of inclination to the actuator 206’ installed on the support frame 203, in order to carry out the movement of the photovoltaic panel 201 , and to subsequently check, by means of the inclination sensor, that the optimal angle of inclination is effectively reached.

[0118] In order to enable the position controller 403 to identify the optimal angle of inclination, the control unit 4 may optionally comprise a solar radiation detector 407, shown schematically in Figure 3, which can be rotated about an axis of rotation Y by means of a relative motor, not shown, and is connected to the position controller 403.

[0119] The position controller 403 is configured to position the solar radiation detector 407 in a plurality of predetermined angular positions, calculated from the angle of elevation of the sun, and to verify for each of these predetermined angular positions the respective energy production value of the solar radiation detector 407. For this purpose, solar radiation detector 407 is associated with a respective field inverter (not illustrated), which is capable of converting the direct current generated by the solar radiation detector 407 and providing the respective energy production value over the specified time interval.

[0120] The position controller 403 is configured to identify as the optimal angle of inclination the angular position, among the predetermined angular positions,which is capable of maximising the energy production value of the solar radiation detector 407 while minimising the movement consumption of the photovoltaic panel 201 , and it is also configured to provide the optimal angle of inclination detected by the solar radiation detector 407 to the field control device 206 of the photovoltaic panel 201.

[0121] For example, on a cloudy day, where the solar radiation is diffused, the optimal angle of inclination identified by the solar radiation detector 407 may be an angle other than the theoretical angle based on the angle of elevation of the sun. For example, it might be advantageous to arrange the photovoltaic panel 201 horizontally to receive the maximum quantity of solar radiation.

[0122] The activation of the solar radiation detector 407 in the plurality of predetermined angular positions to identify the optimal angle of inclination may be carried out each time the photovoltaic panel 201 must be rotated by means of the actuation device, to follow the angle of elevation of the sun, or it may be activated only a few times during daytime sun exposure, at predetermined intervals.

[0123] However, the position controller 403 is able to select a current angle of inclination for the photovoltaic panel 201 between the theoretical angle of inclination (based on the angle of elevation of the sun) and the optimal angle of elevation, calculated by moving the solar radiation detector 407, in relation to the energy consumption that a movement in the optimal angle of inclination would result in for the photovoltaic panel 201 if such a movement were carried out.

[0124] In the event that such an energy consumption is excessive in relation to the production advantage, the position controller 403 is configured to confirm the theoretical angle of inclination as the current angle of inclination to the field control device 206” without moving the photovoltaic panel 201 to the optimal angle of inclination.

[0125] In addition, the angular configuration may also comprise the angle of inclination, measured by the inclination sensor installed on the supportframe 203, which is the angle effectively reached by the photovoltaic panel 201 whenever the photovoltaic panel 201 completes a movement commanded by the actuator 206’ of the field control device 206”.

[0126] It should be noted that there may be only one solar radiation detector 407 in photovoltaic plant 2, and may serve to control the optimal angle of inclination of all the photovoltaic panels 201. However, it may be necessary, if the photovoltaic plant 2 extends over a large geographical area, to have a plurality of solar radiation detectors 407, each of which is associated or associable with a group of photovoltaic panels 201 , which are exposed to equal solar conditions.

[0127] The solar radiation detector 407 is, however, optional and, if the solar radiation detector 407 is absent, the position controller 403 is configured to confirm the theoretical angle of inclination as the current angle of inclination for the field control device 206”. This optimal angle of inclination could, however, be different from the theoretical angle of inclination if the model 502 is able to calculate, by means of interpolation between two theoretical angles of inclination, considering the respective environmental parameters stored over time, a more advantageous angle of inclination for the purpose of maximising the energy production that differs from the theoretical angle of inclination.

[0128] For this reason, in the operational production step, the processing component 501 is configured to receive from the position controller 403 the current angular configuration and to use the model 502, created in the learning step, in order to obtain an optimal inclination configuration to be provided to the position controller 403 of the photovoltaic panel 201 so as to maximise the current energy production on the basis of the current angular configuration and / or the current environmental parameters and / or the weather maps (if present) and / or the real-time weather events (if the image acquisition device 210 and the image acquisition manager are present).Ultimately, in the operational production step, the model 502 created in the learning step is used, configuring the model 502 with current production data (current angular configuration, current environmental parameters and / or weather maps and / or weather events in real time) to maximise energy production.

[0129] For acquiring the environmental parameters associated with each production data, the environmental meteorological unit 402 comprises a plurality of detection sensors, not illustrated, which are capable of acquiring the environmental conditions in which the meteorological plant is operating. The environmental meteorological unit 402 comprises:

[0130] - a thermometer to acquire the air temperature from the surrounding environment, expressed in °C, which is configured to also acquire daily a minimum temperature Tmin and a maximum temperature Tmax;

[0131] - a solarimeter for acquiring the radiation power of the solar radiation, and optionally also the angle of elevation of the sun if the GPS antenna is absent, or the predefined tabular data for the geographical area are not available.

[0132] The environmental meteorological unit 402 may additionally comprise:

[0133] - a hygrometer for acquiring the humidity of the surrounding air, expressed as a percentage;

[0134] - a dew point sensor, expressed in °C;

[0135] - a barometer for acquiring the atmospheric pressure, expressed in hPa;

[0136] - an atmospheric rainfall sensor comprising a rainfall meter, for acquiring the amount of rainfall, expressed as a daily quantity and as an intensity in mm / h, and / or a snow gauge for acquiring the quantity of snow, expressed as a daily quantity and as an accumulation in mm / h;

[0137] - an anemometer for acquiring the wind speed and / or pressure and deriving an average wind speed from them in km / h;- a directional vane for acquiring the wind direction;

[0138] - a luminance sensor for detecting haze or fog;

[0139] - a spectrometer for detecting UV radiation.

