Method and system for monitoring a power grid, method for training an artificial intelligence to predict a future state of a power grid, computer program and computer-readable data carrier

Abstract

Method for monitoring a voltage network (12), comprising the steps: - Recording at least one time series of a first characteristic parameter of the voltage network (12), - Recording at least one time series of a second characteristic parameter of the voltage network (12) that is different from the first characteristic parameter of the voltage network (12), - Feeding the time series of the first characteristic parameter and the time series of the second characteristic parameter into a processor unit (16), which processes the two time series of the different characteristic parameters together in such a way that a future state of the voltage network (12) is predicted based on the two time series of the different characteristic parameters, whereby correlations between the two different characteristic parameters are exploited, on the basis of which a future state of the voltage network (12) is reliably predicted, and Output of the predicted future state of the voltage network (12), wherein the voltage network (12) is a building network assigned to a private household or an industrial building and connected to a supply network (13).

Description

[0001] The invention relates to a method for monitoring a power grid. Furthermore, the invention relates to a method for training an artificial intelligence to predict the future state of a power grid. In addition, the invention relates to a system for monitoring a power grid, a computer program, and a computer-readable data carrier.

[0002] Today's power grids are significantly more volatile than in the past. This is due, among other things, to the constantly increasing but fluctuating energy demand, as well as the progressive and increasingly irregular integration of renewable energy sources into the grid. Furthermore, it is expected that power grids will become even more volatile in the future, as the share of renewable energy fed into the grid will continue to rise, while at the same time the grids will face greater short-term loads, for example, due to the rise of electric vehicles, particularly in the automotive sector. This increasing volatility of the power grids, however, leads to a decline in grid quality, which can result in more frequent disruptions and even outages – something that must be avoided.

[0003] For network operators and electricity users / customers, it is therefore already important to monitor the current state of the power grid, analyzing past grid disturbances to identify potential faults. Smart devices like the DEHNrecord are used for this purpose, recording and analyzing a characteristic parameter of the power grid to determine its current power quality. Such smart devices are used, for example, by network operators and electricity users / customers.

[0004] Therefore, the voltage network being analyzed could be the local voltage network, also known as the building network, i.e., the voltage network at the customer's premises, for example, a private household or an industrial building. In other words, a low-voltage network is being monitored.

[0005] From EP 3 336 995 A1, a method and system are known for determining the operating state of a subnetwork of an energy supply network. For this purpose, several input values ​​from different sources are fed into an artificial intelligence, which processes these input values ​​to output state values ​​in order to determine the operating state.

[0006] From EP 3 751 699 A1 a method and system are known in which a network state of a distribution network is predicted based on active power using a naive Bayes classifier.

[0007] However, due to the increasing volatility of the power grids in the future, the analysis of the current grid quality will probably no longer be sufficient, so improved analysis or monitoring of the power grid is needed.

[0008] The purpose of the invention is to provide a means of monitoring the power grid in a cost-effective and high-quality manner. - The problem is solved according to the invention by a method for monitoring a voltage network according to claim 1.

[0009] The basic idea of ​​the invention is that future grid quality is estimated or predicted based on historical and / or current data of the power grid, enabling the grid operator and / or the electricity customer / user to detect future disturbances at an early stage. In other words, the grid operator and / or the electricity customer / user can predict the time of an emerging grid disturbance and its intensity. This makes it possible to initiate appropriate countermeasures to avert the emerging grid disturbance, at least reduce its intensity, or generally minimize its impact on specific areas.

[0010] The invention is based on the understanding that, due to the volatile nature of power grids, the future performance of the power grid in terms of its quality (grid state) is often more crucial for the grid operator and / or the electricity customer / user than the current state, to which no action can be taken. Looking into the future allows for the timely initiation of appropriate measures to counteract the predicted grid disruption, i.e., the deterioration of the power grid's condition.

[0011] In other words, the network operator and / or the electricity customer / user is enabled to design the voltage network to be more resilient to future expected disturbances.

[0012] The future state of the power grid can be predicted for the next hours, days or weeks, giving the grid operator and / or the electricity customer / user enough time to initiate any necessary countermeasures in a timely manner, so that they can still counteract a predicted disruption or deterioration.

