Simulation optimized monitoring of power distribution and transmission assets

Abstract

The present method refers to a method of monitoring power distributions assets for high voltage and power transmission assets utili zing a fully transient thermal prediction, wherein the improved method very ef ficiently provides precise and reliable predictions. Additionally, the present invention refers to a computer program product including instructions operable to cause a computing entity to execute such method. Furthermore, the present invention refers to a storage device for providing such computer program product. Additionally, the present invention refers to a monitoring device containing at least one processor and at least one data storage containing the inventive computer program product to execute the inventive method.

Description

[0001] Description

[0002] Simulation optimi zed monitoring of power distribution and transmission assets

[0003] The present method refers to a method of monitoring power distribution assets for high voltage and power transmission assets utili zing an improved method . Additionally, the present invention refers to a computer program product including instructions operable to cause a computing entity to execute such method . Furthermore , the present invention refers to a storage device for providing such computer program product . Additionally, the present invention refers to a monitoring device containing at least one processor and at least one data storage containing the inventive computer program product to execute the inventive method .

[0004] Power distribution and power transmission assets are an integral part of modern society . Enabling the utili zation of electricity is required for modern li fe . Herein, transmitting such electrical power over long distances and distributing said electrical power into a local grid like a city requires highly reliable devices and systems .

[0005] While respective devices and systems are already known and in use for a long time , they are still subj ect to improvements and further developments . Herein, not only the devices and systems themselves need to be improved and adapted to new requirements like changes from the corresponding environment they are operating in . Also , their monitoring and control needs to be adapted to suit the respective needs . In the past highly skilled experts having an abundant and experienced insight into the respective devices and systems were tasked with monitoring the respective unit . However, there is a change that a respective expert needs to monitor a signi ficantly increased number of respective units , wherein the detailed insight cannot be ensured for all units to be monitored as the number of di f ferent units continuously increases . Also , it becomes increasingly di f ficult to keep track of included upgrades and the like resulting in slight di f ferences of respective units to be monitored in a plurality of units to be monitored . Resulting in the problem that the skilled person tasked with such monitoring has problems to correctly identi fy the state and possible problems at an early stage as the same behavior and data provided by two units may indicate completely di f ferent states said units are in .

[0006] This and further problems are solved by the products and methods as disclosed hereafter and in the claims . Further beneficial embodiments are disclosed in the dependent claims and the further description and figures . These benefits can be used to adapt the corresponding solution to speci fic needs or to solve additional problems .

[0007] According to one aspect the present invention refers to a method of monitoring at least one power distributions asset for high voltage and / or at least one power transmission asset for high voltage , the method contains utili zing a digital twin model of the at least one power distribution and / or the at least one transmission asset utili zing a data driven neural network to provide a fully transient thermal prediction over relevant time ranges , wherein the digital twin model simulates a transient temperature distribution of the at least one power distribution and / or the at least one transmission asset based on at least thermal di f fusion, thermal advection and thermal radiation . Utili zing such method allows to signi ficantly simpli fy the monitoring of power distribution assets and power transmission assets by an expert . Taking into account the di f ferent types of thermal heat trans fer allows to determine a state of respective assets very reliably . Allowing the expert to very early and reliable determine any problem and opportunities to make better use of respective assets as , for example , the practical range said asset can be used for is not yet reached . Herein, it was surprisingly noted that the quite demanding inclusion of the thermal radiation is very beneficial and provides an increase in the reliability of respective models signi ficantly outbalancing the ef fort required to include it in such models . Relevant time ranges as referred to herein equals no short time periods like seconds or a few minutes . Preferably, it refers to at least one hour, more preferred at least two hours , even more preferred at least four hours .

[0008] According to a further aspect the present invention refers to a computer program product , tangibly embodied in a machine- readable storage medium, including instructions operable to cause a computing entity to execute an inventive method .

[0009] According to a further aspect the present invention refers to a storage device for providing an inventive computer program product , wherein the device stores the computer program product and / or provides the computer program product for further use .