[0140] It should be noted that not all the detection sensors listed above are necessary, for the purposes of the present invention, as the thermometer and the solarimeter are sufficient. However, the greater the number of detection sensors present, the greater the number of environmental parameters acquired and therefore the possible variables considered by the processing component 501 when creating the model 502.

[0141] In particular, the model 502 is particularly efficient if the environmental meteorological unit 402 contains at least a thermometer, an anemometer, the wind vane and the solarimeter, for acquiring, respectively, the air temperature, the wind speed and / or pressure, the wind direction, the irradiation power and, optionally, also the angle of elevation of the sun, to obtain the respective environmental parameters.

[0142] Also in this case, as with the solar radiation detector 407, there may be just one environmental meteorological unit 402 in the photovoltaic plant 2 if all the photovoltaic panels 201 have the same environmental parameters, or there may be several environmental meteorological units 402, each associated with a respective group of photovoltaic panels 201 if, due to the particular shape of the terrain and the geographical area, the environmental parameters could vary for each group of photovoltaic panels 201.

[0143] As mentioned above, thanks to the model 502, the control system 1 is configured to control the photovoltaic panel 201 so as to position it into an optimal configuration in order to optimise the production.

[0144] The optimisation of the production is also achieved because, thanks to the model, it is possible to carry out additional movements of the photovoltaic panel 201 in certain atmospheric conditions.

[0145] In fact, the control system 1 is configured to position the photovoltaic panel 201 in predetermined positions, under certain atmospheric conditions detected by the environmental meteorological unit 402 and / or obtained fromthe weather events in real time, in order to preserve photovoltaic panel 201 and / or to carry out maintenance on it and / or to perform particular cycles. For example, the control system 1 may be configured to horizontally position the photovoltaic panel 201 if the current wind speed is greater than or equal to a threshold value. In this way, wind damage can be limited if the winds are very strong.

[0146] The threshold value can be set a priori, but it can be determined much more advantageously thanks to the model 502.

[0147] The control system 1 may also be configured to position the photovoltaic panel 201 in a first end inclination (31 , for example equal to +55°, or in a second end inclination [32, for example equal to -55°; or to move the photovoltaic panel 201 between the first end inclination [31 and the second end inclination [32, or vice versa, if the current amount of rain indicates heavy rainfall, or if the current amount of snow indicates heavy snowfall, so as to prevent excessive accumulation of rain and / or snow on the photovoltaic panel 201 ; or so as to perform a cleaning cycle through the rain and / or snow when the rain and / or snow slides on the photovoltaic panel 201.

[0148] A threshold value to define heavy rain, or heavy snowfall can be set in advance, but much more advantageously it can be determined thanks to the model 502.

[0149] For this purpose, the control system 1 can be configured to stop the photovoltaic panel 201 in a recovery position, for example having an angle of inclination equal to 0°, for a predetermined time, when the photovoltaic panel is moved between the first end inclination [31 and the second end inclination [32 in the case of rain and / or snow, and vice versa, to allow a controlled accumulation of rain and / or snow to subsequently perform the cleaning cycle. In fact, since the rain and / or snow are free of limescale, they can be used to perform an efficient cleaning of a photovoltaic plant 2.

[0150] Also in this case, the predetermined time can be established a priori, but it can be determined far more advantageously using the model 502.Thanks to the cleaning cycle, it is possible to take advantage, from the first day of useful sunshine, after the rain or snow event, of a cleaner capture surface 204 of the photovoltaic panel 201 , without necessarily waiting for periodic manual cleaning treatments. This consequently allows a reduction in the number of periodic cleaning treatments, lengthening the intervals between interventions and thereby ensuring economic savings.

[0151] In addition, the control system 1 may be configured to move the photovoltaic panel 201 to a cooling angle (not shown) so as to arrange a capture surface 204 of the photovoltaic panel 201 in the shade for a given time interval, when the detected temperature is higher than, or equal to, a predetermined maximum temperature, in order to cool the photovoltaic panel.

[0152] In addition, the control system 1 may be configured to move the photovoltaic panel 201 also on the basis of weather events in real time determined by the image acquisition manager 410, if the latter and the related image acquisition device 210 are present.

[0153] For example, if the image acquisition manager 410 detects the impending arrival of a cloud front, despite the temperature being higher than or equal to the predetermined maximum temperature and therefore it would be appropriate to position the photovoltaic panel 1 in the shade in the cooling angle, the control system 1 can be configured to postpone the cooling movement and check the temperature again after a predetermined waiting time, if the cloud cover continues to be present or when the sky becomes clear.

[0154] In addition, the control system 1 may be configured to position the photovoltaic panel 201 vertically, in the case of a defective panel.

[0155] For example, if the image acquisition manager detects in real time an event such as hail (and even more advantageously in the presence of a hail “weather alert”), the control system 1 may be configured to calculate in real time a hail fall angle, for example, which can be detected by the image acquisition manager 410, and to move photovoltaic panel 201 to a protective angle (not illustrated), for example opposite the hail fall angle.Thanks to the movement to the protection angle, the control system 1 is able to preserve as much as possible the photovoltaic modules 202 of the photovoltaic panel 201 from the risk of breakage, caused by hailstones, or in any case to limit their damage.

[0156] It should be noted that the storage platform 3 may be implemented by means of a storage server comprising a number of replicated discs, for example magnetic and / or solid-state discs with RAID redundancy technology. Among the discs on which the storage platform 3 resides, one can for example be removable "hot", that is, without interrupting an operation of the remaining discs and also without the storage component 406 interrupting its own operation.

[0157] Alternatively, the storage platform 3 could be implemented as a cloud platform accessible remotely.

[0158] The inverter controller 401 may be connected by means of direct wiring to the field inverter 207 of the photovoltaic panel 201 and may be connected via a communication network 408, for example, Ethernet or Wi-Fi, to the acquisition component 405 and / or the storage component 406 of the control unit 4.

[0159] Usually, the field inverter 207 is positioned in the vicinity of the photovoltaic panels 201 , as shown in Figure 6. It should be noted, however, that it can also be positioned near the inverter controller 401.