[0013] The countermeasures to be taken can, for example, consist of avoiding further loads during a predicted period of low voltage in the power grid, thereby ensuring the reliable operation of devices and / or machines whose operation should not be disrupted. In this respect, the countermeasure can consist of disconnecting non-essential machines or devices from the grid during the predicted period of low voltage in order to reduce or prevent further increases in the load.

[0014] In principle, the voltage network can be a local network, also known as a building network, assigned to a private household or an industrial building. The voltage network can therefore be a low-voltage network, for example, a building network connected to a supply network.

[0015] For forecasting purposes, the first characteristic parameter can be recorded multiple times, i.e., at least at two different points in time. The values ​​of the first characteristic parameter recorded at these different points in time are then fed into the processor unit for evaluation. Thus, the future course of the first characteristic parameter of the power grid can be predicted from a historical time series recorded up to the present day. This future course can refer to the first characteristic parameter, so that its future course is predicted, which corresponds to the predicted future state of the power grid.

[0016] The first characteristic parameter alone allows for a characterization of the current state of the voltage network. Since the processor unit jointly evaluates at least two values ​​of the first characteristic parameter, which were recorded at different times, it is now possible to predict future states of the voltage network accordingly, which would not be possible based solely on the evaluation of a single characteristic parameter.

[0017] In particular, correlations between the at least two different points in time of the first characteristic parameter are used, based on which a future state of the voltage network can be reliably predicted.

[0018] A second parameter, different from the first, can also be used. This second parameter can be a separate parameter that is not necessarily a parameter of the voltage network.

[0019] The second parameter can be obtained from a database to include, for example, environmental data, in particular weather data, time data (weekday, weekend, holiday, day and / or night) or usage data of systems in the environment, such as a train timetable or usage data of electric charging systems in the environment.

[0020] Therefore, the second parameter can provide information about the difference between the two different points in time. In particular, it makes it possible to establish a relationship between two different values ​​of the first characteristic parameter that were recorded at the two different points in time.

[0021] The second quantity can also be a second characteristic parameter of the voltage network, which differs from the first characteristic parameter of the voltage network.

[0022] In particular, the time course of the first characteristic parameter, which has been recorded at least at the first and second time points, can be evaluated together with the second parameter, which differs from the first characteristic parameter of the voltage network. Therefore, a time series of the first characteristic parameter can be evaluated together with other data, for example, data obtained from a database such as weather data, time data, and / or usage data from systems in the vicinity.

[0023] This means that a data fusion of two different quantities can be used to provide insights into the future state of the power grid. In other words, the joint processing of these two different quantities exploits a correlation between them, allowing conclusions to be drawn about the future behavior of the power grid. Therefore, the future state of the power grid can be predicted accordingly.

[0024] In principle, the procedure for monitoring a voltage network can also include the following steps: - Recording at least one first, characteristic parameter of the voltage network, in particular at a first time and a second time, - Recording at least one second characteristic parameter of the voltage network that differs from the first characteristic parameter of the voltage network, - Feeding the first characteristic parameter, in particular the values ​​recorded at the first and second time points, and the second characteristic parameter into a processor unit that processes both different characteristic parameters together in such a way that a future state of the voltage grid is predicted based on the two different characteristic parameters, and - Output of the predicted future state of the power grid.

[0025] One aspect stipulates that the first characteristic parameter and / or the second characteristic parameter is a voltage, a current, a power, a frequency, a distortion (Total Harmonic Distortion - THD), a harmonic (up to the 50' harmonic), a reactive power, and / or an energy value. The corresponding characteristic parameter can be recorded for each phase of a multi-phase voltage network, e.g., for phases L1 to L3.

