EP4721233A1 - System and method for the enhanced control of an electrical network - Google Patents
System and method for the enhanced control of an electrical networkInfo
- Publication number
- EP4721233A1 EP4721233A1 EP24733032.7A EP24733032A EP4721233A1 EP 4721233 A1 EP4721233 A1 EP 4721233A1 EP 24733032 A EP24733032 A EP 24733032A EP 4721233 A1 EP4721233 A1 EP 4721233A1
- Authority
- EP
- European Patent Office
- Prior art keywords
- maintenance
- outage
- components
- intervention
- maintenance intervention
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Classifications
-
- H—ELECTRICITY
- H02—GENERATION; CONVERSION OR DISTRIBUTION OF ELECTRIC POWER
- H02J—ELECTRIC POWER NETWORKS; CIRCUIT ARRANGEMENTS OR SYSTEMS FOR SUPPLYING OR DISTRIBUTING ELECTRIC POWER; SYSTEMS FOR STORING ELECTRIC ENERGY
- H02J13/00—Circuit arrangements for providing remote monitoring or remote control of equipment in a power distribution network
- H02J13/12—Monitoring network conditions, e.g. electrical magnitudes or operational status
-
- H—ELECTRICITY
- H02—GENERATION; CONVERSION OR DISTRIBUTION OF ELECTRIC POWER
- H02J—ELECTRIC POWER NETWORKS; CIRCUIT ARRANGEMENTS OR SYSTEMS FOR SUPPLYING OR DISTRIBUTING ELECTRIC POWER; SYSTEMS FOR STORING ELECTRIC ENERGY
- H02J2103/00—Details of circuit arrangements for mains or AC distribution networks
- H02J2103/30—Simulating, planning, modelling, reliability check or computer assisted design [CAD] of electric power networks
Landscapes
- Engineering & Computer Science (AREA)
- Power Engineering (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
- Power Sources (AREA)
- Supply And Distribution Of Alternating Current (AREA)
Abstract
Computer-implemented system for the control of a power transmission and/or distribution system, wherein the power transmission and/or distribution system includes a plurality of components that include a plurality of nodes and a plurality of line segments connecting the plurality of nodes. The computer- implemented system comprises a controller that includes : an impact determination module configured to determine an impact of an outage of each of the components of the power transmission and/ or distribution system, and a maintenance intervention determination module configured to determine a maintenance intervention or that no maintenance is required for one or more, or all, of the plurality of components. The computer-implemented system further comprises an artificial intelligence (Al ) model of an outage of the power transmission and/or distribution system configured to provide as output an outage probability for each of the plurality of components based on an input data set for each of the plurality of components. The controller further includes an outage prediction module configured to generate an outage probability for each of the plurality of components by feeding the respective input data set to the AI model and obtaining the output from the AI model. The maintenance intervention determination module is configured to determine a maintenance intervention for a given component using a first process using the AI model by : modifying an input relating to the maintenance included in the input data set for the given component; determining an updated outage probability based on the modified input by using the Al model of the outage of the power transmission and/ or distribution system; determining a difference between the updated outage probability and the outage probability as previously determined for the given component; and determining a maintenance intervention related to the modified input if the determined difference is above a predetermined threshold.
Description
SYSTEM AND METHOD FOR THE ENHANCED CONTROL OF AN ELECTRICAL NETWORK
Cross-Reference to Related Applications
This Patent Application claims priority from Italian Patent Application No . 102023000011319 filed on June 5 , 2023 , the entire disclosure of which is incorporated herein by reference .
Field of the Invention
Example aspects herein relate to the monitoring and the control of electrical power systems , in particular electrical power distribution and/or transmission systems , including e . g . medium and/or low voltage, and more speci fically to a computer- implemented system and a method for control and monitoring of the electrical power systems .
Background
Everyday electricity distribution companies are committed to solving grid outages in the shortest possible time , to ensure electrical continuity for their customers and to avoid the penalties imposed by the regulatory authority for long systems downtime .
To this end, electricity distribution companies or Distribution System Operators , DSOs , periodically plan a series of long- and short-time interventions to reduce network critical ities , prevent failures , and guarantee power continuity to its customers . Nowadays this is managed by a professional team of network planners , helped by a tool that defines the performance of the
electrical grid through speci fic QoS indicators at line granularity . However, due to the complex structure of the grids and the several involved factors , identi fying the interventions with the highest benefits for the network and its customers is a challenging task .
There is therefore a need to improve the monitoring and the control of electric power system and of the components therein .
In this context , CN 106 384 210 A discloses a power transmission and trans formation equipment maintenance priority ordering method based on a maintenance risk premium . The method comprises the steps of acquiring information of power transmission and trans formation equipment , collecting and analyzing environment information, system information and equipment information of equipment required in an equipment risk source identi fication step and an equipment risk evaluation step; identi fying a power transmission and trans formation equipment risk source , performing identi fication and classi fied quanti zation on uncertain factors and safety hazards in the equipment by means of the information which is acquired in the information acquisition step , establishing an equipment operation stopping model ; performing equipment fault result severity evaluation, performing topology analysis and trend calculation on a power grid according to an equipment operation stopping condition and fault probability of each set of operation equipment , simulating a power grid operation scene , and evaluating comprehensive result severity of the power grid after
the fault of the equipment ; and performing equipment maintenance ordering based on the maintenance risk premium for setting an equipment maintenance priority ordering problem on the condition of limited maintenance resource , and supplying theoretical and technological supports for maintenance arrangement .
Summary of Invention
According to a first example aspect herein, there is provided a computer-implemented system for the control of a power transmission and/or distribution system, wherein the power transmission and/or distribution system includes a plurality of components that include a plurality of nodes and a plurality of line segments connecting the plurality of nodes .
The computer-implemented system comprises a controller that includes :
• an impact determination module configured to determine an impact of an outage of each of the components of the power transmission and/or distribution system; and
• a maintenance intervention determination module configured to determine a maintenance intervention or that no maintenance is required for one or more , or all , of the plurality of components .
The computer-implemented system further comprises an arti ficial intelligence (Al ) model of an outage of the power transmission and/or distribution system configured to provide as output an outage probability for each of
the plurality of components based on an input data set for each of the plurality of components .
The controller further includes an outage prediction module configured to generate an outage probability for each of the plurality of components by feeding the respective input data set to the Al model and obtaining the output from the Al model .
The maintenance intervention determination module is configured to determine a maintenance intervention for a given component using a first process using the Al model by :
• modi fying an input relating to the maintenance included in the input data set for the given component ;
• determining an updated outage probability based on the modi fied input by using the Al model of the outage of the power transmission and/or distribution system;
• determining a di f ference between the updated outage probability and the outage probability as previously determined for the given component ;
• determining a maintenance intervention related to the modi fied input i f the determined di f ference is above a predetermined threshold .
According to a further aspect , the impact determination module is configured to determine an impact of the outage based on a determini stic algorithm associated to each of the components , wherein the output of the deterministic algorithm includes at least one of : a number of af fected customers per interruption time and/or a number of a f f ected customers per interruption
time divided by the number of the customers of the
Distribution System Operator, DSO . further aspect , the maintenance
ination module is configured to determine a level of priority of a maintenance intervention for a component based on the outage probability determined by the outage prediction module for that component and/or based on the impact determined by the impact determination module for that component , preferably based on a coef ficient obtained by multiplying the outage probability and an index associated to the outage impact .