[0160] The environmental meteorological unit 402 may also be connected via the connection network 408 to the acquisition component 405 and / or the storage component 406.

[0161] The position controller 403 may be connected to the movement device of the photovoltaic panel 201 , and more precisely to the field control device 206” by means of a radio link 409 and may be connected by means of the communication network 408 to the acquisition component 405 and / or to the storage component 406.The position controller 403 may be connected to the solar radiation detector 407 by means of the same radio link 409, or by means of the communication network 408.

[0162] The remote access device 404 may be configured with a mobile network Internet connection to access the weather maps and may be connected via the communication network 408 to the acquisition component 405 and / or to the storage component 406.

[0163] The image acquisition manager 410 may be connected by direct wiring to the image acquisition device 210 of the photovoltaic panel 201 , in a similar manner to the direct wiring between the inverter controller 401 and the field inverter 207. The image acquisition manager 410 may also be connected via the connection network 408 to the acquisition component 405 and / or to the storage component 406.

[0164] It should be noted that the remote access device 404 might not be included in the control unit 4 but may be included in the management server 5, for accessing, for each energy data, the respective weather maps and for storing the weather maps in association with the energy data and the environmental parameters acquired by acquisition component 405. However, the fact that control unit 4 also includes the remote access device 404 may allow the position controller 403 to receive short-term forecasts of particular weather conditions so as to adjust the angle of inclination of the photovoltaic panel 201 before those weather conditions occur.

[0165] If, conversely, the image acquisition manager 410 and the image acquisition device 210 are present, the remote access device 404 may be absent, as the image acquisition manager 410 itself can provide those short-term weather forecasts, as described above, which are useful for adjusting the angle inclination of the photovoltaic panel 201 before the predicted weather events occur.

[0166] The remote access device 404 can also be configured, in a similar way, to access daily evapotranspiration data, expressed in mm, relating to the geographical area in which the photovoltaic plant 2 is positioned. In thiscase, the environmental parameters may also include the daily evapotranspiration.

[0167] It should be noted that the storage platform 3, the control unit 4, the management server 5, the acquisition component 405 and the storage component 406 of the control unit 4, the processing component 501 and the model 502 of the management server 5, and / or the cloud, may also be connected to each other via a communication network 408, for example an Ethernet network.

[0168] It should be noted that in this patent application the following terms, such as “server”, “plant”, “component”, “module”, “platform”, “device” and “manager”, refer to one or more entities connected to, or form part of, a computing apparatus with one or more specific functions, in which these entities can be hardware, a combination of hardware and software, or exclusively software. For example, a component, module, or manager may be, without limitation, a process running on a processor; a hard disc unit; multiple storage units (optical or magnetic storage supports), including a solid state storage unit; an object; an executable; a computer executable program; and / or a computer.

[0169] By way of illustration, both an application running on a computer server and the computer server itself can be considered a component, or a module, or a manager. One or more components and / or modules and / or managers may reside within an execution process and a component and / or a module and / or a manager may be located on a computer and / or distributed between two or more computers. Further, the components and / or modules and / or managers as described herein may be executed from various computer-readable storage media having various data structures stored thereon. It should be noted that, in this patent application, the term ”in connection” is intended to mean that a “device”, a “manager”, a "server”, a “component”, a “module” and a “platform” can interact with each other through local and / or remote processes using the communication network 408. The communication network 408, as specified above, can be a local Ethernetnetwork or, more generally, the Internet, without restricting the scope of the invention.

[0170] The management server 5 may comprise a user interface, not illustrated, to allow an external user to perform the prediction linked to the energy production value of the photovoltaic plant 2, once the model 502 is configured with the dimensional characteristics of the photovoltaic plant 2 and the geographical area in which the plant 2 is installed.

[0171] In use, the control system 1 of the present invention is configured to perform a control method in accordance with the present invention.

[0172] The control method of the photovoltaic plant 2, which comprises at least one photovoltaic panel 201 , includes a learning step comprising:

[0173] - processing production information 301 acquired in a massive manner from the photovoltaic panel 201 , the production information 301 comprising a plurality of energy production data, associated with the energy production value and, for each energy production data, the respective production parameters that determined the energy production data. The production information 301 has already been defined above and, for simplicity, their definition is not repeated here.

[0174] The control method, in the learning step, also includes,

[0175] - creating, using machine learning, a model 502 of the photovoltaic plant 2 suitable for analysing the production information in order to identify a set of production parameters correlated with each other, and correlated with the respective energy production data, which have over time influenced the energy production values of the photovoltaic panel 201.

[0176] In order to create the model 502, the machine learning step includes further steps.

[0177] Initially, the control module may normalise the production data 301 (the energy production data of the photovoltaic panel 201 and the respective production parameters, including the angular configuration of the photovoltaic panel 201 and environmental parameters acquired in the vicinity of the photovoltaic plant 2), in order to filter out any spurious and / orincomplete and / or missing and / or incorrect production data 301 and to ensure all the production data 301 are consistent and normalised with each other before further processing. The control method may, for example, perform a normalisation step by means of the normalisation module described above.

[0178] Subsequently, the control method comprises the step of carrying out supervised training on a first group of normalised production information (equal to approximately 80% of the total) and using the energy production value associated with each normalised production information 301 as the target comparison.

[0179] In a subsequent validation step, the control method also comprises the step of selecting a model, amongst one of the plurality of possible and already known models which could be suitable for processing the normalised production information 301 , and carrying out a test step, following the validation step, to apply a second group of normalised information (equal to approximately 20% of the total) to the model selected during the validation step.

[0180] The control method may, for example, carry out the validation step by means of the validation module described above, and the test step by means of the test module described above.

[0181] The training step also provides, in the test step, for comparing, for each normalised production information, the energy production value obtained from the model selected in the validation step with a real energy production value acquired and contained in the corresponding normalised production information 301.