[0026] Another aspect involves recording the first characteristic parameter multiple times, resulting in a time series of more than two points in time for this parameter, which is then processed. Alternatively or additionally, the second parameter, particularly the second characteristic parameter, can also be recorded multiple times, thus generating a time series for this parameter, which is then processed. In this way, the respective parameter—the first characteristic parameter or the second parameter—is recorded multiple times sequentially, especially periodically, to obtain a time series or sequence of the corresponding parameter. Based on this time series or sequence, the future state of the voltage grid is predicted. In other words, at least the historical behavior of the first characteristic parameter is recorded and considered when predicting its future state.This can also be done analogously for the second variable, for example, the second characteristic parameter. For instance, a change in the first characteristic parameter can be detected, which, together with the change over time of the second variable, is uniquely representative of a specific future state of the power grid. The change over time can be a gradual deterioration, but also, more generally, characteristic fluctuations or patterns that, considered on their own, would initially have no significance. The change in the second variable could, for example, be due to a different day; a changing weather situation; a change due to the timetable; a change in charging system usage; or similar factors. By combining the two variables, it is possible to obtain information that is relevant to the future state of the power grid, provided the...Sizes influence each other.

[0027] According to one embodiment, the processor unit comprises an artificial intelligence that receives at least the first characteristic parameter recorded at at least two different times as an input and outputs the future state of the power grid as an output. The artificial intelligence can be trained to recognize the corresponding correlation between the values ​​of the first characteristic parameter recorded at the different times and to learn corresponding technical relationships within the power grid, thereby enabling reliable and highly accurate predictions regarding the future state of the power grid.

[0028] Furthermore, a method for training an artificial intelligence to predict the future state of a power grid is described. The training method comprises the following steps: - Providing a training dataset for the artificial intelligence that includes at least a first characteristic parameter of the voltage network at a first time point, the first characteristic parameter of the voltage network at a second time point, and an actual state of the voltage network at a third time point, which is later in time than the first and second time points. - Feeding the first characteristic parameter at the first time point and the first characteristic parameter at the second time point into a processor unit that has the artificial intelligence to be trained, wherein the processor unit having the artificial intelligence processes the first characteristic parameter recorded at the different times point in time together and outputs a predicted future state of the voltage network at the third time point, wherein correlations between the at least two different times point in time of the first characteristic parameter are used, on the basis of which a future state of the voltage network is reliably predicted, - Comparing the predicted future state of the voltage network at the third time point with the actual state of the voltage network at the third time point, which is part of the training data set, to determine any deviation between the predicted future state of the voltage network at the third time point and the actual state of the voltage network at the third time point, and - Feeding back the deviation between the predicted future state of the voltage network at the third time point and the actual state of the voltage network at the third time point in order to adjust weighting factors of the artificial intelligence to be trained, provided that the deviation is outside a tolerance range.

[0029] The artificial intelligence can therefore be trained using at least one training dataset that includes the first characteristic parameter at both a first and a second time point, in particular a corresponding time sequence or time series of the first characteristic parameter, as well as an actual future state of the voltage grid that exists at the third time point, which is temporally located after the times at which the first characteristic parameter was determined or recorded, i.e., corresponding values ​​of the first characteristic parameter. In particular, the first and second time points do not coincide, so that the first characteristic parameter is recorded at two different times, in particular the respective values ​​of the first characteristic parameter.

[0030] In principle, when training artificial intelligence, it can also be provided that the artificial intelligence does not output the state of the power grid at a discrete point in time, but at an interval which includes the third point in time, thus predicting the future state of the power grid for a future period.

[0031] The corresponding tolerance range for the deviation can be predefined and / or set by a user, e.g. as a percentage or variance.

[0032] In principle, an artificial intelligence trained using the method described above can be used to predict the future state of the power grid.

[0033] In other words, the processor unit comprises a trained artificial intelligence, in particular an artificial intelligence trained according to the procedure described above.

[0034] Another aspect stipulates that the artificial intelligence must include at least one artificial neural network, such as a recurrent neural network (RNN) or a convolutional neural network (CNN). The artificial intelligence may also employ a long short-term memory (LSTM) network or a gated recurrent unit (GRU). Such neural networks allow for the prediction of a future state based on time series data. The artificial intelligence thus learns from past data (historical data), thereby predicting the future state of the electrical network.