According to a further aspect , the maintenance intervention determination module is configured to determine a maintenance intervention for a given component following a f irst process using the Al model of the outage of the power transmission and/or distribution system by :
-modi fying an input relating to the maintenance included in the input data set for the given component , -determining an updated outage probabi lity based on the modi fied input by using the Al model of the outage of the power transmission and/or distribution system, -determining a di f ference between the updated outage probabil ity and the outage probabi lity as previously determined for the given component ,
-determining a maintenance intervention related to the modi fied input i f the determined di f ference is above a predetermined threshold .
According to a further aspect , the maintenance intervention module is configured to use the first process i f the component is a line segment , wherein the input relating to maintenance that is modi fied is related to a replacement action of the conductors/cables of a line segment and wherein the input relating to maintenance that is modi fied is fed to the Am model of the outage of the line segment to determine the updated outage probability .
According to a further aspect , the maintenance intervention determination module is configured to determine a maintenance intervention using a second process by using a heurist ic, wherein the heuristic includes for each type of component : ( i ) first historical data related to the past maintenance of the components of a speci fic type of component in a first period of time and ( ii ) second historical data related to the past occurrence of an outage for the components of the speci fic type in a second period of time subsequent to the first period of time , wherein the first historical data and the second historical data are stored in association with the first historical data for each component , wherein a first outage rate is calculated based on the first and second historical data for components that had a maintenance intervention in the first period of time and a second outage rate is calculated based on the first and second historical data for components that had no maintenance intervention in the first period of time,
wherein the maintenance intervention determination module is configured to determine a maintenance intervention based on the first and second outage occurrence rates .
According to a further aspect , the first historical data refers to anyone of :
- for overhead conductors : pruning, plant maintenance , replacement of accessories other than the cable / c o ndu ctor,
- for overhead cables : plant maintenance , replacement of accessories other than the cable/conductor ,
- for underground cables : plant maintenance,
- for primary substations : structural maintenance , preventive measures against animals , plant maintenance .
According to a further aspect , the intervention determination module is configured to determine a maintenance intervention using the second process using the heurist ic in case the component is a node of the power transmission and/or distribution system .
According to a further aspect , the intervention determination module is configured to determine a maintenance intervention using the second process using the heuristic i f the component is a line segment of the power transmission and/or distribution system, and the maintenance intervention is not a replacement action related to the conductors/cables of the line segment .
According to a further aspect , the intervention determination module is configured to determine a maintenance intervention using the second process using
the heuristic i f the component is a node of the power transmission and/or distribution system, and wherein, i f the component is a line segment :
-the intervention determination module is configured to determine a candidate maintenance intervention as the maintenance intervention using the second process when the candidate maintenance intervention is not a replacement of a cable/conductor , and
-the intervention determination module is configured to determine a candidate maintenance intervention as the maintenance intervention using the first process when the candidate maintenance intervention is a replacement of a cable/conductor .
According to a further example aspect herein, there is provided computer-implemented system for the control of a power transmission and/or distribution system, wherein the power transmission and/or distribution system includes a plurality of components , wherein the components include a plurality of nodes and a plurality of line segments connecting the plurality of nodes , the computer-implemented system comprising :
- a model of an outage of the power transmission and/or distribution system configured to provide as output an outage probability for each of the plurality of components based on an input data set for each o f the plurality of components ;
-a controller includes an outage prediction module configured to generate an outage probability for each of the plurality of components by feeding the respective
input data set to the model and obtaining the output from the model.
According to an example, the model includes a first model of the outage of the nodes of the power transmission and/or distribution system and a second model of the outage the line segments of the power transmission and/or distribution system.
In an example, the first model and/or the second model is/are a model based on artificial intelligence, Al, preferably a machine learning, ML, model.
In a further example, the first model and/or the second model is/are a ML model trained in a supervised manner .
In an example, the ML model is a classifier, preferably a xgboost classifier.
In a further example, the plurality of components includes a plurality of nodes that includes at least: primary substations, secondary substations, pole-mounted substations, and switchgears.
In an example, the plurality of components includes a plurality of line segments that includes at least: underground cables, overhead conductors, and overhead cables .
In an example, the power transmission and/or distribution system is a medium voltage, MV, power distribution system.
According to a further example, the input data set for each of the plurality of components includes one or more inputs related to a maintenance.
According to a further example , the one or more inputs related to a maintenance includes one or more of : a number of maintenance activities applied to the component , number of days from the last maintenance activity, plant maintenance , remote control maintenance , replacement , structural maintenance , pruning, and preventive measures against animals that could af fect the power transmission and/or distribution system.
According to a further example , the input data set for each of the plurality of components includes one or more inputs related to a topology .
According to a further example , the one or more inputs related to a topology includes one or more of : geo-coordinates , line of origin, customers connected to the line to which the component belongs , component type , organi zational unit of origin, construction type , conductor length, section and material , edge of origin and the number of line segments of that edge , short-circuit current , maximum operating current .
According to a further example , the input data set for each of the plurality of components includes one or more inputs related to past outages .
According to a further example , the one or more inputs related to past outages includes one or more of : number of outages incurred by the component , number of days from the last outage , wherein each the number of outages and of number of days is determined for at least each of the following outage
causes : weather events , external causes , mechanical failure , power outage , plant or tree collision .
According to a further example , the input data set for each of the plurality of components includes one or more inputs related to one or more weather indexes .
According to a further example , the weather indexes are based on one or more of : wind, temperature and precipitations .
According to a further example , the input data set for each of the plurality of components includes one or more inputs related to vegetation and/or soil indexes .
According to a further example , the vegetation and/or soil indexes are based on one or more of : vegetation density, leaf type , and soil type .
According to a further example , the input data set for each of the plurality of components includes one or more inputs related to population and territory information .
According to a further example , the one or more inputs related to population and territory information include population density, municipality type ( littoral or insular ) , and altitude .
According to a further example , samples of the input data set are collected periodically over a first time interval to create a set of training data .
According to a further example , the outage probability for a component refers to the probability of an outage of the component in a future time interval , wherein the future time interval is larger than the
period at which the samples of the input data set are collected periodically .
According to a further example , the outage probability for a component refers to the probability of an outage of the component in a future time interval , wherein the future time interval is smaller than the first time interval during which the samples of the input data set are periodically collected .
According to a further example , the outage probability is determined for controlling the power transmission and/or distribution system .
According to a further example , the computer- implemented system is configured to control the power transmission and/or distribution system based on the determined outage probability .
The present disclosure exploits arti ficial intelligence (Al ) models and deterministic algorithms to estimate the outage probability of the network elements , determine their impacts in terms of duration and af fected customers , and finally suggest the best network interventions to minimi ze the outage probability . Speci fically, the present disclosure provides for a novel system and method for planning interventions in a power distribution/ transmission system with the following new features :
-indicators predicting the network' s long-term performance at network element granularity through rule- and machine-learning-based techniques ;
-the accurate detection of the outage probability for each network element of the electrical grid;
-recommendations related to maintenance interventions or substitutions that, looking at the historic behavior of the network and the network composition itself, are proven to reduce the outage probability .
Accordingly, it is possible to predict grid outages in advance, understand their impact and prevent them through targeted network interventions, thereby improving the monitoring and the control of the electrical power system.