[0182] It should be noted that the control method may provide for the use of an absolute value (and therefore directly indicate the energy production value) or a percentage value (and therefore indicate the energy production value in a derived manner, if the percentage value indicates a coefficient of efficiency with respect to a nominal energy production value under optimal atmospheric conditions) as the output target.All the possible models are chosen from algorithms with regression curves, so the training module applies an algorithm like this type.

[0183] For this reason, after the execution of the test step with the model selected in the validation step, the evaluation of the reliability of the selected model itself is performed, applying one or more reliability evaluation methods chosen from a first evaluation method of the "rSquared (r2)" type, a second evaluation method of the "Adjusted rSquared" type, or a third evaluation method of the "RMSE (Root-Mean-Square Error in machine Learning)" type. The control method can, for example, carry out the reliability assessment step by means of a reliability module among those previously described. In fact, the reliability assessment methods have been mentioned previously and are not repeated here, for the sake of brevity.

[0184] Based on the reliability assessment, the model selected in the validation step can be modified and / or replaced if it does not have a high reliability. Consequently, the training step, the validation step, the test step and the evaluation step of the reliability of the selected model can be executed several times in succession, precisely in order to identify in an iterative manner, among the various models theoretically possible and selectable in the validation step, the most reliable model and consequently select this most reliable model as the model 502 created by machine learning for the photovoltaic plant 2.

[0185] The control method further comprises an operational production step, which is after the learning step, comprising the steps of

[0186] - maximising the energy production of the photovoltaic plant 2 over time, using the model 502 created in the learning step, configuring the model 502 with the current production data to obtain an optimal angular configuration of the photovoltaic panel 201.

[0187] Thanks to the model 502 of the photovoltaic plant 2, it is, in other words, possible to predict the optimal angular configuration of the photovoltaic panel 201 , in the environmental conditions detected (current environmental parameters of the production data, acquired during the operationalproduction step) so that photovoltaic panel 201 does not have an angle of inclination equal only to the theoretical angle, or for which, if present, it is necessary to activate the solar radiation detector 407 in the plurality of predetermined angular positions in order to identify the optimal angle of inclination. The optimal angular configuration is, in other words, obtained directly thanks to the model 502 created for the photovoltaic plant 2, and it is no longer necessary to identify the optimal angle of inclination.

[0188] Further, the model 502 also allows, in addition or alternatively, the calculation of a theoretical energy production of the same photovoltaic plant 2 over a different time interval, or the calculation of a theoretical energy production of a different photovoltaic plant, located in a different geographical position during the different time interval, provided that historical weather maps are processed for the specified different time interval and / or geographical location, in the case in which the production parameters comprise, in addition, the weather maps relating to the geographical area in which the photovoltaic plant is located.

[0189] For example, the model 502 can be used to carry out theoretical production forecasts, for example daily, for a different photovoltaic plant, or for the same photovoltaic plant 2 knowing the historical weather maps of the different geographical area and making assumptions about possible environmental parameters in the different geographical area.

[0190] If the production parameters also include the weather maps, the model 502 can also be configured with them to achieve the optimal angular configuration.

[0191] The control method comprises, in fact, the additional step of processing the weather maps (provided by government agencies) to extract from them both the same environmental parameters that can be acquired in the vicinity of the photovoltaic plant 2 and further environmental parameters, more strictly related to the physics of the Earth's atmosphere, as described above.

[0192] The control method can also be configured to compare the environmental parameters contained in the “official” weather maps with those acquired inthe vicinity of the photovoltaic plant 2 within the same time interval for the same geographical area, in order to increase the reliability efficiency of the model 502.

[0193] The control method may also include the step of receiving ‘weather alerts’ from the relevant government agencies and, in the case of a “weather alert”, of receiving and processing the weather maps more frequently in order to extract the ‘official’ environmental parameters from the weather maps themselves and compare them with those acquired in the vicinity of the photovoltaic plant 2 in order to determine the onset of any critical conditions and intense weather phenomena.

[0194] The control method can also extract the additional environmental parameters related to the physics of the Earth's atmosphere from the weather maps, so that the weather maps, as a whole, are considered among the production parameters, in addition to the environmental parameters.

[0195] Even more advantageously, however, if the production parameters also comprise, alternatively or in addition to the weather maps, the weather events in real time which can be obtained by the image acquisition device 210 and the image acquisition manager 410 described above, the model 502 can also be configured with them so as to obtain, in the operational production step, the optimal angular configuration of the photovoltaic panel 201 in order to maximise the energy production of photovoltaic plant 2 over time.

[0196] It should be noted that both in the operational production step and in the learning step, the control method may comprise the additional step of processing images (and / or videos, as described above) acquired of the photovoltaic panel 201 and / or of the sky to obtain real time weather events relating to the geographical area in which the photovoltaic plant 2 is positioned.The control method may also comprise the additional step of acquiring these images at predetermined time intervals, in order to obtain the weather events in real time.

[0197] Alternatively, or in addition, the control method may perform the step of comparing between them a plurality of images of the sky acquired with the same angle and magnification, but in a time sequence, in order to determine a speed of movement of the clouds and the type of clouds, in order to determine a time of arrival of a cloud cover above the photovoltaic plant 2 and the possibility that such cloud cover may bring precipitation.

[0198] Again, in the production step, following the learning step, the control method may further include the step of maximising the energy production of the photovoltaic plant 2 over time, using the model 502 created during the learning step, and configuring the model 502 with the environmental parameters and / or processing the weather events in real time, since the environmental parameters may optionally also include the weather events in real time, to achieve an optimal angular configuration of the photovoltaic panel.

[0199] The control method of the present invention also comprises the step of - acquiring in a massive manner the current production information 301 at predetermined time intervals to obtain a plurality of energy production data of the plant and the relative production parameters;

[0200] - processing the production information 301 acquired in a massive manner by the photovoltaic panel 201. These acquisition and processing steps are valid both during the learning step and during the production step.

