Data processing system and data processing method for substation system, and power system substation

Abstract

A data processing system (20) and a data processing method are provided for a substation system, where the substation system includes a primary system and a secondary system. The data processing system receives and processes data acquired or generated by the secondary system to perform a state assessment based on at least one digital twin (31). The data processing system performs at least one function (32) that performs a check of the state evaluation or supports the state evaluation. Redundant information included in the received data may be used to perform the check.

Description

Technical Field

[0001] Embodiments of the present invention relate to systems and methods operable to process measurements acquired in a substation system. Embodiments of the present invention particularly relate to such data processing systems and methods operable to detect anomalies and / or faults in a substation system. Background Art

[0002] The power system is an important infrastructure component. The power system (such as power generation, transmission and / or distribution system) includes power system components in the primary system.

[0003] Modern power systems use protection systems to operate power system components. Intelligent electronic devices (LEDs) are one implementation of protection devices in such protection systems.

[0004] Conventional substation protection systems are based on technologies in which human experts specify the protection logic or its settings. This conventional approach requires numerous settings and is prone to malfunctioning due to incorrect parameterization. Conventional protection systems can be challenged by scenarios such as high-impedance faults, incipient faults, and other transient phenomena that may appear to be faults.

[0005] Many conventional protection systems have protection logic designed for conventional generation, transmission, and / or distribution systems. While such protection logic can be used for fault or anomaly detection when there is a certain degree of renewable energy penetration, the increased deployment of resources interfaced with power electronic converters (such as renewable energy and battery energy storage systems) modifies the fault characteristics of the grid, making fault or anomaly detection more challenging.

[0006] Conventional protection systems may also tend not to react adequately to so-called "invisible errors" in the data acquisition system. Conventional protection systems do not provide safeguards against such invisible error phenomena within the protection time scale, which may be as short as milliseconds after the event. Summary of the Invention

[0007] It is an object of the present invention to provide apparatus, systems and methods that provide enhanced techniques for detecting anomalies and / or faults. It is also an object of the present invention to provide apparatus, systems and methods that address one or more of the challenges described above. In view of the above, it is desirable to provide systems and methods that can be used to supplement existing protection and / or monitoring systems with apparatus that can reduce possible errors made by human engineers during protection system configuration and / or commissioning. It is also desirable to provide systems and methods that can be used to supplement existing protection and / or monitoring systems in a manner that enhances fault or anomaly detection to take into account higher penetration rates of renewable energy sources, incorrect instrumentation, or other unseen errors in data acquisition systems.

[0008] According to the invention, a data processing system and method are provided as described in the independent claims. The dependent claims define preferred embodiments.

[0009] According to one aspect of the present invention, a data processing system for a substation system is provided, wherein the substation system includes a primary system and a secondary system, the primary system including a plurality of primary system components. The data processing system includes at least one interface operable to receive data acquired or generated by the secondary system and related to the primary system. The data processing system further includes at least one processing circuit operable to continuously and during field operation of the primary system perform a status assessment on at least one primary system component of the plurality of primary system components or on the primary system of the substation system based on the received data and at least one digital twin. The at least one processing circuit is further operable to perform at least one function that performs a check on the status assessment or supports the status assessment based on one or more of the following: redundant information included in the received data, domain knowledge data and / or a trained machine learning (ML) model.

[0010] Various effects and advantages are achieved through the data processing system. The data processing system can be implemented as an additional component separate from, but interfacing with, the intelligent electronic devices (IEDs) of a substation. This provides enhanced reliability. The data processing system can operate to supplement the protection and / or monitoring functions that may be implemented in the IEDs.

[0011] The data processing system utilizes at least one digital twin. Furthermore, the function utilizes redundant information (such as redundant measurements) included in the received data, domain knowledge, and / or trained ML models to further enhance the state assessment based on the at least one digital twin. This provides additional reliability for detecting anomalies and / or faults.

[0012] The data processing system may be operable such that the state assessment is a non-predictive state assessment relating to a current time.

[0013] Thus, the data processing system may be used to assess the current status of the primary system and / or its components.

[0014] The data processing system may be operable such that the status assessment comprises status assessment using currently available measurement values ​​without accessing an internal state of the IED.

[0015] Thereby, the data processing system may be used to perform and check status assessments even in situations where access to the internal state of the IED may not be available (eg for security, communication bandwidth and / or data protection reasons).

[0016] The data processing system may be operable to perform an action, such as an output action and / or a control action, based on the output and status evaluation of at least one function. The action may include generating an alarm or warning. Alternatively or additionally, the action may include selectively overriding a decision of logic (e.g., protection logic) of the IED.

[0017] The data processing system may be a centralized data processing system of the substation.

[0018] This allows for improved decision making regarding anomalies and / or faults based on a centralized approach that integrates various measurements acquired or generated within the substation.

[0019] The data processing system may be a centralized data processing system of a power system substation, such as a substation of a power grid or other power generation, transmission and / or distribution system.

[0020] Thereby, improved detection of faults and / or anomalies in the electrical power field is provided, wherein undetected faults and / or anomalies may have potentially catastrophic effects.

[0021] The at least one digital twin may include a plurality of first digital twins. Each of the plurality of first digital twins may model an associated one of the plurality of primary system components.

[0022] Thus, component-level fault and / or anomaly detection is facilitated. Such first digital twins (each of which is associated with only a single one of the multiple primary system components) can be implemented using a model customized for the respective component. Component-level approaches also require limited measurements associated with the respective primary system components.

[0023] The plurality of first digital twins may include at least two first digital twins, both of which model the same primary system component.

[0024] This allows several model-based condition assessments to be run in parallel for the same primary system component.

[0025] At least one function may check outputs of at least two first digital twins, both of which model a same primary system component.

[0026] Thus, resolution of potentially conflicting findings may be performed using redundant measurements, domain knowledge, and / or trained ML models.

[0027] The first digital twins may be operable to perform electrical modeling.At least some of the first digital twins may be operable to model thermal and / or mechanical characteristics of components of the primary system.

[0028] Thereby, not only electrical anomalies or faults can be identified, but also thermoelectrical, thermal and / or mechanical anomalies or faults.

[0029] The first digital twin may include a first digital twin operable to model one, several, or all of the following: switchgear (such as a circuit breaker or power switch), a transformer, a generator, a converter (such as an AC / DC, DC / AC, and / or DC / DC converter), an energy storage system (such as a battery storage system or a flywheel energy storage system), but is not limited thereto. Different models customized for the respective primary system components may be used.

[0030] Thus, a component level model can be implemented for the primary system components of the substation.

[0031] The at least one function may include a plurality of first component level functions.

[0032] Thus, each of the first digital twins can be supplemented with supervisory or auxiliary functions that check or supplement condition assessments based on, for example, redundant measurements, domain knowledge, or trained ML models. Such component-level functions can be implemented using limited measurements from the substation data acquisition system to facilitate their implementation.

[0033] Each of the first component-level functions may be operable to: cause an adjustment to a process executed by the first digital twin to perform the condition assessment based on the inspection; and / or supervise the condition assessment based on domain knowledge and / or a trained ML model.

[0034] Thus, the state assessment technique is improved based on redundant measurements, domain knowledge and / or trained ML models.

[0035] Adjusting may include discarding unreliable portions of data used in the condition assessment when performing the condition assessment.

[0036] Thereby, at least one function may be operable to improve the reliability of a condition assessment performed based on the first component level digital twin.

[0037] The first component level functionality may be operable to oversee decisions made by an IED associated with a respective component.

[0038] The first component level function may be operable to perform its supervisory and / or assistance functions using only a subset of the received data (in particular, only a subset of the available measurements). This allows for a simple implementation. The first component level function may be operable to perform its supervisory and / or assistance functions based on measurements acquired or generated at or near the respective component of the primary system.

[0039] The at least one digital twin may include a second digital twin modeling at least a primary system of the substation.

[0040] Thus, optionally in addition to component level modeling, modeling at substation level can also be performed. This allows anomaly and / or fault detection to be performed in a more reliable manner.

[0041] Although modeling at the substation level is more complex than modeling at the component level, it provides further enhanced reliability in identifying anomalies and / or faults.

[0042] The second digital twin may be operable to receive outputs generated by the plurality of first digital twins.

[0043] Thus, substation-level modeling can be combined with component-level modeling. This allows for more reliable anomaly and / or fault detection. The second digital twin can use the outputs it receives from the multiple first digital twins to perform system-level processing (taking into account the components modeled by the multiple first digital twins), thereby facilitating anomaly and / or fault detection.