In the first aspect of the present disclosure, the system detects the network elements that are likely to experience an outage. More specifically, a machine learning model predicts the outage probability of the network elements in a certain time horizon. The peculiarity of this mode.1 is that it exploits not only endogenous features, like the intrinsic characteristics of a network element or its previous outages and maintenance activities but also the exogenous ones, like the information on demography, weather, and vegetation related to its surrounding area. The inventors considered the weather and vegetation features because these factors have been found to affect the outage rate of the electrical grids.
In a second aspect of the present disclosure, a deterministic algorithm estimates the impact of an outage based on a network performance indicator, called SAIDI (described in the publication F. Amadei, E.Valigi, C. D'Adamo, "Enel's way to SAIDI", Proceedings CIRED 2019) , which considers the number of affected customers
and the duration of the outage . This algorithm mimics the sequence of all the operations performed on a power line to find and fix an outage .
In a third aspect of the present disclosure, it has been developed an algorithm that recommends maintenance actions or the replacement of network elements to reduce the outage probability . This algorithm estimates the outage probabi l ity reduction by exploiting both heuristics based on the maintenance and outage history of the network elements and the machine learning models defined in the first aspect .
According to a further aspect , the present disclosure provides a computer-implemented method for controlling of a power transmission and/or distribution system, wherein the power transmission and/or distribution system includes a plurality of components , wherein the components include a plurality of nodes and a plurality of line segments connecting the plurality of nodes , the computer-implemented method comprising : providing a model of an outage of the power transmission and/or distribution system configured to provide as output an outage probability for each of the plurality of components based on an input data set for each of the plurality of components ;
- generating an outage probability for each of the plurality of components by feeding the respective input data set to the model and obtaining the output from the model .
According to a further aspect , the present disclosure provides a computer-implemented method for
controlling of a power transmission and/or distribution system, wherein the power transmission and/or distribution system includes a plurality of components , wherein the components include a plurality of nodes and a plurality of line segments connecting the plurality of nodes , the computer-implemented method comprising determining an impact of the outage of each of the components of the power transmission and/or distribution system .
According to a further aspect , the determining an impact of the outage is based on a deterministic algorithm associated to each of the components , wherein the output of the deterministic algorithm includes at least one of : a number of af fected customers per interruption time and/or a number of af fected customers per interruption time divided by the number of the customers of the Distribution System Operator, DSO .
According to a further aspect , the computer- implemented method further comprises determining a maintenance intervention or that no maintenance is required for one or more , or all , of the plurality of components .
According to a further aspect , the computer- implemented method further comprises providing an arti ficial intelligence , Al , model of an outage of the power transmission and/or distribution system configured to provide as output an outage probability for each of the plurality of components based on an input data set for each of the plurality of components ;
wherein the method further includes generating an outage probability for each of the plurality of components by feeding the respective input data set to the model and obtaining the output from the model .
According to a further example aspect herein, there is provided a computer program comprising instructions which, when executed by one or more processor, cause the one or more processors to perform any of the methods of as summari zed above .
List of Figures
Embodiments of the present invention, which are presented for better understanding the inventive concepts , but which are not to be seen as limiting the invention, will now be described with reference to the figures in which :
Figure 1 is a schematic diagram showing an example of a power distribution and transmission system in example embodiments ;
Figure 2 is a schematic diagram showing an example of a power distribution system;
Figure 3 is a schematic diagram showing an example of a computer-implemented system in example embodiments ;
Figure 4 is a schematic diagram showing an example of an outage prediction model in example embodiments ;
Figure 5 shows a schematic diagram showing an example of a maintenance intervention determination module in example embodiments ;
Figure 6 shows a schematic diagram showing an example of an interaction between the modules of the computer- implemented system;
Figure 7 is a diagram showing the application of the heuristic for the determination of the maintenance intervention;
Figure 8 shows a dashboard displaying the outage probability and the impact for a plurality of components ;
Figure 9 shows a display screen that outputs a maintenance intervention with other related data .
Detailed Description
Although example embodiments will be described below, it will be evident that various modi fications may be made to these example embodiments without departing from the broader spirit and scope of the invention . Accordingly, the following description and the accompanying drawings are to be regarded as illustrative rather than restrictive .
In the following description and in the accompanying figures , numerous details are set forth in order to provide an understanding of various example embodiments . However, it will be evident to those skilled in the art that embodiments may be practiced without these details .
Figure 1 is a schematic diagram showing a power transmission and distribution system 10 ; in the present disclosure , with the term "power" is meant "electrical power" , that may also be referred to as "electricity" , "electrical energy" or the like . Accordingly, a power transmission and/or distribution system can be referred
to as electrical power transmission and/or distribution system . The power transmission and distribution system
10 may include a high voltage power transmission system
11 and medium voltage power distribution system 12 and low voltage power distribution system 15 . The power transmission system 11 may operate at high voltage , e . g . 115 kVAC to 230 kVAC, or any voltage above 69kVAC . Electrical energy in a medium voltage electrical network, such as the medium voltage power distribution system 12 , usually is distributed at a voltage in the range of 2 . 4 kVAC to 69kVAC . Electrical energy in the medium voltage power distribution system 12 is trans formed to low voltage electrical and provided to a low voltage electrical network, such as the low voltage power distribution system 15 , for distribution to users . The voltage in the low voltage power distribution system 15 is usually below 2 . 4 kVAC, most commonly below I kVAC . For example , in Italy the low voltage is normally up to I kVAC, the medium voltage is normally between I kVAC and 30kVAC and the high voltage is normally above 30kVAC ; di f ferent voltage ranges may be applicable to other countries .
Figure 2 shows a schematic view of a power distribution system, such as the medium voltage power distribution system 12 . The power distribution system 12 includes a plurality of components , wherein the components include a plurality of nodes 14 and a plurality of line segments 13 connecting the plurality of nodes . In the following disclosure , reference is made to a medium voltage power distribution system, but the
skilled person will understand that the present disclosure and, in particular, the computer-implemented system for the control of the electric power system, is applicable for the control also of other power systems , such as e . g . a power transmission system 11 or a low voltage power distribution system 13 , or any combination of electrical power systems .
In figure 2 only nine nodes 14 and eight l ine segments 13 are shown, however the skilled person will understand that the number of nodes and line segments might be much higher, e . g . in the order of hundreds , thousands or even millions of nodes and line segments depending on the extension of the power distribution system .
The plurality of components of the power system may include a plurality of nodes that includes at least : primary substations , secondary substations , pole-mounted substations , and switchgears .
The plurality of components of the power system may include a plurality of line segments that includes at least : underground cables , overhead conductors , and overhead cables .
These two kinds of network elements are subj ect to network interventions like maintenance activities and replacements . The medium voltage distribution system may be connected to the high voltage transmission system and to the low voltage distribution system by means of trans former, as known in the art .
Other terms that could be found in the present disclosure and refer to an aggregation of network elements and are defined as follows:
-Edge: sequence of one or more line segments flanked by two nodes;
-Line: aggregation of consecutive and alternating medium voltage (MV) nodes and edges. Each line starts from a primary substation and contains one or more branches made of nodes and edges.
In figure 3, examples of modules of a computer- implemented system 70 according to the present disclosure are represented.