[0010] According to a further aspect the present invention refers to a monitoring device containing at least one processor and at least one data storage , wherein the at least one data storage contains an inventive computer program product , wherein the at least one processor is adapted to execute an inventive method .

[0011] The present invention allows to signi ficantly simpli fy the monitoring or decision making actively supporting an expert with the task to keep track and react reasonable while simply lacking more and more the time to review the speci fic cases in detail . Making use of the described solution allows to provide a comprehensive overview and provision of recommendations based on the data available for such power generation assets or power distribution assets enabling to utili ze such calculation engine to provide data enabling the method for such application to run smoothly and reliably . To simpli fy understanding of the present invention it is referred to the detailed description hereafter and the figures attached as well as their description . Herein, the figures are to be understood being not limiting the scope of the present invention as they merely disclosing preferred embodiments explaining the invention further .

[0012] Fig . 1 shows a scheme of a system reali zing the inventive method .

[0013] Fig . 2 shows a scheme of a cross section of an exemplarily power transmission asset being an extension module of a gas- insulated switch gear .

[0014] Preferably, the embodiments hereafter contain, unless specified otherwise , at least one processor and / or data storage unit to implement the inventive method .

[0015] Unless speci fied otherwise terms like "calculate" , "process" , "determine" , "generate" , "configure" , "reconstruct" and comparable terms refer to actions and / or processes and / or steps modi fying data and / or creating data and / or converting data, wherein the data are presented as physical variables or are available as such .

[0016] The term "data storage" or comparable terms as used herein, for example , refer to a temporary data storage like RAM (Random Access Memory) or long-term data storage like hard drives or data storage units like CDs , DVDs , USB sticks and the like . Such data storage can additionally include or be connected to a processing unit to allow a processing of the data stored on the data storage .

[0017] The term "data driven neural network" as used herein refers to a data driven neural network as known to the skilled person . Herein, such neural network has been trained with provided data . Examples of such data driven neural network are neural operator networks , recurrent neural networks (RNN) , convolutional neural networks ( CNN) , graph neural networks (GNN) and so on . An example of neural operator networks is a general neural operator trans former ( GNOT ) .

[0018] Power distribution assets are assets utili zed for distribution power in a grid like a substation attached to an overhead like distributing electrical power in a power grid of a city . For example , such power distribution assets include trans formers including instrument trans formers for power distribution applications and switchgears . For example , substations contain such power distribution assets .

[0019] Power transmission assets are assets utili zed for transmitting electrical power, for example , gas insulated switchgears , air insulated switchgears , instrument trans formers for power transmission applications , coils , bushings , disconnector, high voltage cables , overhead lines .

[0020] According to one aspect the presented invention refers to a method as described above .

[0021] Including the thermal heat trans fer via radiation is preferably done by respective modeling of the thermal radiation . According to further embodiments it is preferred that a thermal heat trans fer by means of the thermal radiation is included in the digital twin model by including connections for transmitting radiation information following the radiation rays in the digital twin model . Herein, such connections for transmitting radiation information model thermal energy emitting from a speci fied location to a speci fied other location . Utili zing such model to describe the thermal heat trans fer via radiation was surprisingly ef ficient utili zing such connections for transmitting radiation information in a respective data driven neural network .

[0022] While the time ranges the current invention refers to no short time period like at least one hour it was noted that the present invention is very beneficial for not too long time periods for many application cases . According to further embodiments it is preferred that the relevant time ranges are at most 24 hours , more preferred at most 18 hours , even more preferred at most 16 hours .

[0023] Furthermore , it was noted that certain models are very suitable for typical application cases . According to further embodiments it is preferred that results of computational fluid dynamics ( CFD) simulations are included in a training of the digital twin model to utili zed to provide the fully transient thermal prediction . Such simulations are typically providing very good descriptions of the considered application cases taking into account the processing power required resulting in an overall very well combination of processing speed even for complex systems and reliable simulations .