[0044] The second digital twin may be operable to receive and process redundant information included in the data.

[0045] Thereby, the protection system can take advantage of redundant measuring instruments typically deployed in substations.

[0046] The secondary system may include a first measurement device and a second measurement device, both of which are operable to measure the same electrical characteristic. At least one function may be operable to process both a first measurement value obtained or generated by the first measurement device and a second measurement value obtained or generated by the second measurement device to supervise the second digital twin.

[0047] Thereby, the protection system may utilize redundant measuring instruments typically deployed in a substation to supervise anomaly and / or fault detection based on at least one digital twin.

[0048] The redundant information may comprise redundant measurements. Alternatively or additionally, the redundant information may comprise at least two measurements of the same substation parameter taken or generated by different measuring instruments of the secondary system.

[0049] Thus, the protection system can utilize the redundant measuring instruments typically deployed in the substation. This is useful for supervising the detection of anomalies and / or faults based on at least one digital twin.

[0050] The at least one processing circuit may be operable to perform a substation state assessment based on the second digital twin.

[0051] Thus, component level modeling can be enhanced by substation level modeling.

[0052] Substation status assessment may include assessing whether the substation is healthy, has an anomaly (such as an incipient fault), or has a fault.

[0053] At least one digital twin may include a third digital twin that models at least a secondary system.

[0054] Thereby, identification of errors in the data acquisition system (ie, the secondary system) is facilitated.

[0055] At least one function may be operable to identify an instrument error based on, for example, expert knowledge, plausibility checks or other processing, and optionally based on a third digital twin.An instrument error may be detected.

[0056] The at least one processing circuit may be operable to identify an instrument error in the secondary system based on the inspection.

[0057] Thereby, anomalies and / or faults in the primary system can be more easily distinguished from incorrect instrumentation.

[0058] The identified instrument errors may include incorrect instrument settings, such as an incorrect winding ratio set for at least one current transformer. The incorrect instrument settings may include instrument settings of at least one IED.

[0059] The processing circuitry may be operable to identify, based on the inspection, a root cause of a discrepancy between the data and the at least one digital twin.

[0060] This further improves the reliability of anomaly and / or fault detection. For example, anomalies and / or faults in the primary system can be distinguished from anomalies and / or faults in the secondary system. Depending on the root cause identified, improved corrective and / or mitigating actions can be taken.

[0061] The processing circuitry may be operable to determine a root cause of a discrepancy based on a comparison of the output(s) of at least one digital twin with received data (such as measurements). The comparison may include residual analysis or other processing that compares errors between the received data (such as measurements) and the digital twin output(s) to identify the most likely cause of the discrepancy.

[0062] Thereby, the reliability of abnormality and / or fault detection is further improved.

[0063] At least one processing circuit may be operable to distinguish, based at least on redundant information, a root cause selected from the group consisting of: an anomaly of at least one primary system component of a primary system; an inappropriateness of a model or model parameterization of at least one digital twin; an anomaly of a measuring instrument.

[0064] This further improves the reliability of anomaly and / or fault detection. For example, anomalies and / or faults in the primary system can be distinguished from anomalies and / or faults in the secondary system. Depending on the root cause identified, improved corrective and / or mitigating actions can be taken.

[0065] The at least one interface may be operable to receive an IED logic output from at least one IED of the substation.

[0066] Thereby, the data processing system may interface with one or several IEDs deployed in a substation.The data processing system may be operable to supplement decision making and analysis performed in the IEDs using traditional techniques.

[0067] The at least one processing circuit may be operable to check the IED logic output based on the at least one digital twin and the received data.

[0068] Thereby, at least one function may be used to check decisions made by the IED, in particular using at least one digital twin.

[0069] The at least one interface may be operable to receive auxiliary data indicative of a problem with the data.The at least one processing circuit may be operable to process the auxiliary data indicative of a problem with the data.

[0070] Thus, when performing the status evaluation, less reliable data can be identified and potentially discarded by at least one functional check status evaluation.

[0071] The at least one processing circuit may be operable to trigger at least one action based on the inspection.

[0072] Thus, the data processing system can take appropriate actions based on the information available to it, including redundant measurements, expert knowledge data, and / or trained ML models.

[0073] The at least one action may include one or more of an output action via a human machine interface (HMI), a corrective action, and / or a mitigating action.

[0074] Thus, the results of the data processing system can be operated to automatically implement alarms or other information outputting warnings via the HMI, and / or automatic corrective and / or mitigating actions.

[0075] The at least one action may include selectively overriding a decision made by an IED of the substation based on the status evaluation and the output of the at least one function.

[0076] Thereby, the results of the data processing system can be operated to supplement the operation of the IEDs of the substation.

[0077] The at least one action may include a corrective or mitigating action that affects a power system protection function and / or a power system monitoring function.

[0078] Thus, the results of the data processing system can operate to enhance the reliability of anomaly and / or fault detection in a power system substation.

[0079] The data processing system may be a data processing system operable to be installed in a substation.

[0080] Thus, the functionality of the data processing system may be implemented locally at the substation, thereby avoiding potential adverse effects that may be associated with data transmission delays on a data transmission path traversing multiple communication nodes.

[0081] The data processing system may be a data processing system operable to be installed locally at the substation.The data processing system may be a data processing system operable to be installed remotely from a regional or national control centre.

[0082] Thus, the functions of the data processing system can be implemented locally at the substation, thereby avoiding potential adverse effects that may be associated with data transmission delays on the data transmission path traversing multiple communication nodes. This facilitates the execution of control functions (such as mitigation functions), especially for functions with critical timing requirements.

[0083] The data processing system need not be located in the substation but may be located remotely from the substation. For the purpose of illustration, the data processing system may be located in a control centre when operable to perform monitoring functions.

[0084] According to another aspect of the present invention, there is provided a power system substation comprising: a primary system comprising a plurality of primary system components; a secondary system operable to acquire data related to the primary system; and a data processing system of any aspect or embodiment operable to receive and process the acquired or generated data.

[0085] Such power system substations provide improved reliability.The data processing system implements enhanced, non-traditional techniques to augment anomaly and / or fault detection.

[0086] According to another aspect of the present invention, there is provided a power system (such as a power grid) including a power system substation according to an embodiment.

[0087] Such power systems provide improved reliability.The data processing system implements enhanced, non-traditional techniques to augment anomaly and / or fault detection.

[0088] The power system substation of the power system may further include a plurality of IEDs interfaced with the data processing system.

[0089] Thus, the data processing system is operable to augment and enhance decisions made by multiple IEDs.

[0090] A power system substation or a power system may include at least one converter and a component interfaced with the converter. The component may include a renewable energy source.

[0091] Thereby, the effects and advantages of data processing systems can be utilized in a context where conventional conservation techniques may be particularly prone to making incorrect decisions in view of renewable energy penetration rates.

[0092] According to another aspect of the present invention, there is provided a use of a data processing system according to any aspect or embodiment disclosed herein for checking anomalies and / or fault detection in a power system substation.

[0093] Thus, the data processing system provides enhanced reliability in detecting anomalies and / or faults in a power system substation.

[0094] According to another aspect of the present invention, a data processing method for a substation system is provided, wherein the substation system includes a primary system and a secondary system, the primary system including a plurality of primary system components. The data processing method includes: receiving data acquired or generated by the secondary system and related to the primary system; and continuously and during field operation of the primary system, performing a status assessment on at least one primary system component of the plurality of primary system components or the primary system of the substation system based on the received data and at least one digital twin; and performing at least one function that performs a check on the status assessment based on at least one of the following: redundant information included in the received data, domain knowledge, and / or a trained machine learning (ML) model.

[0095] The method may be executed by a data processing system according to an embodiment.

[0096] The steps of performing the state evaluation and executing the at least one function may be performed by a processing circuit. The processing circuit may comprise one or several integrated circuits.

[0097] Optional features of the method and the effects achieved thereby correspond to the more detailed features described with reference to the data processing system.

[0098] Various effects and advantages are achieved through the data processing method. The data processing method can be performed to supplement decision logic processing implemented in intelligent electronic devices (IEDs) in a substation. This provides enhanced reliability. The data processing method can also be used to supplement protection and / or monitoring functions that may be implemented in IEDs.