In one example, the computer-implemented system 70 includes a controller 20 (i.e. a control logic unit, or a control unit) that performs one or more functions by means of a corresponding module. The controller 20 may comprise one or more processors (e.g. a single/multiple core CPU, one or more microprocessors etc.) , one or more working memories (e.g. random-access memory, RAM, flash memory etc.) and one or more non-volatile instructions stores (e.g. read-only memory (ROM) , programmable ROM (PROM) , erasable PROM (EPROM) , electrically erasable PROM (EEPROM) , flash memory, etc.) storing computer- readable instructions. The controller 20 may be implemented with a distributed architecture including a plurality of processing units at different location, even remote from each other, and/or be implemented in a cloud architecture. Alternatively, the controller 20 may also be implemented with a single centralized processing device or unit. The present disclosure is not intended
to be limited to any hardware and/or software implementation of the controller 20.
In an example, the computer-implemented system may include a model 40, 41 of an outage of the power distribution system 12 configured to provide as output an outage probability for each of the plurality of components 13, 14 based on an input data set for each of the plurality of components 13, 14. Furthermore, the controller 20 may include an outage prediction module 21 configured to generate an outage probability for each of the plurality of components 13, 14 by feeding the respective input data set to the model 40, 41 and obtaining the output from the model 40, 41. The model 40, 41 may be stored in a storage unit of the system that may include a memory register, a RAM, a non-volatile memory or the like. The storage unit may be included in the controller 20 or be external with respect to the controller 20. The outage probability can be expressed by a binary value or, alternatively, by a probability value within a continuous or discrete range, e.g. as a percentage. The outage probability can be used for the control and/or the monitoring of electrical power distribution system, e.g. to determine a maintenance intervention of one of more of the components (e.g. remote maintenance or on-field maintenance) , but also for any other type of control of the power system, e.g. for load balancing control, switchgear control, or the like. The control based on the outage probability may be performed by the controller of the computer-implemented system, by an external controller and/or by an operator
of the system based on recommended interventions and/or type of control outputted by the computer-implemented system based on the determined outage probability.
Advantageously, the model 40, 41 includes a first model 40 of the outage of the nodes 14 of the power distribution system 12 and a second model 41 of the outage the line segments 13 of the power distribution system. Each model 40, 41 may have a different set of inputs data, e.g. 50, 51; the inputs data set 50 may be fed to the model 40, while the input data set 51 may be fed to the model 41 (see figure 4) ; the input data set 50 and input data set 51 may include overlapping pieces of information or data item, i.e. data items that are included in first input data set 50 may be included also in the second input data set 51. However, the first input data set 50 is different from the second input data set 51. Specifically, at least a type of data included in the first input data set 50 may not be included in the second input data set 51. Furthermore, at least a type of data included in the second input data set 51 may not be included in the first input data set 50.
It has been found advantageous to provide two models, one for the nodes and one of the line segments so as to improve the performance of the model and of the outage prediction of the related component.
In an example, the first model 40 and/or the second model 41 is/are a model based on artificial intelligence, Al, preferably a machine learning, ML, model. In an example, the first model and/or the second model is/are a ML model trained in a supervised manner. In a preferred
embodiment, it has been found that a decision tree may be selected as ML model to achieve a satisfactory performance of the model, i.e. a good prediction of the outage. Hence, the ML model 40 and/or 41 may be a decision tree. Preferably, the ML model 40 and/or 41 may be a classifier, most preferably a xgboost classifier.
The input data set 50 and/or 51 may include endogenous data and exogenous data.
Endogenous data refers to data that is related inherently to the component, or is an attribute of the component of the power distribution system; endogenous data may be defined as data that only depend from the component, its deployment, and/or its history. Examples of endogenous data is data related to topology, data related to past outages, data related to past maintenance or the like. Accordingly, the input data set may include one or more inputs related to a topology, one or more inputs related to past outages, and/or one or more inputs related to a maintenance.
The inputs related to a maintenance may e.g. include one or more of: a number of maintenance activities applied to the component, number of days from the last maintenance activity, plant maintenance, remote control maintenance, replacement, structural maintenance, pruning, and preventive measures against animals that could affect the power transmission and/or distribution system.
The inputs related to a topology may e.g. include one or more of: geo-coordinates, line of origin, customers connected to the line to which the component
belongs, component type, organizational unit of origin, construction type, conductor length, section and material, edge of origin and the number of line segments of that edge, short-circuit current, maximum operating current .
The inputs related to past outages may e.g. include one or more of: number of outages incurred by the component, number of days from the last outage, wherein each the number of outages and of number of days is determined for at least each of the following outage causes: weather events, external causes, mechanical failure, power outage, plant or tree collision.
Exogenous data may be referred to as data relating to a component that is not endogenous data. Exogenous data can be referred to as data that depend from an external agent. Examples of exogenous data are weather indexes, data related to vegetation and/or soil indexes, data related to population and territory information or the like. Accordingly, the input data set may include one or more inputs related to weather indexes (e.g. indexes based on wind, temperature and/or precipitations) , one or more inputs related to vegetation and/or soil indexes (e.g. indexes based on vegetation density, leaf type, and/or soil type) , and/or one or more inputs related to population and territory information (e.g. inputs related to population density, municipality type (littoral or insular) , and/or altitude) .
Exogenous data may be further defined as follows:
•Weather indexes created according to Climdex©
(see Lisa Alexander, Markus Donat, Yoichi Takayama, Hongang Yang, "The CLIMDEX project: Creation of long-term global gridded products for the analysis of temperature and precipitation extremes", Climate Change Resource Center & The University of New South Wales) and based on weather variables, like wind, temperature, and precipitations. For example, some of the weather indexes adopted are the number of annual windy days based on a speed threshold, the warm spell duration index, and the number of annual days with more than 20 mm of precipitations. The data source adopted is called CORDEX Copernicus , with a 12 kmx 12 km granularity.
•Vegetation and soil indexes like vegetation density, leaf type, and soil type. The data source adopted is the ERA5 Copernicus .
•Population and territory information, like population density, municipality type (littoral or insular) , and altitude. This information has been taken from the ISTAT source.
The outage probability is determined by the outage probability module 21 by using the model 40, 41. Specifically, the outage probability is determined for each of the components 13, 14 of the power system. Each component 13, 14 has an associated input data set. Figure 4 shows the input data set 50 relating to a node and the input data set 51 relating to a line segment. The input data set 50 relating to a node is fed to the ML model 40 of the outage of a node, the input data set 51 relating to the line segment is fed to the ML model 41 of the
outage of a line segment. The outage probability output from the outage probability determination model is referred as 60 in figure 4 and includes a value of outage probability for each of the components 13, 14 of the power system.
Advantageously, the input data set for each component type (i.e. for nodes and line segments) includes both endogenous and exogenous data. Most preferably, the input data set for each component type (i.e. for nodes and line segments) includes all type of input data as above described, i.e. inputs related to maintenance, inputs related to a topology, inputs related to past outages, inputs related to one or more weather indexes, inputs related to vegetation and/or soil indexes and inputs related to population and territory information.
The input data set 50, 51 may include one or more samples of input data. Each sample of input data may be taken at a certain point in time. Each sample of input data may include a plurality of types of input data, preferably all types of input data as above described. The samples of input data may be collected periodically, e.g. every year or every six months. For example, each sample of the input data set (or datasets) is a snapshot of a network element (or component) in a specific instant in time. In an example, for each network element the dataset may contain samples from 2010 to 2019, one sample per year. Besides, each sample has a target that indicates whether the network element will have (or not have) an outage in the next three years. Referring to an
example of the Italian grid, since the Italian grid has about 2 million network elements ( or components ) , the dataset is made of about 20 million samples .