[0024] However, to increase the processing speed it is typically preferred to utili ze computational fluid dynamics simulations to train the data driven neural network . According to further embodiments it is preferred that the data driven neural network has been trained with computational fluid dynamics simulation data . It was surprisingly noted that the data driven network is able to provide a slightly less accurate simulation, however, simultaneously provides a signi ficantly higher processing speed enabling a very beneficial application for power distribution assets and power transmission assets . As it was noted that the synergistic ef fect acquired by such trans fer allows to provide a very ef ficient solution taking care of the needs of respective operators tasked with monitoring such assets .

[0025] To define a starting point for the simulation it is preferred to utili ze a initial state describing the state of the considered assets at a starting point . According to further embodiments it is preferred that the method contains the digital twin model receiving a initial state , wherein the data driven neural network utili zes the initial state to simulate a state of the at least one power distributions asset for high voltage and / or the at least one power transmission asset for high voltage . Providing such speci fied state instead of , for example , sensor data to be processed to let the digital twin model deduct an initial state is surprisingly beneficial to speed up the processing .

[0026] However, alternatively or additionally, it can be preferred to provide the creation of a initial state by the digital twin model . According to further embodiments it is preferred that the method contains the digital twin model receiving sensor data related to the at least one power distributions asset for high voltage and / or the at least one power transmission asset for high voltage , wherein the sensor data is processed by the digital twin model to automatically identi fy a current state of the at least one power distributions asset for high voltage and / or the at least one power transmission asset for high voltage simulated by the digital twin model , wherein the digital twin model automatically creates a initial state to be utili zed by the data driven neural network , wherein the data driven neural network utili zes the initial state to determine a current state of the at least one power distributions asset for high voltage and / or the at least one power transmission asset for high voltage at least taking into account the initial state , wherein the fully transient thermal prediction of the at least one power distributions asset for high voltage and / or the at least one power transmission asset for high voltage of the digital twin model is based on the current state . While such embodiments require an increased complexity and more processing power for the digital twin model it was noted that for typical application such automatic provision of the initial state based on sensor data allows to simpli fy the work of the expert monitoring the assets being under real working conditions typically being less experienced with such task and benefitting signi ficantly by such a solution . Additionally, it was noted that it is beneficial for typical applications to include an automatic correction mechanism . According to further embodiments it is preferred that the digital twin model receives sensor data related to the at least one power distributions asset for high voltage and / or the at least one power transmission asset for high voltage , wherein the sensor data is processed by the digital twin model to automatically identi fy deviations of a current state of the at least one power distributions asset and / or the at least one power transmission asset from the fully transient thermal prediction of the at least one power distributions asset for high voltage and / or the at least one power transmission asset for high voltage simulated, wherein the digital twin model automatically adapts the fully transient thermal prediction to reduce the deviation of the fully transient thermal prediction from the current state of the at least one power distributions asset and / or the at least one power transmission asset . Including such solution also increases the complexity and processing power required for the digital twin model . However, it was noted that such solution allows to automatically create alerts and directives for the expert tasked with monitoring at a very early stage . It was noted that respective deviations indicate possible problems at an early stage and respective information provided in this context provide a signi ficant benefit for the skilled person tasked with monitoring the respective assets . Also , it was noted that under real li fe conditions a corresponding feedback loop can be every beneficial as tests showed that some unexpected actions during use like turning of f respective devices containing the digital twin model and resetting such device to a saved condition after some interruption or the like results in deviations that are to be kept track of . Respective feedback to the expert and requesting his veri fication that a signi ficant deviation from the known state took place provides a surprisingly easy yet beneficial solution to take care of such problem . A typically very beneficial application case of the inventive method are high voltage cables . According to further embodiments it is preferred that the at least one power transmission asset is a high voltage cable , preferably wherein the high voltage cable is an overhead line high voltage cable . It was noted that the fully transient thermal prediction acquired according to the inventive method allow to very ef ficiently determine chances to improve the use of the respective cables . Like determining whether under the speci fic conditions it is possible to increase the electric power transmitted through some overhead line .