[0099] The data processing method utilizes at least one digital twin. Furthermore, the function utilizes redundant information (such as redundant measurements), domain knowledge, and / or trained ML models included in the received data to further enhance the state assessment based on the at least one digital twin. This provides additional reliability for detecting anomalies and / or faults.

[0100] The data processing method may be operable such that the state assessment is a non-predictive state assessment relating to a current time.

[0101] Thus, the data processing method can be used to evaluate the current status of the primary system and / or its components.

[0102] The data processing method may be operable such that the status evaluation comprises status evaluation using currently available measurement values ​​without accessing an internal state of the IED.

[0103] Thereby, the data processing method may be used to perform and check status assessments even in situations where access to the internal status of the IED may not be available (eg for security, communication bandwidth and / or data protection reasons).

[0104] The method may further include performing an action, such as an output action and / or a control action, based on the output and status evaluation of the at least one function. The action may include generating an alarm or warning. Alternatively or additionally, the action may include selectively overriding a decision of logic (e.g., protection logic) of the IED.

[0105] The processing may be performed by a data processing system, which may be a centralized data processing system of the substation.

[0106] This allows for improved decision making regarding anomalies and / or faults based on a centralized approach that integrates various measurements acquired or generated within the substation.

[0107] The processing may be performed by a data processing system, which may be a centralized data processing system of a power system substation, such as a substation of a power grid or other power generation, transmission, and / or distribution system.

[0108] Thereby, improved detection of faults and / or anomalies in the electrical power field is provided, wherein undetected faults and / or anomalies may have potentially catastrophic effects.

[0109] The at least one digital twin may include a plurality of first digital twins, each of which may model an associated one of the plurality of primary system components.

[0110] Thus, component-level fault and / or anomaly detection is facilitated. Such first digital twins (each of which is associated with only a single one of the plurality of primary system components) can be implemented using a model customized for the respective component. Component-level approaches also require limited measurements associated with the respective primary system components.

[0111] The plurality of first digital twins may include at least two first digital twins, both of which model the same primary system component.

[0112] This allows several model-based condition assessments to be run in parallel for the same primary system component.

[0113] The at least one function may check outputs of at least two first digital twins, both of which model a same primary system component.

[0114] Thus, resolution of potentially conflicting findings may be performed using redundant measurements, domain knowledge, and / or trained ML models.

[0115] The first digital twin may perform electrical modeling.At least some of the first digital twin may model thermal and / or mechanical properties of components of the primary system.

[0116] Thereby, not only electrical anomalies or faults can be identified, but also thermoelectrical, thermal and / or mechanical anomalies or faults.

[0117] The first digital twin may include, but is not limited to, a first digital twin that models one, several, or all of the following: switchgear (such as a circuit breaker or power switch), a transformer, a generator, a converter (such as an AC / DC, DC / AC, and / or DC / DC converter), and an energy storage system (such as a battery storage system or a flywheel energy storage system). Different models customized for the corresponding primary system components may be used.

[0118] Thus, a component level model can be implemented for the primary system components of the substation.

[0119] The at least one function may include a plurality of first component level functions.

[0120] Thus, each of the first digital twins can be supplemented with supervisory or auxiliary functions that check or supplement condition assessments based on, for example, redundant measurements, domain knowledge, or trained ML models. Such component-level functions can be implemented using limited measurements from the substation data acquisition system, facilitating their implementation.

[0121] Each of the first component-level functions may cause: adjusting a process performed by the first digital twin to perform a state assessment based on the inspection; and / or supervising the state assessment based on domain knowledge and / or a trained ML model.

[0122] Thus, state assessment techniques are improved based on redundant measurements, domain knowledge and / or trained ML models.

[0123] Adjustments may include discarding unreliable portions of data used in the condition assessment when performing the condition assessment.

[0124] Thereby, the at least one function may be operable to improve the reliability of a condition assessment performed based on the first component level digital twin.

[0125] The first component level function may oversee decisions made by the IEDs associated with the respective components.

[0126] The first component-level function may use only a subset of the received data (in particular, only a subset of the available measurements) to perform its supervisory and / or assistance functions. This allows for a simple implementation. The first component-level function may perform its supervisory and / or assistance functions based on measurements acquired or generated at or near the respective component of the primary system.

[0127] The at least one digital twin may include a second digital twin modeling at least one primary system of the substation.

[0128] Thus, optionally in addition to component level modeling, modeling at substation level can also be performed. This allows anomaly and / or fault detection to be performed in a more reliable manner.

[0129] Although modeling at the substation level is more complex than modeling at the component level, it provides further enhanced reliability in identifying anomalies and / or faults.

[0130] The second digital twin may receive outputs generated by the plurality of first digital twins.

[0131] Thus, modeling at the substation level can be combined with component level modeling, and anomaly and / or fault detection can be performed in a more reliable manner.

[0132] The second digital twin can receive and process redundant information included in the data.

[0133] Thereby, the protection system can take advantage of redundant measuring instruments typically deployed in substations.

[0134] The secondary system may include a first measurement device and a second measurement device, both of which are operable to measure the same electrical characteristic. The at least one function may process both a first measurement value obtained or generated by the first measurement device and a second measurement value obtained or generated by the second measurement device to supervise the second digital twin.

[0135] Thereby, the protection method may utilize redundant measuring instruments typically deployed in a substation to supervise anomaly and / or fault detection based on the at least one digital twin.

[0136] The redundant information may include redundant measurement values. Alternatively or additionally, the redundant information may include at least two measurement values ​​of the same substation parameter acquired or generated by different measuring instruments of the secondary system.

[0137] Thus, the protection method can make use of redundant measuring instruments typically deployed in substations. This is useful for supervising the detection of anomalies and / or faults based on the at least one digital twin.

[0138] The data processing method can perform substation status assessment based on the second digital twin.

[0139] Thus, component level modeling can be enhanced by substation level modeling.

[0140] Substation status assessment may include assessing whether the substation is healthy, has an anomaly (such as an incipient fault), or has a fault.

[0141] At least one digital twin may include a third digital twin that models at least a secondary system.

[0142] Thereby, identification of errors in the data acquisition system (ie, the secondary system) is facilitated.

[0143] At least one function may identify an instrument error based on, for example, expert knowledge, plausibility checks, or other processing, and optionally based on a third digital twin. An instrument error may be detected.

[0144] Data processing methods can identify instrument errors in secondary systems based on inspection.

[0145] Thereby, anomalies and / or faults in the primary system can be more easily distinguished from incorrect instrumentation.

[0146] The identified instrument errors may include incorrect instrument settings, such as an incorrect winding ratio set for at least one current transformer. The incorrect instrument settings may include instrument settings of at least one IED.

[0147] The data processing method may include identifying a root cause of a discrepancy between the data and at least one digital twin based on the inspection.

[0148] This further improves the reliability of anomaly and / or fault detection. For example, anomalies and / or faults in the primary system can be distinguished from anomalies and / or faults in the secondary system. Depending on the root cause identified, improved corrective and / or mitigating actions can be taken.

[0149] The data processing method may include: distinguishing a root cause selected from the group consisting of: an abnormality of at least one primary system component of the primary system; inappropriateness of a model or model parameterization of the at least one digital twin; an abnormality of a measuring instrument based at least on redundant information.

[0150] This further improves the reliability of anomaly and / or fault detection. For example, anomalies and / or faults in the primary system can be distinguished from anomalies and / or faults in the secondary system. Depending on the root cause identified, improved corrective and / or mitigating actions can be taken.

[0151] The at least one interface may receive an IED logic output from at least one IED of the substation.

[0152] Thus, the data processing method may provide for interaction of a substation-level data processing system with one or several IEDs deployed in a substation. The substation-level data processing system may be operable to supplement decision making and analysis performed in the IEDs using conventional techniques.

[0153] The method may include checking the IED logic output based on the at least one digital twin and the received data.

[0154] Thereby, the at least one function may be used to check decisions made by the IED, in particular using the at least one digital twin.

[0155] The at least one interface may receive auxiliary data indicating a problem with the data.The data processing method comprises: processing the auxiliary data indicating a problem with the data.

[0156] Thereby, less reliable data may be identified and potentially discarded when the execution status assessment passes the at least one functional check status assessment.

[0157] The data processing method may include triggering at least one action based on the inspection.

[0158] Thus, appropriate action is taken based on the information available to the data processing system, including redundant measurements, expert knowledge data, and / or trained ML models.

[0159] The at least one action may include one or more of an output action via a human machine interface (HMI), a corrective action, and / or a mitigating action.