The machine learning models 40 and 41 are prepared through training, validation and test , using training data, validation and test data sets . The training data set is a data set that is the gold standard, using which the model pairs input data with expected output data . The validation data set is a data set used to tune the model hyperparameters and check performances repetitively ( e . g . prevent under-/overf itting) . The test data set is a data det that the model has never seen before , used to veri fy how the model will perform when making inference .
Since the datasets are time-based, the training set refers to samples with a target period from 2010 to 2015 , instead, the validation and the test sets refer to samples with a target period of 2016-2018 and 2019-2021 , respectively . This target breakdown has been designed to avoid overlaps between the model training and the targets . Moreover, in an example of implementation, the model hyper-parameters have been configured through Amazon Web Services (AWS ) Sagemaker hyperparameter tuning .
The outage probability refers to the probability of an outage in a future time interval , e . g . a time interval of 3 years . In an example , the Al model 40 , 41 may be variable in time and may derive at any stage of data analysis the past history of a component based on the ID of the component . The input data set may include an ID
of the component. If the Al model is not capable of derive the history of the component based on the ID, it can derive this based on other features of the component. In an alternative embodiment, the sampling period, e.g. the periodicity with which the samples of the input data set are collected, is smaller than the time interval for which the outage probability is determined. The periodicity with which the samples of the input data set are collected may correspond to the periodicity of the outage probability determination by means of the outage probability determination module 21. In other words, every time a new sample of input data set is collected, the outage probability determination module 21 may determine the outage probability by means of the model 40, 41. Preferably, the input data set includes samples taken over a period of time that is longer than the future time interval for which the outage probability is determined. For example, samples are collected periodically (e.g. every year or every 6 months) over a time period of 10 years, while the outage probability refers to a time period of 3 years.
The inference may be calculated periodically by means of the models 40, 41, i.e. the computer implemented system 70 may be configured to be run (e.g. automatically or in reaction to a command) so as to determine the outage probability periodically, e.g. every six months. A new sample of the input data set 50, 51 may also be collected, e.g. from an external database with the same periodicity of the calculation of the inference (e.g. six months) . Every time the inference is calculated,
therefore , a new sample of the input data set is available and can be used to perform a new stage of training and validation . Hence , in the same run, the models 40 , 41 may be re-trained using a last available sample of the input data set , may be further validated, and, after these two steps , may perform the inference to determine the outage probability for each component . In this manner, the models 40 , 41 can be refined progressively and the related performance may be improved over time .
In an example , the computer-implemented system 70 may include the outage probability determination module 21 , an impact determination module 22 and a maintenance determination 23 . The example of figure 3 shows the computer-implemented system 70 including all of the three modules 21 , 22 and 23 . However, the computer- implemented system 70 may also include only one of the above modules or any combination of two of these modules . The modules 21 , 22 , and 23 can be implemented ( or be part of ) by the controller 20 of the computer-implemented system 70 ; the controller 20 may have any architecture , distributed or centrali zed, as above discussed in more detail .
The impact determination module 22 is configured to determine an impact of the outage of each of the components of the power transmission and/or distribution system . For example , the impact determination module 22 is configured to determine an impact of the outage by means of a deterministic algorithm 30 that , in response to a triggering input 52 , outputs the impact determined
in case of outage of each of the components 13, 14 of the power system. The algorithm 30 is determined based on the topology of the network (or power system) , the type of components in the network, the location of each component and the number of user served by each component or part of the network relying on a certain component. The input trigger 52 may be determined by a command by a user or may be automatically generated by the system, e.g. on a periodic basis.
In an example, the impact determination module (or impact simulator module) 22 contains a deterministic algorithm 30 that, given a network element and the configuration of its line of origin, simulates an outage in that specific network element and replicates all the network operations to resupply the electricity for all the affected customers.
The sequence of operations made by the deterministic algorithm can include the following:
1) Automation: operations made automatically by the network, by using specific sensors and strategies that open/close the switches which belong to the affected network section.
2) Remote control: operations made by the worker (s) at the operation center (remotely) .
3) Manual phase:
-Alert: time in which the on-field worker (s) is alerted to the outage.
-Logistic: time in which the on-field worker (s) gets ready to reach the affected network and perform the manual operation.
-Manual selection : operations made by the onfield worker ( s ) .
-Reparation : time to fix the outage in a speci fic network element . This time is considered only when the manual selection is not enough to re-supply all the network customers .
The operations made during the remote control and the manual selection phases are dichotomic . An operator of the system may choose a speci fic element of the line af fected by the outage to split the line into two parts made of approximately the same number of customers . Finally, after the attempts to re-supplying the network, the worker knows exactly in which of the two parts of the line the outage is . These operations are made until the operator identi fies the exact element in which the outage took place .
The duration of the remote control and manual selection has been estimated for each zone distributed in the territory by taking the average time of the remote control and manual selection durations registered in the outage events of the last three years .
In an example , as a result , the deterministic algorithm 30 returns the impact of the outage through one of or both of the two following indexes :
AV20 : number of af fected customers per interruption time
SAIDI : AV20 divided by the number of the DSO customers .
These indexes are output by the impact determination module 22 .
These two indicators have been selected, in an example , for the country Italy, but they could be adapted to any other regulatory framework . An index that considers both the results of the machine learning models and the simulator can be obtained by multiplying the outage probability and its impact (AV20 or SAIDI ) . In this way, the network planners can opt for a unique index to rank all the network elements and allocate a maintenance intervention to the worst-performing ones .
Hence , the output 61 of the impact determination module includes an impact ( or an indication of an impact ) of the outage for each component of the power system . The computer-implemented system 70 may be configured to calculate , for each component , an index obtained by multiplying the outage probability determined by the module 21 and the impact determined by the module 22 . This combined index provides a useful indication relating to the necessity of a maintenance intervention .
The outage impact determined by the module 22 can be used for the control and/or the monitoring of electrical power distribution system, e . g . to determine a maintenance intervention of one of more of the components ( e . g . remote maintenance or on- field maintenance ) , a priority of a maintenance intervention, but also for any other type of control of the power system, e . g . for load balancing control , switchgear control , or the like and/or for design of network extensions or topology modi fications . The control based on the outage impact may be performed by the controller 20 of the computer- implemented system, by an external controller and/or by
an operator of the system based on recommended interventions and/or type of control outputted by the computer-implemented system based on the determined outage impact .
The maintenance intervention determination module 23 is configured to determine a maintenance intervention or that no maintenance is required for one or more of the plurality of components , preferably for all the components 13 , 14 of the power system .
The maintenance intervention determination module 23 aims to suggest to the network planners ( or operators ) a maintenance intervention that may include a set of replacement actions and/or maintenance activities that reduce the outage probability of the network elements . The maintenance intervention determination module 23 may also command a remote maintenance intervention to the network component based on the determination of the maintenance intervention .
A replacement action as a maintenance intervention relates to the conductors/cables ; this type of maintenance intervention applies only to the line segments . The ef fect of a replacement action is determined by using the model 41 as follows :
1 ) Simulating the replacement in an existent line segment by modi fying its features , e . g . , by setting the maintenance and the outage features to zero ;
2 ) Predicting the new outage probability with the new input features ;
3) Calculate the deviation from the original outage probability.