[0035] In particular, a multidimensional vector is generated that includes the first characteristic parameter at different times, i.e., the first characteristic parameter recorded at at least two different times. This multidimensional vector is then processed by the processor unit. The artificial neural network is thus configured to process a multidimensional vector that, for example, includes multiple times of the first characteristic parameter in order to predict the future state of the voltage network. Besides the first characteristic parameter, the multidimensional vector can also include other parameters, such as at least a second parameter, especially at multiple times. This second parameter can be a parameter retrieved from a database or a second characteristic parameter of the voltage network itself.For example, the multidimensional vector has up to eight dimensions, meaning it can contain data on up to eight different characteristic parameters, such as information regarding voltage, current, power, frequency, distortion, harmonics, reactive power, and / or energy value. The additional parameters besides the first characteristic parameter can, in principle, be singular values ​​or (also) corresponding time series or sequences.

[0036] Furthermore, at least a future temporal progression of a characteristic parameter can be predicted as the future state of the power grid. Artificial intelligence can therefore, for example, predict a future voltage profile for a specific time interval. Likewise, artificial intelligence can predict a future current profile for the corresponding time interval. The time interval can be defined by a user, enabling predictions for a targeted period. This is particularly relevant when certain machines or devices connected to the monitored power grid must operate without disruption within a defined timeframe. The operator or grid operator thus has an interest in ensuring that, at least during this period, no or only very minor disruptions to the power grid occur.

[0037] In principle, the future time course of a characteristic parameter can be predicted, for example, the future time course of the previously recorded characteristic parameter. However, it is also possible to predict the future time course of a different characteristic parameter than the one that was recorded and fed into the processor unit.

[0038] Furthermore, the invention solves the problem by providing a system for monitoring a power grid, wherein the system includes a processor unit configured to perform one of the aforementioned methods. The system can be a measuring and analysis device integrated into the power grid to be monitored, for example, on a DIN rail. The aforementioned advantages thus apply analogously to the system.

[0039] In particular, the system may include a device, for example a measuring and / or evaluation device or a protective device such as a surge protection device (SPD), wherein the device includes the processor unit.

[0040] The system, and in particular the device, can also include at least one data transmission device, enabling the system, and in particular the device, to communicate with other devices, for example, and in particular to receive data that is additionally used for evaluating or predicting the future state of the power grid. Furthermore, the output of the predicted future state of the power grid can be provided via the data transmission device, which is configured, for example, as a communication interface.

[0041] Furthermore, the invention provides a computer program which includes program code means to perform the steps of one of the aforementioned methods when the computer program is executed on a processor unit, e.g. the processor unit of the aforementioned system.

[0042] Furthermore, the invention provides a computer-readable data carrier on which the computer program of the aforementioned type is stored.

[0043] The aforementioned advantages thus apply analogously to the computer program and the computer-readable data carrier.

[0044] Further advantages and features of the invention will also become apparent from the following description and the drawings, to which reference is made. The drawings show: - Fig. 1 a schematic representation of a system according to the invention, - Fig. 2 an overview illustrating the inventive method for monitoring a voltage network, and - Fig. 3 an overview that presents the inventive method for training an artificial intelligence.

[0045] In Fig. Figure 1 shows a system 10 which monitors a voltage network 12, which in the illustrated embodiment is a multi-phase building network connected to a supply network 13. In this respect, the voltage network 12 is, for example, a low-voltage network.

[0046] System 10 comprises several measuring devices 14 that are connected to a device 15 for signal transmission, for example, a protection device or a measuring and / or evaluation device. In the embodiment shown, both the measuring devices 14 and the device 15 are connected to or integrated with the voltage network 12 to be monitored. Alternatively, the device 15 can also be designed separately from the voltage network 12, so that it is only connected to the measuring devices 14 for signal transmission.

[0047] The device 15 includes a processor unit 16, which is configured to execute a computer program 18 using program code means to monitor the voltage network 12.

[0048] The computer program 18 can be installed on the processor unit 16 from a computer-readable data carrier 20 on which the computer program 18 is stored, so that the system 10, in particular the device 15 comprising the processor unit 16, is able to carry out a corresponding procedure for monitoring the voltage network 12.

[0049] The device 15 therefore has a corresponding interface via which the data carrier 20 can be connected to the device 15 in order to install the computer program 18.

[0050] The computer program 18 may also have been transmitted via a data transmission device 22, which is configured, for example, as a communication interface, for instance for communication, particularly wireless, with the measuring devices 14 and / or a server. It may also be a LAN interface through which communication of the device 15 is possible.