[0160] Thus, the processing results automatically cause output of warning alarms or other information via the HMI, and / or automatic corrective and / or mitigating actions.

[0161] The at least one action may include selectively overriding a decision made by an IED of the substation based on the status evaluation and the output of the at least one function.

[0162] Thereby, the results of the data processing system can be operated to supplement the operation of the IEDs of the substation.

[0163] The at least one action may include a corrective or mitigating action that affects a power system protection function and / or a power system monitoring function.

[0164] Thus, the results of the data processing system can operate to enhance the reliability of anomaly and / or fault detection in a power system substation.

[0165] The data processing method may be executed by a data processing system installed in a substation.

[0166] Thus, centralized processing may be implemented locally at the substation, thereby avoiding potential adverse effects that may be associated with data transmission delays on a data transmission path traversing multiple communication nodes.

[0167] The data processing method may be performed by a data processing system which is installed locally at the substation and remotely from a regional or national control center.

[0168] Thus, the functionality of the data processing system may be implemented locally at the substation, thereby avoiding potential adverse effects that may be associated with data transmission delays on a data transmission path traversing multiple communication nodes.

[0169] According to another aspect of the present invention, a method for operating a power system substation is provided, which power system substation includes: a primary system, which includes multiple primary system components; a secondary system, which is capable of operating to obtain data related to the primary system, the method including: using the processing technology of any aspect or embodiment disclosed herein to process the acquired or generated data.

[0170] Such an operating method provides improved reliability.The process implements enhanced non-traditional techniques to amplify anomaly and / or fault detection.

[0171] According to another aspect of the present invention, there is provided a method of operating a power system, such as a power grid, comprising processing acquired or generated data using the processing technique of any aspect or embodiment disclosed herein.

[0172] Such an operating method provides improved reliability.The operating method implements enhanced non-traditional techniques to amplify anomaly and / or fault detection.

[0173] The power system substation of the power system may further include a plurality of IEDs interfaced with the data processing system. The operating method may include augmenting decisions made by the plurality of IEDs based on the data processing method according to an aspect or embodiment.

[0174] The operating method may be performed on a power system substation or a power system, which may include at least one converter and a component interfaced with the converter. The component may include a renewable energy source.

[0175] Thereby, the effects and advantages of data processing systems can be utilized in a context where conventional conservation techniques may be particularly prone to making incorrect decisions in view of renewable energy penetration rates.

[0176] According to a further embodiment, machine readable instruction code is provided which, when executed by at least one programmable circuit, causes the programmable circuit to perform a method according to an embodiment.

[0177] According to a further embodiment, a non-transitory storage medium is provided having machine-readable instruction codes stored thereon, which, when executed by at least one programmable circuit, cause the programmable circuit to perform a method according to an embodiment.

[0178] Various effects and advantages are achieved through embodiments of the present invention. For illustration, systems and methods according to embodiments provide enhanced techniques for detecting anomalies and / or faults in, for example, a substation of a power system. These systems and methods can be used to supplement existing protection and / or monitoring systems with devices that can reduce potential errors made by human engineers during protection system configuration and / or commissioning. These systems and methods can be used to supplement existing protection and / or monitoring systems in such a way that fault or anomaly detection is enhanced to account for higher penetration rates of renewable energy, incorrect instrumentation, or other unseen errors in data acquisition systems.

[0179] These systems and methods may be used in conjunction with, but are not limited to, power systems such as the power grid. These systems and methods may be used in conjunction with, but are not limited to, power systems such as the power grid that have renewable energy sources and / or battery-based or mechanical energy storage systems. BRIEF DESCRIPTION OF THE DRAWINGS

[0180] The embodiments of the present invention will be described with reference to the accompanying drawings, wherein similar or identical reference numerals designate elements having similar or identical configurations and / or functions.

[0181] Figure 1 It is a block diagram of a data processing system for a substation.

[0182] Figure 2 It is a block diagram of the components of a substation.

[0183] Figure 3 It is a block diagram of the components of a substation.

[0184] Figure 4 It is a flow chart of the data processing method.

[0185] Figure 5 It is a block diagram of the processing circuitry of a data processing system.

[0186] Figure 6 is a block diagram of another processing circuit of a data processing system.

[0187] Figure 7 is a block diagram of another processing circuit of a data processing system.

[0188] Figure 8 is a block diagram of another processing circuit of a data processing system.

[0189] Figure 9 yes Figures 5 to 8 A block diagram of the functional blocks of the processing circuit.

[0190] Figure 10 yes Figures 5 to 8 A block diagram of the functional blocks of the processing circuit.

[0191] Figure 11 yes Figures 5 to 8 A block diagram of the functional blocks of the processing circuit.

[0192] Figure 12 A substation comprising a data processing system according to an embodiment is shown.

[0193] Figure 13 A substation comprising a data processing system according to an embodiment is shown.

[0194] Figure 14 A portion of a power system including several data processing systems according to an embodiment is shown.

[0195] Figure 15 It is a flow chart of the data processing method.

[0196] Figure 16 It is a flow chart of the data processing method.

[0197] Figure 17 It is a flow chart of the data processing method.

[0198] Figure 18 It is a flow chart of the data processing method.

[0199] Figure 19 It is a block diagram of the functional data processing modules of a data processing system.

[0200] Figure 20 FIGURES illustrate fitting errors identified using a data processing system and method according to an embodiment.

[0201] Figure 21 The diagram shows the traditional IED Figure 20 Decision logic output for the same fault in .

[0202] Figure 22 The current transformer output is shown.

[0203] Figure 23 The data processing system and method according to the embodiment are used to Figure 22 The fitting error identified by the current transformer output.

[0204] Figure 24 It is a flow chart of the data processing method.

[0205] Figure 25 and Figure 26 Illustrated are fitting errors identified using a data processing system and method that allows for root cause analysis to be performed according to an embodiment.

[0206] Figure 27 Residual analysis results obtained using a data processing system and method according to an embodiment that allows for performing root cause analysis are illustrated.

[0207] Figure 28 Illustrated are fitting errors identified using a data processing system and method that allows for root cause analysis to be performed according to an embodiment.

[0208] Figure 29 Residual analysis results obtained using a data processing system and method according to an embodiment that allows for performing root cause analysis are illustrated.

[0209] Figure 30 and Figure 31 Illustrated are fitting errors identified using a data processing system and method that allows for root cause analysis to be performed according to an embodiment.

[0210] Figures 32 to 34 Residual analysis results obtained using a data processing system and method according to an embodiment that allows for performing root cause analysis are illustrated.

[0211] Figure 35 It is a flow chart of the data processing method.

[0212] Figures 36 to 38 Illustrated are fitting errors identified using a data processing system and method that allows for root cause analysis to be performed according to an embodiment.

[0213] Figures 39 to 40 Residual analysis results obtained using a data processing system and method according to an embodiment that allows for performing root cause analysis are illustrated. DETAILED DESCRIPTION

[0214] Embodiments of the present invention will be described with reference to the accompanying drawings. In the accompanying drawings, similar or identical reference numerals designate elements having similar or identical configurations and / or functions.

[0215] Although the embodiments will be described in conjunction with data processing systems and methods associated with power system substations, the embodiments are not limited thereto.

[0216] Unless explicitly stated otherwise, the features of the exemplary embodiments can be combined with one another.

[0217] According to the present invention, data processing systems and methods are provided that provide a condition assessment based on modeling of primary system components or primary systems of a substation. The data processing systems and methods further provide at least one function for using the output of the modeling in combination with additional information. The at least one function may include a supervisory function that checks or otherwise verifies the condition assessment based on at least one digital twin. Alternatively or in addition, at least one function may include an assistance function that is operable to assist in the condition assessment performed using at least one digital twin. The additional information may be or may include redundant measurements or other redundant data (such as redundant event data), domain knowledge data (which may be or may include a configuration description of the substation system, such as a substation configuration description (SCD) file) and / or other data-driven techniques (such as a trained ML model).

[0218] According to conventional terminology in the field of Industrial Automation and Control Systems (IACS), the primary system of a substation includes those components that perform the functions of the substation intended to operate in the system in which the substation is located. For illustration purposes, when the substation is a power system substation, the primary system may consist of those components that carry the high power of the power generation, transmission, and / or distribution system (e.g., of a high-voltage, medium-voltage, or low-voltage power grid). For substations on a utility grid that transfer fluids (such as fresh or used water), the primary system may consist of those components that transfer fluids.