[0201] However, in the production step, the method also includes the step of - processing the current production information 301 ,

[0202] - updating the model 502 of the photovoltaic plant 2 using machine learning, analysing the current production information 301.

[0203] For example, by processing the current production information 301 a correlation between the environmental parameters and / or the weather conditions (for example, forecast maps of the same geographical area inwhich the photovoltaic plant 2 is positioned) and / or the real-time weather events, and the optimal angle of inclination identified by the solar radiation detector 407 can be identified and this correlation can also be used to update the model 502 itself.

[0204] As mentioned above, in fact, the training step, the validation step, the test step and the evaluation step of the reliability of the selected model can be executed several times, even iteratively, after the learning step, when the storage platform 3 has stored new production information 301 such as the current production information 301 (which can also be normalised, if the control method comprises the normalisation step).

[0205] It should be noted that the model 502 relating to a certain geographical area will have specific characteristics but may differ from a further model of a further photovoltaic plant located in another geographical area.

[0206] The method of the present invention comprises the further step of creating by machine learning this further model by processing the respective production information acquired in a massive manner from the further photovoltaic plant and updating by machine learning the model of the photovoltaic plant by means of the further model of the further photovoltaic plant, precisely in order to combine as much as possible the environmental parameters with the weather maps and / or the energy production data. Again, the steps to create this new model are similar to those described above and are not repeated here.

[0207] The step of processing the energy production data comprises the step of analysing the energy produced, the relative consumption necessary for the movement of the photovoltaic panel and the efficiency of the photovoltaic panel, in order to have a complete energy picture that considers the energy yield of the photovoltaic panel and also its related consumption.

[0208] On the other hand, the step of processing the angular configuration may comprise the step of identifying the theoretical angle of inclination of the photovoltaic panel 201, based on the angle of elevation of the sun, and a corresponding optimal angle of inclination of the photovoltaic panel 201,which corresponds to the angle of inclination which is capable of delivering the maximum energy production value in the time interval.

[0209] In order to obtain the optimal angle of inclination, the control method may optionally comprise the steps of positioning the solar radiation detector 407 in the plurality of predetermined angular positions, verifying for each of said predetermined angular positions the respective energy production value of the solar radiation detector 407 and identifying as the optimal angle of inclination, among the plurality of predetermined angular positions, the angle of inclination which is capable of maximising the energy production value while limiting, that is, minimising, the respective consumption necessary for the movement of the photovoltaic panel 201.

[0210] However, the control method may provide for identifying the theoretical angle of inclination as the optimal angle of inclination, or may provide for calculating the optimal angle of inclination by interpolation between two theoretical angles of inclination, considering the respective environmental parameters stored over time, an angle of inclination which is more advantageous for the purpose of maximising the energy production, different from the theoretical angle of inclination.

[0211] The step of processing the environmental parameters comprises analysing one or more of the parameters acquired by the environmental meteorological unit 402 positioned in the vicinity of the photovoltaic plant 2, expressed at least as air temperature and solar radiation irradiation power. The step of processing the environmental parameters may further comprise analysing one or more of the following parameters acquired by the environmental meteorological unit 402, expressed as air humidity, dew point, atmospheric pressure, quantity of rain and / or snow, wind speed and / or pressure, wind direction, brightness, and UV radiation.

[0212] Thanks to the model 502, the control method can also move the photovoltaic panel 201 to position it in an optimal configuration to optimise production under certain atmospheric conditions detected by the environmental meteorological unit 402 and / or obtained from real time weather eventsrelating to the geographical area where the plant is located, so as to preserve photovoltaic panel 201 and / or carry out maintenance and / or perform specific cycles.

[0213] For example, if the current wind speed is greater than or equal to a threshold value, the control method may provide for horizontally positioning the photovoltaic panel 201 , so as to minimise the impact of the wind on the photovoltaic panel 201 itself if the wind is very strong.

[0214] The threshold value can be determined a priori but much more advantageously the control method can determine it thanks to the model 502.

[0215] If, on the other hand, the current quantity of rain is indicative of heavy rainfall, or if the current quantity of snow is indicative of heavy snowfall, the control method may comprise the step of positioning the photovoltaic panel in the first end inclination (31 , for example equal to +55°, or in the second end inclination [32, for example equal to -55°; alternatively, it may also be provided to move the photovoltaic panel between the first end inclination [31 and the second end inclination [32, and vice versa, so as not to accumulate too much rain and / or snow; or so as to carry out a cleaning cycle by means of the rain and / or the snow, when the rain and / or the snow slides on the photovoltaic panel 201.

[0216] For this purpose, the control system can also be configured to stop the photovoltaic panel 201 in a recovery position, for example having an angle of inclination equal to 0°, for a predetermined time, when the photovoltaic panel is moved between the first end inclination [31 and the second end inclination [32, and vice versa, to allow a controlled accumulation of rain and / or snow to subsequently perform the cleaning cycle. In fact, since the rain and / or snow are free of limescale, they can be used to perform an efficient cleaning of a photovoltaic plant 2.

[0217] In order to define heavy rain, or heavy snowfall, a threshold value can be determined a priori, but much more advantageously the control method can obtain it from the model 502.Thanks to the cleaning cycle, it is possible to take advantage, from the first day of useful sunshine, after the rain or snow event, of a cleaner capture surface 204 of the photovoltaic panel 201 , without necessarily waiting for periodic manual cleaning treatments. This consequently allows a reduction in the number of periodic cleaning treatments, lengthening the intervals between interventions and thereby ensuring economic savings.