[0027] A further typically very beneficial application case of the inventive method are power trans formers and high voltage switchgears . According to further embodiments it is preferred that the at least one power distribution asset is a power trans former and / or a high voltage switchgear . In this context , it was noted that an expert monitoring said assets is and will be even more often confronted with the task to assess whether the assets can be utili zed to increase the power intake , for example , originating from renewable energy sources like wind power and solar power providing a changing input into a power grid . Deciding whether the respective power does not increase the burden on the assets too much potentially even threatening said assets and their operation .

[0028] In general , it was noted that it is beneficial to define certain boundary conditions and adapt the inventive method accordingly . According to further embodiments it is preferred that the fully transient thermal prediction of a single asset requires a processing time of less than 10 seconds , more preferred less than 7 seconds , even more preferred less than 5 seconds , to provide a prediction of the at least one power distribution asset and / or the at least one power transmission asset . Under real li fe conditions such speed of prediction enables the skilled person to make fast responses too typical requirements surprisingly rendering the system far better even in case the preciseness is slightly decreased . Comparable to the provision of the fully transient thermal prediction of a single asset it was noted that the fully transient thermal prediction of an asset system to be restricted for special application . According to further embodiments it is preferred that the fully transient thermal prediction of an asset system requires a processing time of less than 2 minutes , more preferred less than 90 seconds , even more preferred less than 80 seconds , to provide a prediction of the at least one power distribution asset and / or the at least one power transmission asset . Also , in this case it was noted that even accepting a slight loss of precision can be preferable to providing the corresponding prediction . Such asset system is , for example , a group of assets at a speci fic location or a group of assets of a speci fic type of power distribution assets or power transmission assets .

[0029] Furthermore , it was noted that the data driven neural network can be beneficially located in a cloud for speci fic applications . According to further embodiments it is preferred that the data driven neural network is located in a cloud . Allowing to make use of a higher amount of processing power and especially in combination with many embodiments as disclosed herein providing even more benefits . Like in combination with the time limitations above ensuring a timely prediction while reducing the loss of precision .

[0030] Alternatively, the data driven neural network can be provided locally . According to further embodiments it is preferred that the data driven neural network is located on a local device . It was noted that taking into account the requirements of typical embodiments using such embodiment is preferred . Especially, in case the monitoring of critical assets is included requiring to instantly react in case of problems . Rendering any delay resulting from a loss of connectivity problematic . The results observed showed that the inventive method does not only allow to monitor single or small numbers of assets . According to further embodiments it is preferred that the at least one power distributions asset for high voltage is a plurality of power distribution assets , more preferred wherein the at least one power distributions asset for high voltage is a fleet of power distribution assets . Despite the increasing complexity and amount of processing power required it was noted that the ef fort required is signi ficantly outbalanced by the simpli fication of the monitoring of such group of assets . Enabling a signi ficant leap of the number of assets being able to be reliable monitored even by a single expert .

[0031] The same applies for power transmission assets . According to further embodiments it is preferred that the at least one power transmission asset is a plurality of power transmission assets , more preferred wherein the at least one power transmission asset is a fleet of power transmission assets . According to further embodiments it is preferred that the inventive method is a method of monitoring a plurality of power distribution assets for high voltage and / or a plurality of power transmission assets for high voltage , more preferred a method of monitoring a fleet of power distribution assets for high voltage and / or a fleet of power transmission assets for high voltage .

[0032] The term "plurality" as used herein refers to as least 5 assets , more preferred at least 7 assets .

[0033] The term " fleet" as used herein refers to at least at least 10 assets , more preferred at least 20 assets .

