Method for determining the expected PV output

Abstract

The invention relates to a method (100) for determining the expected PV power output of a photovoltaic system (6), in particular for a future time or period, wherein a reference image file and a reference PV power output are stored in a memory for at least one time or period, comprising the steps: • Receiving (140) at least one image file, in particular by the optical sensor system, in particular a camera, • Determining (150) the expected PV power output for the future time or period as a function of the received image file and the reference image file and the reference PV power output stored for the time or period.

Description

The invention relates to a method for determining the expected PV power generated by a photovoltaic system, to a computer program, to a machine-readable storage medium and to an energy manager.Prior ArtIn order to drive the energy and heat reversals, the legislator plans that a photovoltaic installation must necessarily be planned both in the case of new buildings and in the case of remanation of old buildings. In addition, in recent years, the legislator has already paid several hundred million euros of deliveries to private households, so that heat pumps and charging devices for battery-electric vehicles are increasingly being created in the private environment.For the joint operation of photovoltaic systems, heat pumps, charging stations for battery-electric vehicles and further consumers, in a next step, care should be taken that these systems are operated together in an optimum manner and a high proportion of the available solar energy is also actually used by the heat pump or the vehicle or further electrical devices in private home instead of the purchase of electrical energy from the power grid.Energy managers are known which look at the current measured values, for example power at the domestic connection point or production of a photovoltaic installation etc., and with corresponding controllers, for example, adjust the charging power of a charging device for a battery-electric vehicle or use a heat pump for warm water treatment only at specific times.Disclosure of the InventionIt is an object of the invention to provide an improved method which brings about a reduction in the power draw from the power grid.This object is achieved by a method according to the invention, a computer program, a machine-readable storage medium and an energy manager.The method according to the invention makes it possible, inter alia, to implement improved utilization of the generated PV current and thus relief of the power grid by means of an improved prediction of the expected PV power. In particular, improvements are possible if the use of electrical consumers can already be matched to one another in the best possible manner in advance. For this purpose, the method according to the invention is helpful since it can predict the power or the amount of energy to be expected for a future time or period. In particular, the method makes it possible to take account of local weather phenomena in a targeted manner and with a sufficiently high temporal resolution. Unfortunately, this has not been possible with the available weather information from the Internet, since both local and temporal resolution is not sufficiently accurate.The invention relates to a method for determining the PV power to be expected by means of a photovoltaic system, in particular for a future time or period. Preferably, for at least one time or period, a reference image file and a reference PV power are stored in a memory.The invention comprises the following process steps:In one method step, at least one further image file is received, in particular by an optical sensor system, in particular a camera. In particular, it can be a single image file, a plurality of image files or a stream. Receiving also includes active polling.Examples of optical sensor systems are surveillance cameras, (fish eye) cameras, central observation station, drones (fleets), etc.In a method step, the PV power to be expected for the future time or period is ascertained as a function of the received image file and the reference image file stored for the time or period and the reference PV power stored for the time or period.An advantageous further development of the invention comprises the method steps listed below:In one method step, at least one image file is received by an optical sensor system, in particular by a camera. The optical sensor element or the camera provides a single image file, a plurality of image files or a stream.In one method step, the PV power of the photovoltaic installation is received, in particular from a measuring device.In one method step, the storing of the at least one received image file in a memory takes place as a reference image file together with the received PV power as a reference PV power for a time or period. The reference image file and the reference PV power are preferably acquired simultaneously, or in the same time period or within a time window.A particularly advantageous development is distinguished in that the stored reference image file and the stored reference PV power are selected which correspond to or come closest to the same time of day point or time of day space. For example, for the current time period 15:00 to 15:30, the stored time period 15:00 to 15:30 is used, in which reference image file and reference PV power are stored. If this is not present, the time used is the next time, for example 14:45 o'clock, or the closest time period, for example 14:00 to 15 o'clock. A period of time that only overlaps, for example, can also be used. 14:45 to 15:05 o'clock.An advantageous development of the method is that the optical