[0051] The procedure, which is executed by computer program 18, is described below using the following examples: Fig. 2 explained.

[0052] In a first step S1, at least one characteristic parameter of the voltage network 12 is recorded at a first time point. This can be done using one of the measuring devices 14 or via the device 15 itself. In any case, it is ensured that the device 15, in particular its processor unit 16, records the first characteristic parameter of the voltage network 12.

[0053] In a second step S2, the first characteristic parameter of the voltage network 12 is recorded at a second time point in time, which is different from the first time point in time.

[0054] In principle, the first characteristic parameter of the voltage network 12 can be recorded at several points in time. The first characteristic parameter is therefore recorded several times consecutively over time, in particular periodically, so that a time series / sequence of the first characteristic parameter exists.

[0055] Additionally, it may be provided that at least one second parameter, for example a second characteristic parameter of the voltage network 12, is recorded, which differs from the first characteristic parameter of the voltage network 12. The second parameter can be recorded independently of the voltage network 12, for example read from a database to which the device 15 has access, in particular via the communication interface 22. Data can be obtained from the database that are included in the analysis, for example environmental data such as weather data, time data (weekday, weekend, holiday, day and / or night) or usage data of systems in the environment, for example a timetable for trains or usage data of electric charging systems in the environment.This allows additional information to be obtained that explains any possible difference between the two values ​​of the first characteristic parameter of the voltage network 12, which were measured at the two times.

[0056] The second quantity obtained from the database is transferred to the processor unit 16 of the device-15, so that the processor unit 16 records the second quantity of the voltage network 12.

[0057] The second quantity can also be a second characteristic parameter of the voltage network 12, which is recorded by one of the measuring devices 14 or by the device 15 itself, wherein the second characteristic parameter is transmitted to the processor unit 16 of the device 15, so that the processor unit 16 records the second characteristic parameter of the voltage network 12. If two characteristic parameters are determined, the two characteristic parameters differ from each other.

[0058] In principle, the first characteristic parameter of the voltage network 12 can be a voltage, a current, a power, a frequency, a distortion, a harmonic, a reactive power and / or an energy value of the voltage network 12, in particular of a phase of the multiphase voltage network 12. This applies analogously to the second characteristic parameter, provided that it is also recorded.

[0059] As mentioned above, a time series / sequence of the first characteristic parameter of the voltage network 12 is recorded. This can also apply to the optionally recorded second parameter.

[0060] For example, the second quantity is always recorded when the first characteristic parameter of the voltage network 12 is recorded, so that the values ​​of the respective quantity are recorded in parallel, in particular simultaneously.

[0061] This means in particular that when recording the first characteristic parameter of the voltage network 12 at the different times, a corresponding data set is also read from the database, which represents the second parameter, for example the weather data available at the respective times.

[0062] In a third step S3, at least the first characteristic parameter recorded at the first time and the first characteristic parameter recorded at the second time are fed into the processor unit 16, which processes the first characteristic parameter recorded at the first time and the first characteristic parameter recorded at the second time together, so that a future state of the voltage network 12 is predicted by the processor unit 16 based on the first characteristic parameter recorded at the at least two different times.

[0063] Additionally, it may be provided that the second quantity, which is recorded at least once and is different from the first characteristic parameter of the voltage network 12, is also processed together in order to predict the future state of the voltage network 12 by the processor unit 16.

[0064] For forecasting purposes, the processor unit 16 has an artificial intelligence 24 which receives at least the first characteristic parameter recorded at at least two different times as an input parameter and outputs the future state of the voltage network 12 as an output parameter.

[0065] The artificial intelligence 24 can include at least one artificial neural network, e.g. a convolutional neural network (CNN) or an artificial recurrent neural network (RNN), such as a Long Short-Term Memory (LSTM) network or a “Gated Recurrent Unit” (GRU).

[0066] Accordingly, the artificial intelligence 24 is able to predict a future state of the voltage grid 12, at least based on the obtained time series, i.e., the time sequence of the first characteristic parameter. To do this, the artificial intelligence 24 processes the multiple values ​​of the first characteristic parameter, which were recorded at different times, by generating it as a multidimensional vector and processing it accordingly.