[0219] Primary system components are components of the primary system. For power system substations, primary system components include those components that deliver the high power encountered in power generation, transmission, and / or distribution systems (e.g., of high-voltage, medium-voltage, or low-voltage power grids). Examples of such primary system components include, but are not limited to, circuit breakers, switches (such as disconnectors and earthing switches), power transformers, instrument transformers, and generators.

[0220] According to conventional terminology in the field of IACS, the secondary system of a substation includes or consists of measuring instruments and associated data acquisition systems. The secondary system may optionally include equipment for controlling, regulating, protecting, and monitoring primary equipment. The secondary equipment and its interconnected circuits are collectively referred to as the secondary system. The secondary system is capable of operating to provide safety and reliability by, for example, performing control and monitoring functions. The secondary system may include, but is not limited to, secondary equipment such as protection relays, automatic reclosers, sensors, fault recorders, and control switches.

[0221] According to non-limiting embodiments, the data processing systems and methods utilize one or more digital twins. The term "digital twin" refers to a digital representation, digital model, or digital "shadow" constructed corresponding to digital information about a physical device or system (such as a primary system component, a primary system, or a primary and secondary system of a substation). That is, the digital information can be implemented as a "twin" of the physical device or system (e.g., a primary system component, etc.) and information associated with and / or embedded within the primary system component.

[0222] A digital twin may include one or more component-level digital twins, also referred to herein as "first digital twins." Each of the first digital twins is operable to model a primary system component of a primary system. The digital twin may include a system-level digital twin that models a primary system that includes several primary system components that interact with each other in the primary system. Thus, the second digital twin may take into account interactions between primary system components that are interconnected within the primary system.

[0223] As used herein, a digital twin of a primary system component (i.e., a "first digital twin") may be operable to model one, several, or all electrical, thermal, chemical, and / or mechanical parameters of the primary system component using physics-based modeling. A digital twin of a primary system (such as a substation) (i.e., a "second digital twin") may be operable to model one, several, or all electrical, thermal, chemical, and / or mechanical parameters of the primary system using physics-based modeling; this may include modeling several primary system components interconnected in the primary system.

[0224] As used herein, a "model" may be or may include a model based on physics-based modeling.

[0225] Throughout the life cycle of a primary system component, a digital twin may be associated with the primary system component, and optionally with the secondary system component. In some examples, a digital twin may include a physical object in real space, a digital twin of the physical object that exists in virtual space, and information that connects the physical object to its digital twin. A digital twin may exist in a virtual space that corresponds to the real space, and may include data flow links from the real space to the virtual space, and information flow links from the virtual space to the real space and virtual subspaces. The data flow or information flow links may correspond to a digital thread that represents the communication framework between the data source and the digital twin model. The digital thread may enable a comprehensive view of asset data throughout the life cycle of the asset. For example, a digital twin may correspond to a virtual model of an asset, and the digital thread may represent the connected data flows between the asset data source and the virtual model.

[0226] According to non-limiting embodiments, a data processing system and method are operable to perform hierarchical monitoring, protection, and supervision of a substation. A substation includes electrically connected components such as transformers, transmission lines, busbars, capacitor banks, reactors, surge protectors, and the like. The data processing system and method utilize both component-level modules and substation-level modules, which work together to monitor each device for internal anomalies.

[0227] According to non-limiting embodiments, a data processing system and a data processing method may be operable to mitigate the risk of false triggering due to hidden failures in a secondary (data acquisition) system based on measurement redundancy at the substation level.

[0228] According to non-limiting embodiments, several digital twins and several supervisory or assistance functions may be provided in a data processing system or data processing method. For illustration purposes, a data processing system and method according to non-limiting embodiments may utilize: several first digital twins, each of which models associated primary system components of a primary system of a substation; and at least one second digital twin, which models the primary system of the substation. At least one function may include a first plurality of first assistance functions, each of which assists in a condition assessment based on one of the first digital twins. At least one function may include a supervisory function that checks a condition assessment performed based on the first digital twin and / or the at least one second digital twin.

[0229] The data processing system and data processing method can be operable to monitor each component based on measurements associated with the component (such as terminal measurements) using a model-driven approach, while overseeing component-level decisions at the substation level through another model-driven approach to mitigate problems associated with data errors. The effect of such a data processing system and data processing method is twofold. First, the model-driven approach can supplement the threshold / setting-based protection philosophy of traditional decision logic and provide additional intelligence to the protection / monitoring of the substation. Second, if implemented in a hierarchical manner, the hierarchical processing at the component level and the substation level provides enhanced robustness against unseen errors in the instrumentation that might otherwise trigger spurious operations, such as relay operations. Such a data processing system and data processing method can also help maintain verified models of substation equipment and calibrate instruments during field operations (e.g., periodically) while providing monitoring and protection functions.

[0230] Although the associated figures describe embodiments that use modeling and processing at both the component level and the substation level, the data processing systems and data processing methods disclosed herein can also advantageously use component level modeling in conjunction with functions that check or support condition assessments obtained through component level modeling (such as supervisory functions or assistance functions). Combining component level modeling with substation level modeling and / or supervisory functions operating at the substation level is beneficial but may not always be required.

[0231] As used herein, the term digital twin refers to a model or model-based technology that models at least one characteristic of at least one component of the primary system of a substation (as will be described for the first digital twin and the second digital twin) or models a measuring instrument (as will be described for the third digital twin (if it exists)). If the substation is a substation of an electric power system (such as, a power grid), the digital twin may model the electrical characteristics. The digital twin may additionally model the thermal and / or mechanical characteristics of the components. The digital twin may additionally or alternatively model characteristics other than electrical, thermal and mechanical characteristics, such as the chemical characteristics of the insulating fluid (e.g., insulating oil) of the power transformer.

[0232] The data processing systems and methods disclosed herein may be operable to interact with protection and / or monitoring devices of a substation. The data processing systems and methods disclosed herein may be operable to interact with protection and / or monitoring devices that use threshold-based methods to detect anomalies and / or faults. Such conventional protection and / or monitoring devices may be intelligent electronic devices (IEDs).

[0233] As used herein, the term IED refers to a device that may be operable in a manner compatible with or according to IEC 61850 (eg according to the latest version of IEC 61850 as available at the filing or priority date of the present application).

[0234] The data processing systems and methods disclosed herein can be operable such that at least one digital twin includes a plurality of first digital twins, each of the plurality of first digital twins models an associated one of a plurality of primary system components, and at least one digital twin includes a second digital twin that models at least one primary system (i.e., several primary system components in the primary system that interact, for example), wherein the second digital twin is operable to receive outputs generated by the plurality of first digital twins. When the first digital twin and the second digital twin are combined in this manner, the reliability and / or accuracy of state assessment can be improved.

[0235] The processing systems and methods disclosed herein can be operable to identify the root cause of discrepancies between data and at least one digital twin based on checks performed by the processing system, thereby distinguishing anomalies and / or faults in the primary system from anomalies and / or faults in the secondary system. Thus, the risk of false alarms or false corrective actions (such as incorrect protection actions not warranted by the primary system state) can be mitigated. Power system availability is thereby enhanced because anomalies and / or faults in the primary system are derived from anomalies and / or faults in the secondary system (such as incorrect measuring instruments).

[0236] The processing systems and methods disclosed herein may be operable to receive IED logic outputs from at least one IED of a substation system and to check the IED logic outputs based on the received data and at least one first digital twin associated with a primary system component and / or a second digital twin associated with the primary system. Discrepancies between the IED logic outputs and actual conditions in the primary system may thereby be detected. Appropriate corrective actions (such as correcting incorrect IED settings) may be triggered.

[0237] The processing systems and methods disclosed herein may be operable to utilize at least one digital twin including a second digital twin modeling at least one primary system, wherein the second digital twin is operable to receive and process redundant information included in the data, wherein the redundant information includes redundant measurements obtained using redundant measuring instruments in the secondary system. Thus, the redundant measurements can be used, for example, to check a condition assessment of the primary system, thereby enhancing the accuracy and reliability of the condition assessment.

[0238] Figure 1 A data processing system 20 is shown which is operable to process data comprising measurements acquired or generated by one or more measuring instruments of the data acquisition system. The data may include event data, but is not limited thereto.

[0239] The data processing system 20 may be implemented as a device including a device housing, Figure 1 The components shown in FIG. 2 are housed in the device housing. The data processing system may be implemented as a combination of several devices communicatively coupled to one another. The data processing system 20 may be operable to be used in conjunction with or in cooperation with a protection and / or monitoring device that implements threshold-based decision logic.