sensor system is directed at a reference object and provides image files of the reference object or of a part of the reference object. The visibility of the reference object in the received image file is compared with the visibility of the reference object in the reference image file. If the visibility in the received image file is greater, an increased PV power than the stored reference PV power can be assumed in particular. If, on the other hand, it is foggy, for example, and thus the visibility is lower, then a lower PV power than the stored reference PV power must be assumed.An advantageous development is that the optical sensor system is directed into the sky and captures and provides image files of the sky. The degree of vaulting and the far vision in the received image file is compared with that of the degree of vaulting and / or the far vision in the reference image file. If the degree of vaulting is greater and / or the distance vision is lower, a reduced PV power than the stored reference PV power can be assumed in particularAn advantageous development is characterized in that the optical sensor system, in particular the camera, is directed at at least one partial region of the PV panels of the photovoltaic installation. Preferably, the camera is directed at at least one PV panel or a part of a panel. The expected PV power for the future time or period can be determined on the basis of the image files. The determination is made by comparing it with the reference image file. It can be determined as a function of the degree of soiling, the degree of snow and / or the degree of ice coverage. If there is a reduced degree of soiling, snow and / or ice coverage in the received image file compared to the reference image file, a greater PV power than the reference PV power can be assumed.An advantageous development is that the reference image file is replaced by the received image file for a time or period when the received PV power is greater than the reference PV power. Advantageously, the reference values are thereby adapted to changing seasons. An adaptation also takes place if it becomes smaller and smaller over a defined time. This can also be adapted depending on the date or time of year.An advantageous development is that the tag is divided into a plurality of time periods or that a plurality of times are defined distributed over the tag. Preferably, for each time or period of time, a reference image file and a reference PV power are stored, in particular stored in a memory.The memory, in particular data memory, can be designed in particular as a local data memory, a storage medium or as a cloud memory.The invention further relates to a computer program which is configured and designed to execute all steps of the method. The computer program can preferably also be executed in a cloud environment or a server. The method can preferably be executed on any computing unit, in particular a central solution, preferably a cloud or a local solution. In the case of a local solution, everything from a minicomputer to a server can be used in this case.The invention furthermore relates to a machine-readable storage medium on which the computer program is stored.The invention further relates to an energy manager which is configured and designed to carry out all steps of the method. The energy manager preferably has a computing unit which is designed to carry out the method. Furthermore, the energy manager has an interface which allows it to communicate with a photovoltaic system and / or a measuring device. The energy manager can also represent current PV powers and the future PV powers generated by means of the method. The energy manager can preferably control at least one load, in particular depending on the determined future PV power.The communication takes place in particular in a wired manner, in particular by means of a bus system, modbus, powerline, M-bus, Ethernet, EIB, CAN, KNX, EMS, OpenTherm or wirelessly, preferably Bluetooth, NFC, RFID, ANT+, Dash 7, GPRS, EDGE, UMTS, 5G, LTE, WIMAX, Zigbee, Thread, Matter, Z-Wave, 868 MHz or WIFI.A measuring device is in particular designed to record the generated PV power. Preferably, the measuring device can be part of an inverter. In particular, the inverter detects the PV power. A measuring device is in particular designed to detect the consumption of a consumer. A measurement device, in particular a further measurement device, is in particular designed to record the consumption of the electrical device.Exemplary embodiments of the invention are explained with reference to the following drawings. In the drawings, there are shown: FIG. 1 shows an exemplary building with an optical sensor system, in particular a camera, FIG. 2 shows a schematic model; and FIG. 3 shows the method according to the invention.FIG. 1 shows, by way of example, a building 5 with a photovoltaic system 6. Preferably, only the PV panel of a photovoltaic system 6 is shown. Furthermore, three optical sensor systems 10 a, 10 b, 10 c, in particular cameras, are shown by way of example.A first optical sensor system 10 ais directed at a reference object, for example a tree 8. The camera preferably captures at least a part of the reference object. The camera captures a tree 8, for example. Depending on the visibility of the reference object recorded in an image file relative to the visibility in the reference image file, fog or snow or other weather factors influencing the PV performance can be inferred. The poorer the visibility, the lower the PV power will be compared to a reference PV power linked to the reference image file.A second optical sensor system 10 bis exemplarily directed into the sky. The camera system captures the sky. Depending on the degree of