[0218] If an air temperature that is greater than or equal to a predetermined maximum temperature is detected and there is no wind, the control method may provide for a cooling cycle of the photovoltaic panel 201 itself. The predetermined maximum temperature may, for example, correspond to, or in any case be related to, a maximum working temperature indicated in the technical specifications of the photovoltaic panel 201 , which is the temperature beyond which the photovoltaic panel 201 reduces the energy production value despite being positioned at the optimal angle of inclination. In other words, when the efficiency of photovoltaic panel 201 may decrease (as per the design specifications of the various manufacturers of photovoltaic panels) due to high environmental temperature and the absence of wind, the control method may move the photovoltaic panel 201 to a cooling angle, opposite the theoretical angle of inclination, so as to position the capture surface 204 of the photovoltaic panel 201 in the shade for a determined time interval, thus allowing it to cool. The time interval needed to cool the photovoltaic panel 201 may, however, also depend on other environmental factors such as wind speed and / or direction. In fact, if the presence of wind is already able to sufficiently cool the capture surface 204, the movement to the cooling angle, in the shade, of the photovoltaic panel 201 may not be necessary, or, it may be necessary but for a very short time interval compared to that necessary in the absence of wind. If the control method comprises the additional step of processing the acquired images of the photovoltaic panel 201 and / or of the sky to obtain the real time weather events relating to the geographical area in which the photovoltaic plant 2 is positioned, and in the event that the arrival soon of acloudy front is detected, the control method may decide to postpone the cooling movement and wait for a predetermined period of time. In fact, the cloud cover could itself be sufficient to lower the temperature of the photovoltaic panel 201, even in the presence of modest wind, without the movement in the cooling angle being necessary. The control method can check again the temperature of the photovoltaic panel 201 at the end of the waiting time or if in the meantime the sky has cleared.

[0219] In the case, on the other hand, of a defective photovoltaic panel 201 , however, the control method may provide for the step of vertically positioning the panel 201.

[0220] In the case that the control method detects in real time a hail event (and even more advantageously in the presence of a hail "weather alert"), the control method can advantageously calculate in real time a hail fall angle, processing the acquired images of the photovoltaic panel 201 and / or of the sky, and move the photovoltaic panel 201 to a protection angle, for example opposite to the hail fall angle.

[0221] Thanks to the movement to the protection angle, the control method can preserve as much as possible the photovoltaic modules 202 of the photovoltaic panel 201 from the risk of breakage, caused by hailstones, or in any case limit their damage.

[0222] It should be noted that the control system 1 may be realised by means of a computer program including instructions such that, when the program is executed in the control system 1 itself, the control system 1 executes one or more steps of the control method described above.

[0223] Therefore, thanks to the present invention, a simple control method is provided that delivers an immediate economic benefit which is able to maximise the value of energy production and, at the same time, enables increasingly accurate energy forecasts t be guaranteed.

Claims

CLAIMS1. A method for controlling a photovoltaic plant (2) positioned in a geographical area, wherein the photovoltaic plant (2) comprises at least one photovoltaic panel (201), which is configured to rotate about an axis of inclination (X) according to a respective angular configuration and to provide a predetermined energy production value; wherein the method comprises: a learning step comprising:- processing production information (301) acquired in a massive manner from the photovoltaic panel (201), wherein the production information (301) comprises a plurality of production energy data, associated with respective energy production values, which are acquired in a predetermined time interval, and wherein, for each production energy data, the production information (301) comprises in addition respective production parameters, which have determined the production energy data, the production parameters comprising the angular configuration and environmental parameters acquired in the vicinity of the photovoltaic plant;- creating, using machine learning, a model (502) of the photovoltaic plant (2) suitable for analysing the production information (301 ) in order to identify a set of production parameters correlated with each other, and correlated with the respective production energy data, which have overtime influenced the energy production values of the photovoltaic panel (201).

2. The control method according to claim 1 , wherein the production parameters comprise in addition weather events in real time relating to the geographical area in which the photovoltaic plant is positioned and wherein the control method also comprises an operational production step, after the learning step, which comprises the steps of- maximising the energy production of the plant over time, using the model (502) created during the learning step and configuring the model (502) with the current environmental parameters and / or processing the weather events in real time to obtain an optimal angular configuration of the photovoltaic panel (201).

3. The control method according to claim 2, and wherein in the learning step and in the operational production step the control method further comprises the steps of acquiring, at predetermined time intervals, images of the photovoltaic panels and / or the sky and processing these images to obtain weather events in real time relating to the geographical area in which the plant is located.

4. The control method according to any one of the preceding claims, wherein the production parameters comprise in addition the weather maps relating to the geographical area in which the photovoltaic plant is positioned and wherein the control method also comprises an operational production step, after the learning step, which comprises the steps of- maximising the energy production of the plant over time, using the model (502) created during the learning step and configuring the model (502) with the environmental parameters and / or processing the current weather maps to obtain an optimal angular configuration of the photovoltaic panel (201) and / or- calculating a theoretical energy production of the photovoltaic plant (2), in a different time interval, and / or calculating a theoretical energy production of a different plant, in a different geographical position in the different time interval, processing historical weather maps in the predetermined different time interval and / or in the different geographical position.

5. The control method according to any one of the preceding claims, wherein the control method further comprises the steps of- acquiring in a massive manner the current production information (301) at predetermined time intervals to obtain a plurality of energy production data of the plant and the relative production parameters;- processing the production information (301) acquired in a massive manner by the photovoltaic panel (201);and wherein in the operational production step the control method further comprises the step of- processing the current production information (301);- updating the model (502) of the photovoltaic plant (2) using machine learning, analysing the current production information (301).

6. The control method according to any one of the preceding claims, and comprising the further step of- creating by machine learning a further model of a further photovoltaic plant (2) positioned in a further geographical area, processing respective production information acquired in a massive manner from the further photovoltaic plant;- updating by machine learning the model (5) of the photovoltaic plant (2) using the further model of the further plant.

7. The control method according to any one of the preceding claims, wherein the step of processing the energy production data of the photovoltaic panel (201) comprises the step of analysing the energy production value, the relative energy consumption necessary for the movement of the photovoltaic panel and the efficiency of the photovoltaic panel (201).

8. The control method according to any one of the preceding claims, wherein the step of processing the angular configuration comprises identifying a theoretical angle of inclination of the photovoltaic panel (201), based on an angle of elevation of the sun.