The maintenance intervention determination module 23 may there follow a procedure using the Al model 41 to determine the maintenance intervention for line segments, wherein the maintenance intervention includes a replacement action of a cable and/or a conductor.
More in general, the maintenance intervention determination module 23 may be configured to determine a maintenance intervention for a given component using the Al model 41 of the outage of the power distribution system by:
-modifying an input relating to the maintenance included in the input data set for the given component (e.g., by setting the maintenance and the outage features to zero) ,
-determining an updated outage probability based on the modified input by using the Al model 41 of the outage of the power distribution system,
-determining a difference between the updated outage probability and the outage probability as previously determined for the given component,
-determining a maintenance intervention related to the modified input (e.g. the cable/conductor replacement) if the determined difference is above a predetermined threshold (the threshold could be e.g. 0 in order to assess whether the outage probability is reduced or not) .
For example, referring to figure 5, the modified input 51a relating to the maintenance refers to cable or
a conductor; more specifically, the maintenance and the outage features of the cable/conductor may be set to zero, because it is supposed that the cable/conductor is new in case of performing the maintenance intervention (this simulates the case in which no outage occurred in the past and no maintenance has been performed) . This means, that the model 41 is fed with a modified input data set 51a, 51b to take into account a situation in which the cable/conductor has been replaced, in order to determine the outage prediction in case the maintenance intervention has been performed. The other input data of the input data set may remain unchanged and are referred to as 51b. Hence, the updated outage probability is determined using a modified input data set 51a, 51b that differs from the original input data set 51 only in the modified input data referring to the maintenance. In this manner, the maintenance intervention determination module 23 can determine whether a specific maintenance (i.e. the cable/conductor replacement) has a positive consequence, in the sense that the outage probability is e.g. reduced, when the threshold is zero to assess the difference between the outage probability and the updated outage probability. The threshold to assess the chance in the outage probability may be set also to a value different from zero, e.g. to 10% or 30%, in order to select the maintenance intervention only if a significant change in the risk of outage is found.
It has been found that the above process using the Al model is particularly suited to determine the
cable/conductor replacement as the maintenance intervention for a line segment .
In case of a node or in case of a di f ferent intervention for a line segment that is not a cable/conductor replacement , the maintenance determination module 23 is configured to use a second process by using a heuristic to determine the maintenance intervention .
In an example of implementation, the procedure to calculate the benefit of one or more maintenance activities can be described with the example represented in Figure 7 . In this figure , first the observation point is set , and then the outage coef ficients for the next year are estimated both for the network elements that have ( 1-^5 activities ) and not have ( 0 activity) maintenance activities in the previous two years . The outage coef ficients ( or outage rates ) are calculated as the percentage of network elements that will have an outage in the next year and had a certain amount of maintenance in the previous two years . For example , %0MI-5 refers to the percentage of network elements that will have an outage in the next year and had from 1 to 5 maintenance activities in the previous 5 years . %0MI-5 is defined as follows :
Oi %0MI-5 — 7 — U1 + (J2 where Oi is the number of network elements that will have ( Oi ) or will not have ( O2 ) an outage in the next year, knowing that those elements had from 1 to 5 maintenance activities in the previous two years .
As a result, the percent deviation between the two outage coefficients represents the benefit of going from 0 maintenance activities to 1-5.
The percent deviation is calculated with the following formula:
Moreover, the coefficients represented in the figure have been averaged considering all the available historical years of maintenance activities and outages (i.e., from 2008 up to now) .
Accordingly, the heuristic is based (or includes) on a database 42 that stores the historical data regarding the maintenance and the outage of all components of the network .
The database 42 includes (i) first historical data related to the past maintenance of the components of a specific type of component in a first period of time and (ii) second historical data related to the past occurrence of an outage for the components of the specific type in a second period of time subsequent to the first period of time. The first historical data and the second historical data are stored and associated to each component; in other words, the data referring to an outage for a certain component (that could be absence of an outage or occurrence of an outage) is stored in association to the data referring to the past maintenance for that component. For each type of components (e.g. a primary substation, secondary substations, pole-mounted substations or the like) a first outage rate (or
coefficient) is calculated based on the first and second historical data for components that had a maintenance intervention in the first period of time and a second outage rate is calculated based on the first and second historical data for components that had no maintenance intervention in the first period of time. The maintenance intervention determination module 23 is then configured to determine a maintenance intervention based on the first and second outage occurrence rates. Specifically, by using the database 42 and the stored associations for each type of component, the maintenance determination module 23 can determine which maintenance intervention can be most effective to reduce the outage risk.
The heuristics have been applied only to certain network elements and only with a subset of maintenance activity types, as follows:
•Overhead conductors: pruning, plant maintenance, replacement (only accessories, not the cable /conductor)
•Overhead cables: plant maintenance, replacement (only accessories, not the conductor)
•Underground cables: plant maintenance
•Primary substations: structural maintenance, preventive measures against animals, plant maintenance
Hence, in an example, the first historical data stored in the database 42 refers to anyone of:
-for overhead conductors: pruning, plant maintenance, replacement of accessories other than the cable /conductor
- for overhead cables : plant maintenance , replacement of accessories other than the cable/conductor , - for underground cables : plant maintenance - for primary substations : structural maintenance , preventive measures against animals , plant maintenance .
In case the maintenance that can be determined by the module 23 does not include the cable/conductor replacement , the module 23 may operate also in the absence of an outage probability determination module 21 ; in this case , the maintenance determination may be performed based on the heuristic only .
In one example, referring to figure 6 , the maintenance intervention determination module 23 may be configured to determine a level of priority 63 of a maintenance intervention for a component among a plurality of components based on the outage probability determined by the outage prediction module 21 for that component and/or based on the impact determined by the impact determination module 22 for that component . Also another module of the system, e . g . module 21 or 22 may calculate the level of priority . For example , the priority may be expressed in terms of a priority index obtained based on the multiplication of an index representing the outage impact and an index representing the outage probability of a component . Based on the priority index, the priority of the maintenance intervention determined for a plurality of components may be output . Accordingly, an order of the determined maintenance interventions may be determined and output to the operator .
In an example, when the difference between the outage probabilities of different components is lower than a predetermined difference threshold (i.e. the outage probabilities are comparable as being included in a relatively narrow range) , the maintenance intervention determination module 23 may be configured to generate an order of the maintenance interventions relating to the different components based on the impact determined by the impact determination module 22 for the different components, e.g. giving priority to the maintenance of components whose outage impact is larger. This allows to improve the control and efficiency in the operation of the power system, by improving continuity and quality of service .
The computer-implemented system 70 may include a display for outputting information to an operator. For example, the display may display the outage probability for each component, the outage impact for each component, and the maintenance intervention determined for each component, or any other output data determined by the computer-implemented system and described herein. An example of a display is shown in figures 8 or 9.
The maintenance intervention outputted by the module 23 can be used for the control and/or the monitoring of electrical power distribution system, e.g. to carry out a maintenance intervention of one of more of the components (e.g. remote maintenance) or to provide recommendations relating to an on-field maintenance. A control based on the determined maintenance intervention may be performed by the controller of the computer-
implemented system, by an external controller and/or by an operator of the system based on recommended interventions and/or type of control outputted by the computer-implemented system .