[0067] The multidimensional vector can have up to eight dimensions, for example, eight dimensions per phase of a multiphase power grid. These eight dimensions can be represented by voltage, current, power, frequency, distortion, harmonics, reactive power, and energy. In principle, however, the vector's dimensions can be higher, depending on the application.

[0068] The artificial intelligence 24 may have been previously trained using a method in which the artificial intelligence 24 was trained to predict the future state of the voltage grid 12 based on values ​​of the first characteristic parameter that were recorded at at least two different times. Consequently, the artificial intelligence 24 is a trained artificial intelligence 24.

[0069] The corresponding procedure for training the artificial intelligence 24 is in Fig. Figure 3 is shown, which will be referred to below.

[0070] In a first training step T1, a training dataset is provided for the artificial intelligence 24, which includes at least the first characteristic parameter of the voltage network 12 at a first time point, the first characteristic parameter of the voltage network 12 at a second time point, and an actual state of the voltage network 12 at a third time point. The third time point is later than the first and second time points, so that it is a future time point to be predicted based on the first and second time points.

[0071] Additionally, the training dataset may contain a second variable that differs from the first characteristic variable, so that the training dataset comprises at least two different variables. The second variable may be a second characteristic variable of the voltage network 12, which differs from the first characteristic variable.

[0072] In particular, the training data set can include a time series or time sequence of the first characteristic parameter, such that several values, preferably more than two values, of the first characteristic parameter are available, which have been measured or recorded at different times.

[0073] The optional second quantity can also be included as a time series or sequence in the training dataset. The training dataset can therefore comprise data from at least two different quantities, in particular two different characteristic parameters of the voltage network 12, for a specific period, as well as information on the state of the voltage network 12 that was obtained at a later time than the specified period.

[0074] However, it is also possible that the training data set includes corresponding information from more than just two different quantities, in particular from more than two different characteristic parameters of the voltage network 12, thereby providing more information or data overall, making the training more comprehensive and the predictive power of the correspondingly trained artificial intelligence 24 higher.

[0075] In a second training step T2, at least the first characteristic parameter recorded at the first time point and the first characteristic parameter recorded at the second time point, in particular the time series or sequence of the first characteristic parameter, are fed into the processor unit 16, which contains the artificial intelligence 24 to be trained. The processor unit 16, containing the artificial intelligence 24, processes the first characteristic parameter recorded at the different time points, in particular the time series or sequence, together and outputs a predicted future state of the voltage network 12 at the third time point, at which the training data set includes the actual state of the voltage network 12.

[0076] Accordingly, during training, the artificial intelligence 24 learns corresponding relationships between the first characteristic parameter recorded at the two different times and its effect(s) on the later state of the voltage network 12, so that the artificial intelligence - 24 is trained to predict the future state of the voltage network 12 based on the past or current data.

[0077] Additionally, the second parameter can optionally be included, which allows correlations between the different parameters to be identified during training, i.e., correlations between the values ​​of the first characteristic parameter, which were recorded at the different times, and the second parameter, which was recorded once or also at the different times.

[0078] In a third training step T3, the predicted future state of the power grid 12 at the third time point is compared with the actual state of the power grid 12 at the third time point, the latter being included in the training dataset. This comparison identifies any deviation between the predicted future state of the power grid 12 and the actual state of the power grid 12. Thus, the training determines how accurate the artificial intelligence 24's prediction already is, i.e., how well the prediction reflects the actual state.

[0079] In a fourth training step T4, the detected deviation between the predicted future state of the voltage network 12 and the actual state of the voltage network 12 is fed back into the artificial intelligence 24 being trained in order to adjust the weighting factors of the artificial intelligence 24 being trained, provided that the deviation lies outside a tolerance range. This is in Fig. 3 is indicated by the corresponding arrow. The tolerance range may be predefined and / or set by a user.

[0080] Subsequently, at least training step T3 is repeated, progressively reducing the deviation detected during the comparison in training step T3. After a certain number of repetitions (iterations), the deviation becomes so small that it falls within the tolerance range and is no longer fed back into the system. At this point, the artificial intelligence has reached at least a (pre-)trained state for the training dataset, making it usable.