9. The control method according to claim 8, when dependent on claim 7, wherein the step of processing the angular configuration further comprises the step of identifying, for each theoretical angle of inclination, a corresponding optimal angle of inclination of the photovoltaic panel (201), which corresponds to the angle of inclination which is able to deliver the maximum energy production value in the time interval; and wherein optionally, to obtain the optimal angle of inclination, the method comprises the steps of positioning a solar radiation detector (407) in a plurality of predetermined angular positions, verifying for each of these predetermined angular positions the respective energy production value of the solar radiation detector (407) and identifying as the optimal angle of inclination,between the plurality of predetermined angular positions, the angle of inclination which is able to maximise the energy production value whilst limiting the respective energy consumption necessary for moving the photovoltaic panel (201).

10. The control method according to any one of the preceding claims, wherein the step of processing the environmental parameters comprises analysing the following parameters acquired from an environmental meteorological unit (402) in the vicinity of the photovoltaic plant (2): air temperature and solar radiation power; and wherein optionally the step of processing the environmental parameters further comprises the step of analysing one or more of the following parameters acquired from the environmental meteorological unit (402): air humidity; dew point; atmospheric pressure; rain and / or snow quantity; wind speed and / or pressure; wind direction; environmental brightness; UV radiation.

11. The control method according to claim 10, and comprising the step of horizontally positioning the photovoltaic panel (201) if the current wind speed is greater than or equal to a threshold value.

12. The control method according to claim 10 or 11 , and comprising the step of positioning the photovoltaic panel (201) in a first end inclination ([31), for example equal to +55°, or in a second end inclination ([32), for example equal to -55°; or moving the photovoltaic panel (201) between the first end inclination ([31 ) and the second end inclination ([32), if the quantity of current rain is indicative of heavy rain, or if the amount of current snow is indicative of heavy snowfall.

13. The control method according to any one of claims 10 to 12, comprising the step of moving the photovoltaic panel (201) into a cooling angle so as to position a capture surface (204) of the photovoltaic panel (201 ) in a shade for a predetermined period of time, when a temperature greater than or equal to a predetermined maximum temperature is detected, to cool the photovoltaic panel.

14. The control method according to any one of claims 10 to 13, andcomprising the step of vertically positioning the photovoltaic panel (201), in the case of a defective panel.

15. The control method according to claim 2 or 3, or according to any one of claims 4 to 14, when dependent on claim 2 or 3, wherein, if the detected event in real time is hail, the method further comprises the step of calculating a hail fall angle and moving the photovoltaic panel to a protective angle, the protective angle being opposite to the hail fall angle.

16. The control method according to claim 4, or according to any one of claims 5 to 15, when dependent on claim 4, wherein the step of processing the weather maps relative to the geographical area in which the photovoltaic plant (2) is positioned comprises analysing the weather maps for extracting from these weather maps the environmental parameters contained in them and comparing the latter with the environmental parameters acquired in the vicinity of the photovoltaic plant (2) in the same period of time and for the same geographical area.

17. A system (1) for controlling a photovoltaic plant (2) positioned in a geographical area, wherein the photovoltaic plant (2) comprises at least one photovoltaic panel (201 ), which comprises a movement device (206’, 206”), configured to rotate the photovoltaic panel (201) about an axis of inclination (X) according to a respective angular configuration, and a field inverter (207), configured to provide a predetermined energy production value; wherein the control system (1) comprises:a storage platform (3) configured for storing production information (301) acquired in a massive manner from the photovoltaic panel (201) which comprises production energy data, associated with respective energy production values and acquired in a predetermined time interval, and wherein, for each production energy data, the production information (301) in addition comprises the respective production parameters, which have determined the production energy data; the production parameters comprising the angular configuration and environmental parameters acquired in the vicinity of the photovoltaic plant (2);a control unit (4) which comprises- an inverter controller (401), connected to the field inverter (207) of the photovoltaic panel (201) and configured for acquiring the plurality of production energy data of the photovoltaic panel (201 ), in the predetermined time interval;- an environmental meteorological unit (402), positioned in the vicinity of the photovoltaic plant (2) and configured for acquiring, for each production energy data, relative environmental parameters of the production parameters;- a position controller (403), connected to the movement device (206’; 206”) of the photovoltaic panel (201) and configured to determine the angular configuration of the photovoltaic panel (201) in the predetermined time interval;and wherein the control system comprises a management server (5) which comprises a processing component (501) which is configured, in a learning step, to process said production information (301) stored in the storage platform (3) and to create, by machine learning, a model (502) of the photovoltaic plant (2) suitable for analysing the production information (301) in order to identify a set of production parameters correlated with each other, and correlated with the respective energy production data, which have over time influenced the energy production values of the photovoltaic panel (201).

18. The control system according to claim 17, wherein the processing component (501) is configured to receive from the environmental meteorological unit (402) the environmental parameters; and wherein in an operational production step, subsequent to the learning step, the processing component (501) is configured:- to receive the current environmental parameters from the environmental meteorological unit (402) and- for using the model (502) created in the learning step in order to obtain an optimal inclination configuration to be supplied to the position controller(403) of the photovoltaic panel (201) in such a way as to maximise the current energy production on the basis of the current environmental parameters.

19. The control system (1) according to claim 17 or 18, wherein the photovoltaic plant (2) comprises at least one image acquisition device (210) for acquiring images of the photovoltaic panel (201) and / or the sky and wherein the control unit (4) comprises an image acquisition manager (410) connected to the image acquisition device (210) and configured to obtain weather events in real time relating to the geographical area in which the photovoltaic plant (2) is located; and wherein the production parameters comprise in addition said weather events in real time, which are configured to be stored in the storage platform (3) associated with each energy production data.