[0240] The data processing system 20 may generally be operable to provide protection and / or monitoring functions for a substation. The substation may be, but is not limited to, a power system substation. The substation includes a primary system comprising a plurality of primary system components, such as, but not limited to, transformers, circuit breakers, other switchgear, reactors, impedances, generators, or other primary system components. The substation includes a secondary system comprising measuring instruments that acquire data to be processed by at least the data processing system 20 and / or other data acquisition systems. The secondary system may optionally include additional protection and / or monitoring devices (such as IEDs that may implement threshold-based decision logic).

[0241] The data processing system 20 may be implemented at the substation level. The data processing system 20 may be installed at or within the substation. The data processing system 20 may be operable to interact with a control center (such as a national or regional control center for the power grid). The data processing system 20 may be distinct from and separate from the control center (e.g., by being located within the substation). The data processing system 20 may be integrated with the control center.

[0242] As will be explained in more detail below, the data processing system 20 is operable to utilize at least one digital twin 31 that models one or several components of the primary system of the substation or optionally the primary system of the substation as a whole. Several digital twins may be used in combination.

[0243] The data processing system 20 may be operable to additionally perform at least one supervisory and / or assistance function 32. The at least one supervisory and / or assistance function 32 may be or may include a supervisory function operable to review the results of a condition assessment based on at least one digital twin. Alternatively or additionally, the at least one supervisory and / or assistance function 32 may include an assistance function operable to support a condition assessment based on at least one digital twin.

[0244] The data processing system 20 includes at least one processing circuit 30. The at least one processing circuit 30 is operatively coupled to the at least one interface 21. The at least one processing circuit 30 is configured to receive data acquired or generated by data acquisition devices (such as measuring instruments or other sensors) of the substation's secondary system. The at least one processing circuit 30 is operative to perform a condition assessment based on one or more digital twins 31 and to perform at least one supervisory and / or assistance function 32.

[0245] The at least one processing circuit 30 may include any one or any combination of an integrated circuit, an integrated semiconductor circuit, a processor, a controller, an application-specific integrated circuit (ASIC), a field-programmable gate array (FPGA), circuit(s) including quantum bits (qubits) and / or quantum gates, but is not limited thereto.

[0246] At least one processing circuit 30 may be operable to perform a state assessment using measurements associated with a component of the primary system (such as electrical characteristics of the terminals of the component). The electrical characteristics may include time domain samples of voltage, current and / or power (e.g., instantaneous samples of any of these quantities) and / or frequency domain phasors of voltage, current and / or power. The specific implementation of the state assessment will generally depend on the type of component in question. A skilled person may use a variety of techniques at their disposal to perform a state assessment, such as a state of health (SoH) assessment, based on the measurements. For purposes of illustration and not limitation, a Markov chain implementation may be used to model transitions between healthy, incipient anomaly, and fault states of any component of the primary system. The received measurements may be used to evaluate an initial state of the Markov chain modeling. Alternatively or additionally, incipient or developing state anomalies of any component of the primary system of the substation may be detected and classified based on, for example, deviations of the measurements from expected behavior determined based on the digital twin.

[0247] Such deviations can be detected and, optionally, quantified based on the fit error between the digital twin result and the measured value. The fit error over a time interval or frequency range can be quantified using any suitable metric, such as the L2 norm or other L-norm. For example, Figure 17 The implementation of detecting deviations will be described in more detail.

[0248] At least one processing circuit 30 may be operable to use domain knowledge to perform one or more authenticity checks, for example to detect deviations between measurements and expected behavior determined based on the digital twin or to otherwise process the measurements. The domain knowledge depends on the corresponding primary system. For illustration, for a power system substation, the domain knowledge may include electrodynamic rules (such as Kirchhoff's rules), consistency between switch or circuit breaker states and current, voltage and / or power measurements, typical variation limits for voltage and / or current, directionality or polarity of current at the terminals of the primary system components, etc. Any one or any combination of these information elements may be used in data processing. For a fluid grid (such as a fresh water, heating or coolant circuit, wastewater or process fluid system), the domain knowledge may include fluid dynamic relationships, consistency between valve states, and measurements of fluid properties (such as dynamic and / or static pressure, velocity and / or density).

[0249] While a technician may use a variety of techniques at his or her disposal to implement condition assessment, the technician may be free to use a general model having a discrete state space that represents different states of the corresponding asset (such as healthy, incipient fault, fault). EP3923214A1, EP3923213A1 and EP3923101A1 disclose techniques that can be used to perform condition assessment for any asset. Typically, transition probabilities between discrete states of a probabilistic model (such as a Markov chain model space) can be inferred based on measurements (which may be historical, or acquired at a specific substation, and / or may be based on cluster learning methods). (Multiple) digital twins can be used in various ways, such as by inferring transition probabilities of discrete state models based on modeling performed using (multiple) digital twins. Measurements from the secondary system can be used to initialize a simulation that provides as its output: the probability that a primary system component is healthy, or an indicator of degradation of a primary system component (such as remaining useful life).

[0250] Yet another decision making technique that may be used in any of the embodiments disclosed herein is provided in EP4080702A1.

[0251] J. Zhao et al. provide an overview of various exemplary techniques that a technician can use at their disposal to implement state assessment in the following document: "Power system dynamic state estimation: Motivations, definitions, methodologies, and future work", IEEE Transactions on Power Systems (Vol. 34, No. 4, July 2019), pp. 3188-3198, January 23, 2019, IEEE, DOI: 10.1109 / TPWRS.2019.2894769.

[0252] In yet other exemplary embodiments, the state assessment may be based on differences between measured values ​​(e.g., time-domain or frequency-domain electrical characteristics such as voltage(s), current(s), and / or power(s), which may be obtained via phasor measurement units or otherwise) and measured values ​​expected based on the digital twin(s).

[0253] C. Brosinsky et al., “Recent and prospective developments in power system control centers: Adapting the digital twin technology for application in power system control centers” (2018 IEEE International Energy Conference (ENERGYCON), Limassol, Cyprus, June 3–7, 2018, IEEE, DOI: 10.1109 / ENERGYCON.2018.8398846) and Y. Yang et al., “State Evaluation of Power Transformer Based on Digital Twins” (IEEE International Conference on Energy, 2018, DOI: 10.1109 / ENERGYCON.2018.8398846).

[0019] The paper "Power Transformer Condition Assessment Based on Digital Twins" (2019 IEEE International Conference on Service Operations and Logistics and Informatics (SOLI), November 6-8, 2019, Zhengzhou, China, IEEE, DOI: 10.1109 / SOLI48380.2019.8955043) provides further examples of techniques that can be used (also specifically in the field of power systems) to implement condition assessment based on (multiple) digital twins.

[0254] The one or more digital twins 31 are not limited to modeling electrical properties. For illustration purposes, for components of a power system substation, modeling may include electrical, thermal, and / or mechanical modeling for at least some of these components. Additional modeling may be implemented, such as for modeling the properties of the insulating oil of a power transformer, in which case the concentration of soft shell in the insulating oil may be modeled. For further illustration purposes, modeling may include modeling the state of transformer breathers that may be provided on transformers at the substation. For further illustration purposes, modeling may include modeling chemical properties (e.g., to model dissolved gases in the insulating fluid of a power transformer or instrument transformer).

[0255] By way of further illustration, additional details of the digital twin for certain primary system components are discussed below:

[0256] A digital twin of a transformer may have inputs operable to receive electrical properties (which may be obtained from measurements and / or control settings), temperature measurements, and optionally measurements related to insulation condition, such as measurements related to the insulating oil (e.g., dissolved gas analysis (DGA) measurements). The digital twin of the transformer may perform physics-based processing on the inputs. The physics-based processing may model the electrical, thermal, and / or mechanical behavior of the transformer. The digital twin may provide as output information such as thermal properties (e.g., temperature at locations where no measurements are available), insulation condition (e.g., of paper-based insulation surrounding transformer windings), transformer aging, and the like.

[0257] A digital twin of a capacitor bank can have inputs operable to receive information about the bank geometry and / or configuration, electrical parameters related to capacitance and ohmic losses in each element or cell or bank, thermal parameters (such as dielectric loss factor), and thermal coupling between capacitors. The digital twin can perform physics-based processing on the inputs by also considering terminal voltages and currents, and temperature sensor measurements. The digital twin can provide information such as internal branch currents, voltages, and heat distribution throughout the bank (in the core, housing, and ambient, etc.) as outputs.