[0081] The artificial intelligence 24 can then be further trained using the same training steps T1-T4 with corresponding iterations, so that the artificial intelligence 24 is trained, for example, on further characteristic parameters of the voltage network 12, in particular different pairings of characteristic parameters of the voltage network 12.

[0082] In particular, the training of the artificial intelligence 24 can also include the input of more than two different quantities, especially characteristic parameters of the voltage network 12, e.g., up to eight different quantities or more. The corresponding training set used for this purpose therefore contains more data, which is provided and fed in.

[0083] Typically, training steps T1 to T4 are repeated for several different actual states of the voltage network 12 and / or several different data points of the characteristic parameters, in particular for several different characteristic parameters, in order to train the artificial intelligence 24. In the final training step T4, as already described, the weighting factors of the artificial intelligence 24 to be trained are adjusted such that the respective predicted future state of the voltage network 12 is always within the tolerance range.

[0084] The artificial intelligence 24 used in the procedure for monitoring the voltage network 12 has been trained according to the aforementioned training procedure, so that it is a trained artificial intelligence 24 which, based on at least two values ​​of the first characteristic parameter, which have been recorded at different times, predicts the future state of the voltage network 12.

[0085] The processor unit 16, which includes the trained artificial intelligence 24, therefore outputs in a fourth step S4 the predicted future state of the voltage network 12.

[0086] This can involve a time course of a characteristic parameter of the voltage network 12, in particular a characteristic parameter that differs from the first characteristic parameter (and the second characteristic parameter). For example, the artificial intelligence 24 is trained, among other things, to predict the future course of a total harmonic distortion (THD) of the voltage network 12, based on historical data (time series or time sequences) of the current and / or voltage.

[0087] In principle, this makes it possible for the network operator and / or the electricity customer or the electricity user to recognize future disturbances or periods of weakness in the voltage network 12 in a timely manner and to initiate appropriate countermeasures in order to operate the voltage network 12 stably.

Claims

[1] Method for monitoring a voltage network (12), comprising the steps: - Recording at least one time series of a first characteristic parameter of the voltage network (12), - Recording at least one time series of a second characteristic parameter of the voltage network (12) that is different from the first characteristic parameter of the voltage network (12), - Feeding the time series of the first characteristic parameter and the time series of the second characteristic parameter into a processor unit (16), which processes the two time series of the different characteristic parameters together in such a way that a future state of the voltage network (12) is predicted based on the two time series of the different characteristic parameters, whereby correlations between the two different characteristic parameters are exploited, on the basis of which a future state of the voltage network (12) is reliably predicted, and Output of the predicted future state of the voltage network (12), wherein the voltage network (12) is a building network assigned to a private household or an industrial building and connected to a supply network (13). [2] Method according to claim 1, characterized by, that the first characteristic parameter or the second characteristic parameter is a voltage, a current, a power, a frequency, a distortion, a harmonic, a reactive power and / or an energy value, in particular for a phase of a multiphase voltage network (12). [3] Method according to any one of the preceding claims, characterized by , that the processor unit (16) includes an artificial intelligence (24) which receives at least the characteristic parameters that have been recorded at at least two different times as an input and outputs the future state of the voltage network (12) as an output. [4] Method according to claim 3, characterized by, that the artificial intelligence (24) includes at least one artificial neural network, for example an artificial convolutional neural network or an artificial recurrent neural network, in particular wherein the artificial intelligence includes a Long Short-Term Memory (LSTM) network or a Gated recurrent unit (GRU). [5] Method according to any one of the preceding claims, characterized by , that a multidimensional vector is generated which includes the first characteristic parameter and the second characteristic parameter, wherein the multidimensional vector is processed by the processor unit (16). [6] Method according to any one of the preceding claims, characterized by , that at least one future temporal progression of a characteristic parameter is predicted as the future state of the voltage network (12). [7] System (10) for monitoring a voltage network (12), comprising at least one processor unit (16) configured to perform a method according to any of the preceding claims. [8] Computer program (18) with program code means to perform the steps of a method according to any one of claims 1 to 6 when the computer program (18) is executed on a processor unit (16), in particular the processor unit (16) of the system (10) according to claim 7. [9] Computer-readable data carrier (20) on which the computer program (18) according to claim 8 is stored.

Images