20. The control system according to claim 19, wherein the processing component (501) is configured, in an operational production step, after the learning step:- to receive from the image acquisition manager (410) the weather events in real time and- to use the model (502) created in the learning step in order to obtain an optimal inclination configuration to be supplied to the position controller (403) of the photovoltaic panel (201) so as to maximise the current energy production considering the weather events in real time relating to the geographical area in which the panel is located.

21. The control system (1) according to claim 19 or 20, wherein the image acquisition manager (410) is configured to process the acquired images of the photovoltaic panel (201) and / or the sky to obtain the weather events in real time.

22. The control system (1 ) according to any one of claims 17 to 21 , wherein the control system (1) comprises a remote access device (404) to access for each energy production data the respective weather maps relating to the geographical area in which the photovoltaic plant (2) is located, andwherein, during an operational production step subsequent to the learning step, the processing component (501) is configured- to receive the current weather maps from the remote access device (404) and use the model (501) created in the learning step, processing such current weather maps, in order to obtain an optimal inclination configuration to be supplied to the position controller (403) of the photovoltaic panel (201 ) and maximise the current energy production; and / or- for using the model (502) in such a way as to calculate a theoretical energy production of the photovoltaic plant (2), in a different time interval, or calculate a theoretical energy production of a different photovoltaic plant (2), in a different geographical position in the different time interval, acquiring historical weather maps relative to the geographical area in which the photovoltaic plant (2) is positioned in the predetermined different time interval and / or in the different geographical position.

23. The control system (1) according to any one of claims claims 17 to 22, wherein the control unit (4) comprises: an acquisition component (405) configured for acquiring, in the learning step and in the operational production step, in a massive manner at predetermined time intervals, the production information (301), comprising the plurality of current energy production data of the plant and the respective current production parameters; a storage component (406) configured for storing the current production information (301) in the storage platform (3); wherein the processing component (501 ) is configured for processing, in the operational production step, the current production information (301) thus acquired and for updating the model (502) of the photovoltaic plant (2) by means of machine learning, analysing the current production information (301).

24. The control system (1 ) according to any one of claims 17 to 23, wherein the inverter controller (401 ) is configured for acquiring from the field inverter (207) of the photovoltaic panel (201) the energy production value, the respective energy consumption necessary for moving the photovoltaic panel (201) and the efficiency of the photovoltaic panel (201).

25. The control system (1 ) according to any one of claims 17 to 24, wherein the position controller (403) is configured for identifying the angular configuration of the photovoltaic panel (201) comprising a theoretical angle of inclination, based on an angle of elevation of the sun, and an optimal angle of inclination of the photovoltaic panel (201), which corresponds to the angle of inclination which is able to deliver the maximum value of energy production in the period of time; and wherein the position controller (403) is configured to provide to the movement device (206’; 206”) the optimal angle of inclination.

26. The control system (1) according to claim 25, wherein the control unit (4) additionally comprises a solar radiation detector (407) which is rotatable about an axis of rotation (Y) by means of a respective motor to a plurality of predetermined angular positions, wherein the position controller (403) is connected to the solar radiation detector (407) and is configured to verify, for each of the predetermined angular positions, the respective energy production value of the solar radiation detector (407) and identify, as an optimal angle of inclination among the plurality of predetermined angular positions, the angle of inclination which is capable of maximising the energy production value while limiting the respective consumption necessary for moving the photovoltaic panel (201), the position controller (403) being additionally configured to provide the movement device (206’; 206”) of the photovoltaic panel (201) with the selected optimal angle of inclination.

27. The control system (1 ) according to any one of claims 17 to 26, wherein in order to acquire the environmental parameters associated with each production data, the environmental meteorological unit (402) comprises a plurality of detection sensors, including a thermometer for acquiring the air temperature from the surrounding environment, expressed in °C, and a solarimeter for acquiring a power of irradiation of the solar radiation, and optionally also an angle of elevation of the sun; and wherein the environmental meteorological unit (402) optionally comprises a hygrometer for acquiring an air humidity from the surrounding environment, expressedas a percentage; and / or optionally a dew point sensor, expressed in °C; and / or optionally a barometer for acquiring the atmospheric pressure, expressed in hPa; and / or optionally an atmospheric rain sensor comprising a rain gauge, for acquiring a quantity of rain, expressed as daily quantity and as an intensity in mm / h, and / or optionally a snow meter for acquiring a quantity of snow, expressed as a daily quantity and as an accumulation in mm / h; and / or optionally an anemometer for acquiring a wind speed expressed in km / h and / or a wind pressure; and / or optionally a wind vane for acquiring a wind direction; and / or optionally a luminance sensor for detecting haze, or fog; and / or a optionally spectrometer for detecting UV radiation.

28. The control system according to claim 27, wherein the control system (1) is configured for horizontally positioning the photovoltaic panel (201) if the current wind speed is greater than, or equal, to a threshold value; and / or wherein the control system (1) is configured for positioning the photovoltaic panel (201 ) in a first end inclination (|31 ) , for example equal to +55°, or in a second end inclination ([32), for example equal to -55°; or for moving the photovoltaic panel (201) between the first end inclination ([31 ) and the second end inclination ([32), if the quantity of current rain indicates an intense rain, or if the quantity of current snow indicates an intense snow; and / or wherein the control system(1) is configured for moving the photovoltaic panel (201) to a cooling angle in such a way as to position a capture surface (204) of the photovoltaic panel (201) in a shade for a predetermined period of time, when the current temperature is greater than, or equal, to a maximum predetermined temperature, for cooling the photovoltaic panel; and / or wherein the control system (1) is configured for positioning vertically the photovoltaic panel (201), in the case of a defective panel.

29. The control system (1) according to claim 27 or 28, when dependent on one of claims 19 to 21 , wherein the control system (1) is configured to calculate, if the detected real-time event is hail, a hail fall angle and to movethe photovoltaic panel (201) into a protection angle, wherein the protection angle is opposite to the hail fall angle.

30. A computer program including instructions such that, when the program is run in the control system (1) of claims 17 to 29, the control system (1) executes one or more steps of the control method of claims 1 to 16.

Images