[0034] According to further embodiments it is preferred that the data driven neural network creates an output , wherein the output is forwarded to a user interface and / or a database , preferably a user interface . Especially, utili zing a user interface receiving such output is beneficial for typ- ical applications . Allowing to make good use of the speed the prediction is provided and to react in time in case an action is required .

[0035] However, it was further noted that the high speed and preciseness of the inventive method can be utili zed to automi ze a respective action . According to further embodiments it is preferred that the data driven neural network creates an output , wherein the output automatically triggers a control action, wherein the control action influences an operation of at least one of the at least one power distribution asset and / or the at least one power transmission asset . For example , such action can be a control action to decide whether further power can be received and processed by the respective power distribution asset or power transmission asset . Also , such action can be to power down a respective asset in case the prediction indicates a dangerous overheating endangering the operation or even safety of the asset .

[0036] It was quite easy to provide speci fically adapted data driven neural network of each asset . However, it was noted that even typical di f ferences observed for assets of the same product type typically do not require such adaption . According to further embodiments it is preferred that the same data driven neural network is utili zed for the fully transient thermal prediction of the at least one power distributions asset for high voltage and / or the at least one power transmission asset for high voltage providing a similar geometry . Especially, it is preferred that the data driven neural network can be utili zed for di f ferent assets of the same product type . Such similar geometry is , for example , provided to assets of the same product type . Like 8VN1 Blue GIS as an example of a product type of a switchgear as produced by Siemens Energy . Further examples for such switchgears are 8DN8 GIS or 8DN9 GIS or 8DQ1 GIS also product and sold by Siemens Energy . Examples of product types of circuit breakers are 3AV1 Blue DT or 3AP1 EG DT or 3AP1 FI DT or 3AP2 / 3 FI DT as provided by Siemens Energy. It was noted that the digital twin model as described herein are able to generalize the geometry of respective assets within the application field as specified in the inventive method. Allowing the fully transient thermal prediction without retraining. Significantly simplifying the inventive method.

[0037] According to a further aspect the present invention refers to a computer program product, tangibly embodied in a machine- readable storage medium, including instructions operable to cause a computing entity to execute an inventive method.

[0038] According to a further aspect the present invention refers to a storage device for providing an inventive computer program product, wherein the device stores the computer program product and / or provides the computer program product for further use .

[0039] According to a further aspect the present invention refers to a monitoring device containing at least one processor and at least one data storage, wherein the at least one data storage contains an inventive computer program product, wherein the at least one processor is adapted to execute an inventive method.

[0040] The following detailed description of the figure uses the figure to discuss illustrative embodiments, which are not to be construed as restrictive, along with the features and further advantages thereof.

[0041] Figure 1 shows a scheme of a system realizing the inventive method. To simplify the figure, only a small number of the plurality of the power distribution assets 4, 4' , 4' ’ 4, 4' , 4' ’ for high voltage and of the plurality of the power transmission assets 3, 3' , 3' ’ for high voltage are shown. Said power transmission assets 3, 3' , 3' ’ contain high voltage cables of overhead lines. Said power distribution assets 4, 4' , 4' ’ contain power transformers. Said power distribution as- sets 4, 4' , 4' ' and power transmission assets 3, 3' , 3' ' provide sensor data being transmitted to the monitoring device 2 containing a computer program product. While it is not shown in this figure, the digital twin model 5 can also be located in a cloud. The computer program product again contains the digital twin model 5 containing the data driven neural network 1. Said data driven neural network is utilized to provide a fully transient thermal prediction over relevant time ranges of the power distribution assets 4, 4' , 4' ’ and power transmission assets 3, 3' , 3' ’ .

[0042] The digital twin model 5 simulates the temperature distribution of the power distribution assets 4, 4' , 4' ’ and the power transmission assets 3, 3' , 3' ’ based on thermal diffusion, thermal advection and thermal radiation utilizing the data driven neural network 1. Herein, the thermal radiation is included in the digital twin model by connections for transmitting radiation information. To start the simulation an initial state is provided to the digital twin model 5, wherein the data driven neural network 1 utilizes said initial state to simulate the current state of the power distribution assets 4, 4' , 4' ’ and the power transmission assets 3, 3' , 3' ’ .