vaulting and / or the remote view relative to the degree of vaulting and / or the remote view shown in the reference image file, the future power can be deduced together with the reference PV power.A third exemplary optical sensor system 10 bis directed at least to a partial region of the PV panels of the photovoltaic system 6. The future performance may be inferred depending on the level of fouling, the level of snow and / or ice coverage of the PV panel relative to the level of fouling, the level of snow and / or ice coverage shown in the reference image file along with the reference PV performance.In FIG. 2, a schematic model 200 is shown. The model 200 enables the method 100 according to the invention to be understood more easily. The model 200 includes an ideal day sub-model 210 and a weather model 220 that can reduce generated power according to the weather information.The model for the ideal tag 210 is based in particular on a database or a table which calculates the PV power P(t) over the course of an individual tag as a function of the time t. During operation, the PV power P meas(t*) measured at a time t* is used to adapt the value stored in the table or database for the time t*.Preferably, as P meas an average PV power in a certain time window, in particular time period, around the time t* is used. According to a further development, a subdivision into time periods or time slots takes place. Preferably, P meas(t*), in particular in the case of a plurality of measurement values, an average PV power is used in the defined time period or time window.Preferably, a new entry P (ref,new) is calculated on the old entry P (ref,old) for a defined period of time according to the following rule:If the actually measured PV power at this point in time is greater than or equal to the value from the model, then directly the measured solar power P meas is adopted as a new entry for the time period t*. The table model can thus be quickly adapted to the longer days and the greater solar power during the spring.In the autumn, on the other hand, a weighting factor γ ensures that the model can also adapt to shorter days and a weaker PV power with a somewhat delay again. The weighting factor γ is usually between the values zero and one and can be fixedly predefined, for example. It is advantageous that the weighting factor can be selected as a function of a specific error measure e between the profile of table model P ref(t) and the measured values P meas(t) over a day. As a result, a more rapid adaptation is carried out on counter days of the beauty (small error measure) than on counter days of the poor (larger error measure). The following relationship can be used for this purpose:Care must be taken to restrict the quotient to values between zero and one. The parameters gd (=good day) and bd (=bath day) form threshold values which have to be defined in advance as appropriate for characterizing a fine weather day (gd) or a fine weather day (bd). The parameter R describes whether it is a beauty weather day or a bad weather day within these thresholds. Gamma is 1 on good days and 0. In the range between gd and bd, R.R is between gd and bd. Gd and bd are determined depending on the signal quality of the power measurement. For example, gd=0.1 and bd=0.2 may be selected, and then R moves therebetween.P ref( t) corresponds to the expected PV power on an ideal day at a defined time t or in a defined period of time.The database / table for PV power production on an ideal day contains, for example, entries at a time interval of 15 min or 5 min depending on the desired resolution and available storage space. Preferably, any resolution is conceivable, 30s, 1' 2s, 3' etc. Preferably, the periods are 5 or 15 min. long. Preferably, any length of the time period is possible, in particular 30 s, 1' 2 min, 3 min etc.For the PV power P meas it is possible, for example, either to use directly the measured production of the photovoltaic installation or the power actually flowing into a load, in particular an e-vehicle. Due to the fact that the charging process in the e-vehicle has to be monitored by the onboard computer and this process itself requires power, measurable differences can be present here. The same applies to other consumers. The same applies to the submodel described below.The method 100 according to the invention serves for determining the PV power to be expected by means of a photovoltaic system 6, in particular for a defined period of time, which is in particular in the future. For this purpose, the weather is preferably monitored. The weather model 220 is used for monitoring. The weather model 220 is likewise, for example, a multidimensional table model or a database.The weather model describes the efficiency η between the actual solar power P meas and the power P ref. expected on an ideal day (according to the previously ideal day model 210). P ref corresponds to the stored reference PV power. The efficiency essentially depends on the degree of vaulting c, the visibility v and the ice and snow covering s, which can be provided with an optical sensor system 10, in particular a camera, preferably a local camera system, preferably in real time.Providing is understood to mean providing, but also making transmission retrievable. The camera provides at least one image file.During the adaptation of the model 220, it is ensured according to the following formula that the adaptation runs correspondingly with a delay and no brief phenomena are detected:The corresponding weighting factor λ lies in the interval [0,1) and must be fixedly predefined here. In a simplified version, a threshold for s can be defined, starting from which a constant efficiency of 0 is output and no adaptation of the model, in