RESULTS
Examples of results obtained are related to the three modules described in the previous paragraph : machine learning models , the outage simulator, and network interventions .
The performance of the machine learning models is based on the test set , which includes all the network elements on the 1 st of January 2019 . The performance has been estimated for three di f ferent scenarios : 5/ 10/25K interventions at the network elements that are more likely to experience an outage ( the overall network has about 1 . 82 million network elements in 2019 ) . The metric adopted is the precision which is , in this case , the ratio of the correctly predicted outages to the total predicted outages : true positive precision = - — - - true positive + false positive
The results are represented in Table 1 , for both the node and the line segment models . From Table 1 , the model that has the highest precision is the line segment model for a 5K intervention scenario . A higher-level result has been obtained by aggregating the node and line segment information at the line level . The performance of this aggregation is represented in Table 2 . In this case , the aggregated model can correctly identi fy the
line in which there will be an outage in 97 % of the time , with a 5K interventions scenario .
Table 1 - Models Performance
Table 2 - Line Aggregation Performance
The outage simulator has been tested in an AWS machine of type ml . m5 . 12 . xlarge that takes about 3 hours to simulate 1 million network elements . Besides , its proper functioning has been tested in a portion of the Italian territory by a team of on- field workers .
Table 3 - Network Interventions
Lastly, the results of the network interventions module are summari zed in Table 3 .
For the interventions obtained through the machine learning model , the percentage of outage probability reduction was averaged on all the applicable elements . Instead, for the interventions obtained through heuristics , the outage percent deviations was averaged for each applicable maintenance activity . The highest outage probability reduction is obtained with the plant maintenance applied to the nodes and the overhead cables ( 70% ) . Instead, the lowest reduction is related to the replacement of cables or conductors of any type ( 18 % ) .
All the results have been summari zed in a dedicated dashboard made through Mi crosoft PowerBi ( Figure 8 ) . The figure shows the network element predictions from Jan 2022 to Jan 2024 in the Ancona province in Italy . On the top-left , there is the number of network elements selected, the belonging electrical lines , and the SAIDI in case the 174 network elements would experience an outage . Below, some filters are applied to the network elements , and a chart represents the network behavior by relating the outage probability to the impact . On the right , there are a graphical representation of the network elements and a chart that sums up the average performance of the electrical lines by aggregating the outage probability per impact of each network element .
By clicking on the map, detailed information on the outage probability, the impact, and two suggested network interventions for each network element is displayed (Figure 9) . The first suggested network intervention refers to the least effort, i.e., the minimum number of network interventions to get an outage probability improvement. Instead, the second one refers to the max yield, i.e., the number of network interventions to minimize the outage probability.
In an example, the module 21 is operated periodically to output the outage probability of all components, e.g. every six months. In an example, the module 22 is operated periodically to output the outage impact of all components, e.g. every six months. In an example, the module 23 is operated periodically to output the maintenance intervention of all components, e.g. every six months. Preferably, all modules 21-23 are operated substantially at the same time.
In an example, all outputs of the module 21, 22 and 23 as described herein may be output to an operator of the computer-implemented system 70 by means of a user interface, e.g. a display. All modules 21, 22 and 23 may be implemented as software module, hardware module and any combination thereof. The present disclosure is not limited to any specific implementation of the hardware/sof tware architecture of the modules 21, 22 and 23. The modules 21, 22 and 23 may be communicably connected with each other to exchange any of the data as described herein.
In one embodiment , the outage impact may be determined by the module 22 only for those components whose outage probability is determined to be higher than a predetermined threshold by the module 21 . However, in a preferred embodiment , the outage impact is determined for all components of the network regardless of the outage probability . Furthermore , the maintenance intervention is determined by the module 23 only for those components whose outage probability is determined to be higher than a predetermined threshold by the module 21 ; however, preferably, the maintenance intervention is determined for all components of the network .
Modifications and Variations
Although the electrical networks are described to have alternative current (AC ) electrical energy, it would be understood that the medium voltage electrical network MV and/or the low voltage electrical network LV may operate with direct current ( DC ) electrical energy instead .
Claims
1 . Computer-implemented system for the control of a power transmission and/or distribution system, wherein the power transmission and/or distribution system includes a plurality of components that include a plurality of nodes and a plurality of line segments connecting the plurality of nodes ; the computer-implemented system comprising a controller that includes :
• an impact determination module configured to determine an impact of an outage of each of the components of the power transmission and/or distribution system; and
• a maintenance intervention determination module configured to determine a maintenance intervention or that no maintenance is required for one or more , or all , of the plurality of components ; characteri zed by further comprising an arti ficial intelligence model of an outage of the power transmission and/or distribution system configured to provide as output an outage probability for each of the plurality of components based on an input data set for each o f the plurality of components ; wherein the controller further includes an outage prediction module configured to generate an outage probability for each of the plurality of components by feeding the respective input data set to the arti ficial intelligence model and obtaining the output from the arti ficial intelligence model ;
and wherein the maintenance intervention determination module is configured to determine a maintenance intervention for a given component using a first process using the arti ficial intelligence model by :
• modi fying an input relating to the maintenance included in the input data set for the given component ;
• determining an updated outage probability based on the modi fied input by using the arti ficial intelligence model of the outage of the power transmission and/or distribution system;
• determining a di f ference between the updated outage probability and the outage probability as previously determined for the given component ;
• determining a maintenance intervention related to the modi fied input i f the determined di f ference is above a predetermined threshold .
2 . Computer-implemented system of claim 1 , wherein the impact determination module is configured to determine an impact of the outage based on a deterministic algorithm associated to each of the components , wherein the output of the deterministic algorithm includes at least one of : a number of af fected customers per interruption time and/or a number of af fected customers per interruption time divided by the number of the customers of the Distribution System Operator .
3. Computer-implemented system according to claim 1 or 2 , wherein the maintenance intervention determination module is configured to determine a level of priority of a maintenance intervention for a component based on the outage probability determined by the outage prediction module for that component and based on the impact determined by the impact determination module for that component .
4 . Computer-implemented system according to any preceding claim, wherein the maintenance intervention module is configured to use the first process i f the component is a line segment , wherein the input relating to maintenance that is modi fied is related to a replacement action of the conductors/cables of a line segment and wherein the input relating to maintenance that is modi fied is fed to the arti ficial intell igence model of the outage of the line segment to determine the updated outage probability .
5 . Computer-implemented system according to any preceding claim, wherein the maintenance intervention determination module is configured to determine a maintenance intervention using a second process by using a heuristic ; wherein the heuristic includes for each type of component : ( i ) first historical data related to the past maintenance of the components of a speci fic type of component in a first period of time and ( ii ) second historical data related to the past occurrence of an
outage for the components of the speci fic type in a second period of time subsequent to the first period of time ; wherein the first historical data and the second historical data are stored in association with the first historical data for each component ; wherein a first outage rate is calculated based on the first and second historical data for components that had a maintenance intervention in the first period of time and a second outage rate is calculated based on the first and second historical data for components that had no maintenance intervention in the first period of time ; wherein the maintenance intervention determination module is configured to determine a maintenance intervention based on the first and second outage occurrence rates .