[0043] During a continuous provision of the fully transient thermal prediction digital twin model 5 receives sensor data related to of the power distribution assets 4, 4' , 4' ’ and the power transmission assets 3, 3' , 3' ’ . Said sensor data is processed by the digital twin model 5 to automatically identify a current state of the at least one power distribution assets 4, 4' , 4' ’ and the power transmission assets 3, 3' , 3' ’ . Said current state is utilized to further optimize and adapt the fully transient thermal prediction. Additionally, the sensor data is processed by the digital twin model 5 to automatically identify deviations of a current state of the power distribution assets 4, 4' , 4' ’ and the power transmission assets 3, 3' , 3' ’ from the fully transient thermal prediction. The digital twin model 5 automatically adapts the fully transient thermal prediction to reduce the deviation of the fully tran- sient thermal prediction from the current state of the power distribution assets 4, 4' , 4' ’ and the power transmission assets 3, 3' , 3' ' .

[0044] Despite the additional inclusion of the sensor data as specified above the fully transient thermal prediction of a single asset requires a processing time of less than 7 seconds. Additionally, the fully transient thermal prediction of an asset system requires a processing time of less than 2 minutes. Allowing the expert tasked with monitoring the power distribution assets 4, 4' , 4' ’ and the power transmission assets 3, 3' , 3' ’ to review said fully transient thermal prediction of the asset system regularly to identify developing problems. The short time to provide the fully transient thermal prediction of a single asset on short notice to review a specific case and identify the specific problem.

[0045] Herein, the expert enters a corresponding request in a user interface 7 to trigger the generation of an output containing the fully transient thermal prediction. The output created by the data driven neural network 1 is forwarded again to the user interface 7 to be reviewed. However, said output is in addition forwarded to a database 8 to be stored enabling to be reviewed at a later point. Such reviewing can be beneficially utilized thereafter, for example, by respective experts for training and optimize their actions.

[0046] Additionally, the output generated by the data driven neural network 1 is able to trigger a control action. Said automatically generated control action influences the operation of the power distribution assets 4, 4' , 4' ’ and the power transmission assets 3, 3' , 3' ’ . For example, triggering an alert to focus the attention of the expert monitoring the power distribution assets 4, 4' , 4' ’ and the power transmission assets 3, 3' , 3' ’ on a specific topic or asset. Said potentially triggered actions even include shutting down power distributions asset 4, 4' , 4' ’ or power transmission asset 3, 3' , 3' ’ . Figure 2 shows a scheme of a cross section of an exemplarily power transmission asset being an extension module 24 of a gas-insulated switch gear . Herein, the extension module 24 comprises of an extension module hull 21 being its outer casing which is filled with gas and contains three conductors 22 extending though the extension module 24 transmitting the electricity . The figure further indicates the thermal radiation utili zed in the simulation of the temperature distribution . Said thermal radiation is indicated by means of lines representing the rays of thermal energy being transmitted that are translated into connections for transmitting radiation information 23 between the conductors 22 and the extension module hull 21 to be utili zed by the data driven neural network . It needs to be noted that respective rays of thermal energy between di f ferent conductors and from extension module hull 21 to a di f ferent part of the extension module hull 21 are not shown in Figure 2 for not overloading the figure .

[0047] The present invention was only described in further detail for explanatory purposes . However, the invention is not to be understood being limited to these embodiments as they represent embodiments providing benefits to solve speci fic problems or ful filling speci fic needs . The scope of the protection should be understood to be only limited by the claims attached .