[0258] A digital twin of a transmission line can have inputs operable to receive conductor geometry, electrical parameters related to the line (such as resistance per unit length, inductance, and capacitance), and thermal parameters (such as linear mass density, specific heat of the line material, absorption coefficient, etc.). The digital twin can perform physics-based processing on the inputs by also considering terminal voltage and current, temperature, and weather sensor measurements. The digital twin can provide information such as the following as output: line charging current, voltage across the line inductance, and temperature distribution.

[0259] The at least one processing circuit 30 may be operable to use any or any combination of additional information available to the at least one processing circuit 30 and which may be useful for reviewing and / or supporting a condition assessment based on the at least one.

[0260] At least one supervisory and / or assistance function 32 may be operable to check whether a condition assessment that classifies a component of the primary system as abnormal or faulty is correct using redundant measurements included in the data received at the at least one interface 21. The redundant measurements, which may originate from redundant measuring instruments in the secondary system, may be used to check whether the condition assessment based on the at least one digital twin 31 is consistent with the redundant measurements or whether it may be incorrect due to, for example, incorrect measuring instrument settings.

[0261] Various embodiments of the inspection status assessment are described in more detail herein. Figure 10 、 Figure 11 as well as Figures 24 to 40 As will be explained in more detail, the supervisory and / or assistance function 32 may include a supervisory function that uses redundant measurements to determine whether the identification of a fault (which may be identified by the verification function of the data processing system 20 based on the digital twin) is correct (in the sense that the fault exists in the primary system) or whether the identification is incorrect (in the sense that the fault is caused by incorrect instrument settings or any root cause other than a fault in the primary system).

[0262] Alternatively or additionally, the supervisory and / or assistance function 32 may be operable to use domain knowledge data or other data-driven processing techniques (these processing techniques may include, for example, trained machine learning (ML) models) to support condition assessment based on at least one digital twin. Based on the domain knowledge data or using the ML model, the supervisory and / or assistance function 32 may be operable to determine which data to be considered unreliable and ignored when performing the condition assessment.

[0263] As will be discussed in more detail herein (see e.g. Figure 18 ), the systems and methods disclosed herein can use residual analysis to check and / or support condition assessments. Residual analysis can be used to determine whether a fault identified is caused by an actual fault in the primary system. Residual analysis can be used to determine whether a fault identified is caused by an actual fault in the primary system, erroneous measurement channel(s) (such as incorrect measurement instrument settings), or other root causes.

[0264] Various embodiments of assisting with state assessment are described in more detail herein. Figure 10 As will be explained in greater detail, the supervision and / or assistance function 32 may include an assistance function that may be based on domain expertise and / or an ML model to judge events based on patterns learned a priori from previous events (e.g., data collected from IEDs, disturbance recorders, historical databases, etc.). Based on the domain knowledge and previous measurements, the assistance function may provide flags that may indicate, for example, which data should be considered unreliable and ignored when performing a condition assessment, or may indicate, for example, the probability that the condition assessment is correct (based on the ML model output).

[0265] The data processing system 20 may be operable to perform different actions in response to the status assessment and output of the supervisory and / or assist function 32. For illustration, and as explained in more detail herein, the data processing system 20 may be operable to react differently to an incoming signal, data, command, or request depending on whether the supervisory and / or assist function 32 identifies an instrument error in the secondary system.

[0266] The data processing system 20 may be operable to perform output and / or control functions in response to status assessments based on the digital twin(s) 31 and outputs of the supervisory and / or assist functions 32 .

[0267] The data processing system 20 may control a human-machine interface (HMI) to output alarms, warnings, or other information related to the status assessment via the output interface 22 .

[0268] Alternatively or additionally, the data processing system 20 may issue commands via the output interface 22 to perform corrective and / or mitigating actions in response to the status assessment and the output of the supervisory and / or assistance functions 32. The corrective and / or mitigating actions may include selectively overriding decisions made by the IED's decision logic based on a threshold-based evaluation of the measurements. This may be appropriate when the data processing system 20 determines that the anomaly is caused by an incorrect instrument (from which the IED receives its measurements) rather than a fault in the primary system. Overriding decisions made by threshold-based decision logic (which may be implemented in the IED) may also be appropriate when the data processing system 20 detects a fault that remains undetected by the threshold-based decision logic. Then, when the data processing system 20 detects a fault that remains undetected by the associated IED, the data processing system 20 may be operable to cause the switchgear to trip or trigger the IED to cause the switchgear to trip.

[0269] The data processing system 20 may also be operable to communicate with at least one other data processing system installed, for example, in or at another substation of the power grid. To this end, the data processing system 20 may provide processing results to the other data processing system via the output interface 22. The data processing system 20 may be operable to communicate in a manner compatible with or in accordance with IEC TR 61850-90 (e.g., IEC TR 61850-90-1, -2, and -3), for example in a manner compatible with the version of IEC TR 61850-90 in effect as of the filing date or priority date of the present application. The data processing system 20 may share processing results based on the state estimation and at least one support and / or assistance function with other data processing systems or national or regional control centers.

[0270] The data processing system 20 may use redundant measurements captured by redundant measuring instruments of the secondary system of the substation. Such redundant measurements may be used to check condition assessment and / or support condition assessment based on digital twins.

[0271] Figure 2 1 is a block diagram of a substation 10 including a secondary system having several measuring instruments 43, 44 for the same physical parameter. For illustration purposes, the several measuring instruments 43, 44 may measure the same current, voltage, and / or phasor at the same node (e.g., one terminal of a component of the primary system) within the primary system (e.g., multiple terminals of a component of the primary system).

[0272] The data processing system 20 may be operable to perform a digital twin-based condition assessment using at least the measurement values ​​of the measuring instrument 43 .

[0273] The data processing system 20 can be operable to use at least the measurements of the redundant measuring instruments 44 to inspect and / or assist in condition assessments performed based on the digital twin. The data processing system 20 can use the measurements of the redundant measuring instruments 44 to distinguish a fault or anomaly in the primary system from instrument errors, inadequacy of the model of the digital twin, incorrect model parameterization of the model of the digital twin, and / or other root causes of discrepancies between the measurements captured by the measuring instruments 43 and the expected behavior according to the digital twin 31.

[0274] The data processing system 20 may be operable to control the HMI 49 to output a warning alarm or other information in response to a status assessment as checked by the supervisory and / or assistance function 32 , wherein the checking is performed using at least the measurements of the redundant measuring instruments 44 .

[0275] Figure 3 Another specific example of a substation 10 is shown. In the illustrated example, the primary system includes components that carry current. The current may be current from a power grid or other power generation, transmission, and / or distribution system. The primary system may include switchgear 13, which may include circuit breakers (CBs), power switches, or other switchgear.

[0276] Current transformers (CTs) 11 and 12 can measure the current in one or more phases of a line or bus. CTs 11 and 12 can measure the same current (e.g., one or more currents in the same one or more phases) but can have different measurement characteristics, such as different accuracy, different acquisition rates, and / or different response characteristics.

[0277] The substation automation system executes the following: a first decision logic 41 that performs a first function; and a second decision logic 42 that performs a second function. The first function and the second function may be protection and / or monitoring functions. For illustration purposes, at least one of the first function and the second function may be a protection function operable to trip the CB or switch 13. The other of the first function and the second function may be a monitoring function or another protection function. For illustration purposes, the different functions may be related to distance and time domain protection, respectively, but are not limited thereto. The first decision logic 41 uses the measurement values ​​of the first current transformer 11 to perform the first function. The second decision logic 42 uses the measurement values ​​of the second current transformer 12 to perform the second function.

[0278] The data processing system 20 may be operable to perform a condition assessment based on at least one digital twin using measurements of at least one of the CTs 11, 12. The data processing system 20 may be operable to perform a supervisory and / or assist function 32 using measurements of at least another of the current transformers 11, 12.

[0279] Figure 4 4 is a flow chart of a method 50 for processing data acquired or generated by a secondary system of a substation. The method 50 may be automatically executed by the processing device 20.

[0280] At process block 51, the data processing system performs a state assessment. The state assessment is performed based on a model (also called a digital twin) that uses data acquired or generated by the secondary system of the substation as input. The state assessment can be performed based on several models that can model one or more of the following: individual components of the primary system of the substation system (140), the primary system of the substation as a whole, and the secondary system. The data received from the secondary system can include measurements acquired by the secondary system and / or event data generated by the secondary system.