6. Computer-implemented system of claim 5 , wherein the first historical data refers to anyone of :
• for overhead conductors : pruning, plant maintenance , replacement of accessories other than the cable /conductor ;
• for overhead cables : plant maintenance , replacement of accessories other than the cable /conductor ;
• for underground cables : plant maintenance ;
• for primary substations : structural maintenance , preventive measures against animals , plant maintenance .
7 . Computer-implemented system according to claim 5 or 6 , wherein the intervention determination module is configured to determine a maintenance intervention using the second process using the heuristic i f the component is a node of the power transmission and/or distribution system .
8 . Computer-implemented system according to any of claims 5-7 , wherein the intervention determination module is configured to determine a maintenance intervention using the second process using the heuristic i f the component is a line segment of the power transmission and/or distribution system, and the maintenance intervention is not a replacement action related to the conductors/cables of the line segment .
9. Computer-implemented system according to any of claims 5- 6 , wherein the intervention determination module is configured to determine a maintenance intervention using the second process by using the heuristic i f the component is a node of the power transmission and/or distribution system, and wherein, i f the component is a line segment :
• the maintenance intervention determination module is configured to determine a candidate maintenance intervention as the maintenance intervention using the second process when the candidate maintenance intervention is not a replacement of a cable/conductor , and
• the maintenance intervention determination module is configured to determine a candidate maintenance intervention as the maintenance intervention using the first process when the candidate maintenance intervention is a replacement of a cable/conductor .
10 . Computer-implemented system according to any of the preceding claims , wherein the power transmission and/or distribution system is a medium voltage power distribution system .
Applications Claiming Priority (2)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| IT102023000011319A IT202300011319A1 (en) | 2023-06-05 | 2023-06-05 | SYSTEM AND METHOD FOR ADVANCED CONTROL OF AN ELECTRICITY NETWORK |
| PCT/IB2024/055431 WO2024252268A1 (en) | 2023-06-05 | 2024-06-04 | System and method for the enhanced control of an electrical network |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| EP4721233A1 true EP4721233A1 (en) | 2026-04-08 |
Family
ID=87801443
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| EP24733032.7A Pending EP4721233A1 (en) | 2023-06-05 | 2024-06-04 | System and method for the enhanced control of an electrical network |
Country Status (4)
| Country | Link |
|---|---|
| EP (1) | EP4721233A1 (en) |
| CL (1) | CL2025003834A1 (en) |
| IT (1) | IT202300011319A1 (en) |
| WO (1) | WO2024252268A1 (en) |
Family Cites Families (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US7203622B2 (en) * | 2002-12-23 | 2007-04-10 | Abb Research Ltd. | Value-based transmission asset maintenance management of electric power networks |
| CN106384210B (en) * | 2016-10-28 | 2019-08-27 | 贵州电网有限责任公司贵阳供电局 | A kind of power transmission and transforming equipment maintenance prioritization method based on maintenance risk income |
| WO2021076326A1 (en) * | 2019-10-16 | 2021-04-22 | Vit Tall Llc | Resilient decision systems and methods |
| CN115051406A (en) * | 2022-06-15 | 2022-09-13 | 贵州电网有限责任公司 | Multi-stage coordination control risk scheduling method considering source network uncertainty |
| CN116008866B (en) * | 2023-03-10 | 2024-08-27 | 中国电信国际有限公司 | Submarine cable fault monitoring method and device, electronic equipment and storage medium |
-
2023
- 2023-06-05 IT IT102023000011319A patent/IT202300011319A1/en unknown
-
2024
- 2024-06-04 WO PCT/IB2024/055431 patent/WO2024252268A1/en not_active Ceased
- 2024-06-04 EP EP24733032.7A patent/EP4721233A1/en active Pending
-
2025
- 2025-12-04 CL CL2025003834A patent/CL2025003834A1/en unknown
Also Published As
| Publication number | Publication date |
|---|---|
| IT202300011319A1 (en) | 2024-12-05 |
| CL2025003834A1 (en) | 2026-03-06 |
| WO2024252268A1 (en) | 2024-12-12 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US11735912B2 (en) | Systems and methods for ai-assisted electrical power grid fault analysis | |
| Leite et al. | Resiliency assessment in distribution networks using GIS-based predictive risk analytics | |
| Baldick et al. | Vulnerability assessment for cascading failures in electric power systems | |
| JP7257455B2 (en) | Method and monitoring system for monitoring controller decision logic performance | |
| EP4068551B1 (en) | System and method for predicting failure in a power system in real-time | |
| Ferreira et al. | A nonlinear binary programming model for electric distribution systems reliability optimization | |
| US20130138482A1 (en) | Capital asset planning system | |
| Barrett et al. | A risk-based framework for power system modeling to improve resilience to extreme events | |
| EP3750223B1 (en) | Predicting voltage stability of a power system post-contingency | |
| Camilo López et al. | Design and simulation of a centralized self-healing scheme for unbalanced three-phase electrical distribution systems | |
| EP4429070A1 (en) | Method and processing system for processing data related to a power grid, and system comprising a power grid | |
| WO2024252268A1 (en) | System and method for the enhanced control of an electrical network | |
| WO2024252273A1 (en) | Enhanced control of electrical networks | |
| Shadi et al. | Proactive Cable Replacement Planning Using Neural Weibull Proportional Hazard Modeling and Bayesian Nested Monte Carlo Simulation Under Data Deficiencies | |
| Merigeault et al. | Big data applications for improving the reliability of the french electricity distribution grid | |
| Khalyasmaa et al. | Machine learning application for the high-voltage equipment lifecycle forecasting | |
| Sabot et al. | Towards Probabilistic Dynamic Security Assessment and Enhancement of Large Power Systems | |
| Vermigli et al. | Data-driven AI network analysis | |
| Dokic et al. | Optimized Asset Management in Distribution Systems Based on Predictive Risk Analysis | |
| Oliveira et al. | Data-driven methodology to predict distribution lines failure location | |
| Falahati et al. | Power system reliability enhancement considering smart monitoring | |
| Perkin et al. | Performance of probabilistic disturbance forecasts in extreme weather on the Icelandic power system | |
| Zhang et al. | Real-Time Outage Management in Active Distribution Networks Using Reinforcement Learning over Graphs | |
| Yao et al. | Multi-timescale Simulation, Risk Assessment, and Mitigation of Cascading Outages | |
| Gowtham et al. | Implementing Resiliency Upgrades on Power Distribution System Using Graph Theoretic Methods |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| STAA | Information on the status of an ep patent application or granted ep patent |
Free format text: STATUS: UNKNOWN |
|
| STAA | Information on the status of an ep patent application or granted ep patent |
Free format text: STATUS: THE INTERNATIONAL PUBLICATION HAS BEEN MADE |
|
| PUAI | Public reference made under article 153(3) epc to a published international application that has entered the european phase |
Free format text: ORIGINAL CODE: 0009012 |
|
| STAA | Information on the status of an ep patent application or granted ep patent |
Free format text: STATUS: REQUEST FOR EXAMINATION WAS MADE |
|
| 17P | Request for examination filed |
Effective date: 20251127 |
|
| AK | Designated contracting states |
Kind code of ref document: A1 Designated state(s): AL AT BE BG CH CY CZ DE DK EE ES FI FR GB GR HR HU IE IS IT LI LT LU LV MC ME MK MT NL NO PL PT RO RS SE SI SK SM TR |