Claims

Patent claims1 . Method of monitoring at least one power distributions asset for high voltage and / or at least one power transmission asset for high voltage , the method contains utili zing a digital twin model of the at least one power distribution and / or the at least one transmission asset utili zing a data driven neural network to provide a fully transient thermal prediction over relevant time ranges , wherein the digital twin model simulates a temperature distribution of the at least one power distribution and / or the at least one transmission asset based on at least thermal di f fusion, thermal advection and thermal radiation .2 . Method according to claim 1 , wherein a thermal heat trans fer by means of the thermal radiation is included in the digital twin model by connections for transmitting radiation information in the digital twin model .3 . Method according to any of the preceding claims , wherein the method contains the digital twin model receiving an initial state , wherein the data driven neural network utili zes the initial state to simulate a state of the at least one power distributions asset for high voltage and / or the at least one power transmission asset for high voltage .4 . Method according to any of the preceding claims , wherein the method contains the digital twin model receiving sensor data related to the at least one power distributions asset for high voltage and / or the at least one power transmission asset for high voltage , wherein the sensor data is processed by the digital twin model to automatically identi fy a current state of the at least one power distributions asset for high voltage and / or the at least one power transmission asset for high voltage simulatedby the digital twin model , wherein the digital twin model automatically creates a initial state to be utili zed by the data driven neural network , wherein the data driven neural network utili zes the initial state to determine a current state of the at least one power distributions asset for high voltage and / or the at least one power transmission asset for high voltage at least taking into account the initial state , wherein the fully transient thermal prediction of the at least one power distributions asset for high voltage and / or the at least one power transmission asset for high voltage of the digital twin model is based on the current state .

5. Method according to any of the preceding claims , wherein the method contains that the digital twin model receives sensor data related to the at least one power distributions asset for high voltage and / or the at least one power transmission asset for high voltage , wherein the sensor data is processed by the digital twin model to automatically identi fy deviations of a current state of the at least one power distributions asset and / or the at least one power transmission asset from the fully transient thermal prediction of the at least one power distributions asset for high voltage and / or the at least one power transmission asset for high voltage simulated, wherein the digital twin model automatically adapts the fully transient thermal prediction to reduce the deviation of the fully transient thermal prediction from the current state of the at least one power distributions asset and / or the at least one power transmission asset .

6. Method according to any of the preceding claims , wherein the at least one power transmission asset is a high voltage cable , preferably wherein the high voltage cable is an overhead line high voltage cable .7 . Method according to any of the preceding claims , wherein the at least one power distribution asset is a power transformer and / or a high voltage switchgears .8 . Method according to any of the preceding claims , wherein the fully transient thermal prediction of a single asset requires a processing time of less than 10 seconds , more preferred less than 7 seconds , even more preferred less than 5 seconds , to provide a prediction of the at least one power distribution asset and / or the at least one power transmission asset .

9. Method according to any of the preceding claims , wherein the fully transient thermal prediction of an asset system requires a processing time of less than 2 minutes , more preferred less than 90 seconds , even more preferred less than 80 seconds , to provide a prediction of the at least one power distribution asset and / or the at least one power transmission asset .10 . Method according to any of the preceding claims , wherein the at least one power distributions asset for high voltage is a plurality of power distribution assets and wherein the at least one power transmission asset is a plurality of power transmission assets .11 . Method according to any of the preceding claims , wherein the data driven neural network creates an output , wherein the output automatically triggers a control action, wherein the control action influences an operation of at least one of the at least one power distribution asset and / or the at least one power transmission asset .12 . Method according to any of the preceding claims , wherein the data driven neural network can be utili zed for di f ferent assets of the same product type .13 . Computer program product , tangibly embodied in a machine-readable storage medium, including instructions operable to cause a computing entity to execute a method according to any of claims 1 to 12 .14 . Storage device for providing a computer program product according to claim 13 , wherein the device stores the computer program product and / or provides the computer program product for further use .15 . Monitoring device containing at least one processor and at least one data storage , wherein the at least one data storage contains a computer program product according to claim 13 , wherein the at least one processor is adapted to execute a method according to any of claims 1 to 12 .

Images