particular of the entry in the table and / or the database, takes place. Preferably, from a three-dimensional table over v, c, s, an only two-dimensional table over v and c is obtained, which requires less storage space.By using the camera data to determine the local weather precisely, the efficiency parameters are particularly accurate.The expected PV power and energy amount over this period of time are displayed to the user in particular in the form of a curve profile and as numerical value (e.g. per day, per week, per month, per year). In particular when the PV power is used to charge surplus power, the user can decide whether the power flowing into the vehicle is sufficient for the imminent journeys using the e-vehicle.In FIG. 3, the method 100 according to the invention is shown in detail.In an optional method step 110, at least one image file is received by an optical sensor system, in particular by a camera. The optical sensor system provides one or more image files, in particular also as a stream.Providing is understood here to mean providing, in particular so that it can be called up and sending. The image file preferably has metadata which comprise the time of creation.Receiving is understood to mean that the information is received but also called up.Preferably, the optical sensor system may provide a stream consisting of a plurality of image files.In a further optional method step 120, a determined, in particular measured, PV power value P meas. is received. The PV power P meas is measured and provided by a measuring device, in particular a meter or an inverter.Preferably, the PV power and the image file are captured at a similar time, in particular the same time, or in the same time period. Preferably, it is detected within the same time range.In a further optional method step 130, the storing of the at least one generated image file in a memory takes place as a reference image file together with the detected PV power as a reference PV power P ref. Preferably, a reference image file and / or a reference PV power P ref is stored for each time of day or each time of day space.In a method step 140, a further image file is received. The image file is provided by the optical sensor system, in particular the camera.In a further method step, the PV power to be expected is determined 150 for a future time or period as a function of the received further image file, the stored reference image file and the stored reference PV power.For example, to determine the expected PV power for 15 o'clock, the reference image file and reference PV power are used at the same time, i.e., 15 o'clock or a time period, for example, 14:45 o'clock to 15:15 o'clock. In particular, the tag is divided into a plurality of times or periods.For each time or period, a reference image file and a reference PV power are stored. If the future PV power to be expected is to be determined for a time or a period for which there is no reference image file or reference PV power P ref then the closest time or period with stored information is used.The method 100 can be implemented both on an app in the mobile telephone or on an Internet site or a cloud service, wherein a retrieval of the information on different terminals is then possible.In addition to the control of the excess charging, numerous further applications for these methods are possible, such as in particular the selective activation of heat pumps, air conditioners or other terminals which can store thermal energy in a closed space if an energy excess is currently available but a lack of PV energy is more likely to be expected for the next days.If sufficient PV energy is expected even in the next days, then no comfort-reducing measures are necessary, if appropriate.According to a development, the optical sensor system, in particular the camera, can be configured centrally. In particular, an optical sensor system may take over the acquisition of reference image files and image files for an area, a building complex, a city part, community, dock, city or area. The range must not be chosen too large, since otherwise the method becomes too inaccurate.Preferably, the reference image files and image files are centrally acquired and provided. It is not necessary for each user to require a separate optical sensor system. The provision is effected, for example, via a server, the cloud or a storage system which allows remote access to it. The method is extended in particular by receiving the reference image file and the reference PV power.According to a development, the method can be carried out centrally. The expected PV power is then provided and can be retrieved. In particular, energy managers, heat pumps or inverters can retrieve expected PV performance. The energy manager, the heat pumps or the inverter receives the determined expected PV power. This is provided in an additional method step.Correction factors can also be incorporated which correct local differences. In particular, a service can be provided in this way, which makes it possible to carry out the method centrally and make it available to a plurality of users or buildings. In particular, the number of cameras etc. can be reduced by means of a central infrastructure.The method can also be carried out by buildings or users. The expected PV power determined is then provided to at least one neighbor or a plurality of neighbors or an area, or a city part, a community, a dock, a city. The provision takes place in particular via a communication interface, a memory, a server or the cloud.The expected PV powers determined can also be collected and evaluated centrally. In particular, as information to monitor PV performance in areas. The power network can preferably be controlled by means of the information.

Claims

Method (100) for determining the PV power to be expected by means of a photovoltaic system (6), in particular for a future time or period, wherein a reference image file and a reference PV power are stored in a memory for at least one time or period, comprising the steps: • receiving (140) at least one image file, in particular by means of the optical sensor system, in particular a camera, • determining (150) the PV power to be expected for the future time or period as a function of the received image file and the reference image file stored for the time or period and the reference PV power stored for the time or period.Method according to the preceding claim, characterized bythe steps of: • receiving (110), by an optical sensor system (10, 10a, 10b, 10c), in particular by a camera, at least one image file, • receiving (120) the PV power of the photovoltaic installation, • storing (130) the at least one received image file in a memory as a reference image file together with the received PV power as a reference PV power for a time or period.Method according to one of the preceding claims, characterized in that the stored reference image file and the stored reference PV power are selected which corresponds to or comes closest to the same time of day point or time of day space.Method according to one of the preceding claims, characterized in that the optical sensor system (10, 10a), in particular the camera, is directed at a reference object (8) and provides at least one image file which at least partially shows the reference object (8), and in that the PV power to be expected is determined (130) for the future time or period as a function of the visibility of the reference object (8) in the reference image file and the visibility of the reference object (8) in one of the received image files and the reference PV power.Method according to one of the preceding claims, characterized in that the optical sensor system (10, 10b), in particular the camera, is directed into the sky and provides at least one image file which shows at least a section of the sky, and in that the determination (130) of the expected PV power for the future time or period takes place as a function of the degree of bulging and / or the remote view in the reference image file and the degree of bulging and / or the remote view in one of the received image files and the reference PV power.Method according to one of the preceding claims, characterized in that the optical sensor system (10, 10b), in particular the camera, is directed at at least one partial region of the PV panels of the photovoltaic installation (6) and provides at least one image file which shows at least one section of a PV panel, and in that the PV power to be expected is determined for the future time or time interval on the basis of the degree of contamination, the degree of snow and / or ice coverage of the PV panel in the reference image file and the degree of contamination, the degree of snow and / or ice coverage of the PV panel in one of the received image files and the reference PV power.Method according to one of the preceding claims, characterized in that the reference image file is replaced by the received image file for a time or period if the received PV power is greater than the reference PV power for this time or period.Method according to one of the preceding claims, characterized in that the tag is divided into a plurality of time periods or a plurality of times distributed over the tag are defined, and in that a reference image file and a reference PV power are stored for each time period or time period.Method according to one of the preceding claims, characterized in that the efficiency (η) is determined during the determination, and in that the expected PV power is dependent on the determined efficiency.Computer program which is configured to carry out all the steps of the method (100) according to one of the preceding claims.A machine readable storage medium having stored thereon the computer program of the preceding claim.A power manager configured to perform all the steps of the method of any one of claims 1 to 9.

Images