[0281] At process block 52, the data processing system checks the state estimate. The data processing system may use redundant measurements included in the data received from the secondary system to check the state estimate. The data processing system may alternatively or additionally use domain knowledge data and / or a trained ML model to check the state estimate. Checking the state estimate may be performed using a dedicated function distinct from the digital twin that receives inputs other than the digital twin. This dedicated function may be implemented as a supervisory or assistive function.

[0282] At process block 53, a determination is made as to whether action is to be taken. The determination at process block 53 may be performed based on the status assessment and based on the output of the supervisory or assist function.

[0283] If no action is to be taken, the method reverts to process block 51. During field operation of the substation, the condition assessment and the checking of the condition assessment may be repeated continuously (ie, continually).

[0284] If action is to be taken, the action is performed at process block 54. The action may include an output action, wherein alarms, warnings, or other information related to the primary and / or secondary systems of the substation is output to the HMI layer. The action may alternatively or additionally include a corrective or mitigation action. The corrective or mitigation action may include overriding, for example, an IED or another protection logic based on threshold operation, thereby providing improved protection for the substation.

[0285] Will refer to Figure 5 、 Figure 6 、 Figure 7 、 Figure 8 To describe various implementations of digital twins and supervisory and / or assistance functions that may be employed by the data processing system 20 and data processing methods.

[0286] Figure 5 A functional block diagram illustrating processing performed by at least one processing circuit 30 of data processing system 20 is shown.

[0287] The data processing system 20 uses multiple first digital twins 61, 62, and 63. Each of the first digital twins 61, 62, and 63 models an associated component of the primary system. Each of the multiple first digital twins 61, 62, and 63 can model a primary system component selected from the group consisting of: a transformer, a transmission line, a busbar, a capacitor bank, a reactor, and a surge protector. Several first digital twins can model the same component of the primary system. The digital twins can provide electrical, and optionally thermal, and / or chemical modeling of the corresponding component of the primary system.

[0288] A wide variety of techniques are at the disposal of the skilled person to model the primary system components of the substation's primary system.Any known modeling technique may be employed that is executable by the data processing system 20 during field operation and preferably has a computational complexity suitable for real-time modeling.

[0289] The data processing system 20 further executes a plurality of supervisory or assisting functions 65, 66, 67. The plurality of supervisory or assisting functions 65, 66, 67 may each be associated with one of the plurality of first digital twins. The plurality of supervisory or assisting functions 65, 66, 67 may each have inputs that may include not only inputs of the corresponding digital twin, but also use additional information for the purpose of checking and / or supporting a condition assessment based on the associated digital twin. The additional information may include one or more of redundant measurements and / or domain knowledge data that may be employed to check or supplement the condition assessment performed by the associated digital twin.

[0290] The data processing system 20 may further execute an output and / or control function 33. The output and / or control function may be operable to perform actions in response to status assessments based on the plurality of digital twins 61, 62, 63 and the outputs of the plurality of supervisory or assistance functions 65, 66, 67. The actions may include outputting alarms, warnings, or other information related to the primary and / or secondary systems and HMI of the substation purchaser. The actions may also include corrective or mitigating actions taken in response to detected faults or anomalies.

[0291] As reference Figure 5 As explained, the data processing system may utilize a plurality of first digital twins and associated supervisory or assistance functions. Such processing modules are generally easy to implement, since modeling individual components of a primary system is less complex to implement than modeling at least one primary system of a substation as a whole.

[0292] Figure 6 Another functional block diagram illustrating processing performed by at least one processing circuit 30 of data processing system 20 is shown.

[0293] The data processing system may use a plurality of first component level digital twins 61, 63. The data processing system may further use a plurality of component level supervisory or assist functions 65, 67. These processing modules may be as described in reference Figure 5 Carry out as described.

[0294] The data processing system further includes a digital twin 64 of the substation's primary system. The digital twin 64 of the primary system can model electrical characteristics within the substation's primary system. The digital twin 64 of the substation's primary system can model voltages, currents, and / or phasors at all relevant nodes of the substation's primary system. The digital twin 64 of the primary system can optionally model thermal and / or mechanical characteristics. For illustration purposes, the digital twin 64 of the primary system can optionally model the temperature within the transformer tank and / or the mechanical loads on the transformer core or the transformer tank.

[0295] The data processing system further executes a substation-level supervisory function 68. The substation-level supervisory function 68 may receive output from all component-level digital twins 61, 63 and the digital twin 64 of the substation's primary system. The substation-level supervisory function 68 may receive output from its component-level supervisory or assistance functions 65, 67. The substation-level supervisory function 68 may receive additional information (such as redundant measurements taken by secondary systems) to check the consistency of modeling results or perform additional operations.

[0296] In one embodiment, the substation-level supervisory function(s) 68 may be operable to perform a root cause analysis to identify a root cause of a detected discrepancy between data acquired or generated by the secondary system and modeled results. Identifying the root cause may include distinguishing a root cause selected from the group consisting of: an anomaly in the primary system, a fault in the primary system, an incorrect setting of a measuring instrument in the secondary system, an inadequacy of at least one model employed by the digital twin, and inadequate parameterization of at least one digital twin.

[0297] As reference Figure 6 As explained, the data processing system can utilize a second digital twin that models the substation's primary system. This type of digital twin is more complex to implement than the component-level digital twins 61, 63, but can be based on known fundamental equations governing the flow of electrical potential and current, as well as thermal and mechanical responses, in a power system substation.

[0298] The second digital twin 64 and supervisory function 68 operating at the substation level (rather than limited to individual components of the primary system) provide an additional level of reliability in performing substation protection and / or monitoring functions.

[0299] Figure 7 Another functional block diagram illustrating processing performed by at least one processing circuit 30 of data processing system 20 is shown.

[0300] The data processing system may use a plurality of first component level digital twins 61, 63 and a second substation level digital twin 64. The data processing system may further use a plurality of component level supervisory or assisting functions 65, 67 and a substation level supervisory function 68. These processing modules may be as described in reference Figure 5 and Figure 6 Carry out as described.

[0301] The data processing system further includes a digital twin 69 of the substation's secondary system. The digital twin 69 of the secondary system can model the secondary system (such as its measuring instruments). The digital twin 69 of the substation's secondary system can model the response characteristics of current transformers or other measuring instruments in the secondary system. This modeling can include modeling the response characteristics. The modeling can utilize nameplate information from the measuring instruments.

[0302] The data processing system further performs a substation level supervisory function 68, which can generally be performed as described in reference to Figure 6 The substation level supervisory function(s) 68 may also receive and process the output of a digital twin 69 that models the secondary system.

[0303] The provision of a digital twin 69 of the secondary system further facilitates identifying root causes of detected discrepancies between data acquired or generated by the secondary system and modeled results. Identifying the root causes may include distinguishing root causes selected from the group consisting of: an anomaly in the primary system, a fault in the primary system, an incorrect setting of a measuring instrument in the secondary system, an inadequacy of at least one model employed by the digital twin, and inadequate parameterization of at least one digital twin.

[0304] The output and / or control function 33 may in particular use the results of the modeling of the secondary system as performed using the digital twin 69 to perform its actions. The actions may depend on the indirect causes of the identified differences between the measured values ​​and the modeling results.

[0305] Figure 5 、 Figure 6 and Figure 7 An embodiment of the data processing system 20 is shown in which component level assistance functions 65, 66, and 67 are present. However, in yet another embodiment, the data processing system need not have component level assistance functions 65, 66, and 67. Figure 8 Such a data processing system is illustrated in FIG.

[0306] Figure 8 Another functional block diagram illustrating processing performed by at least one processing circuit 30 of the data processing system 20 is shown. The substation-level supervisory function 68 may be operable in response to results obtained from the component-level digital twins 61, 63, the primary system digital twin 64, and the secondary system digital twin 69. Based on these inputs and additional information (such as redundant measurements), the substation-level supervisory function 68 may be operable to check the consistency of the modeling performed by the various digital twins and / or perform root cause analysis on detected discrepancies.

[0307] Figure 5 、 Figure 6 、 Figure 7 and Figure 8 The functional components of the data processing system 20 according to various embodiments are shown. Figure 5 、 Figure 6 、 Figure 7 and Figure 8 The operation of functional modules at the component level and at the substation level used in any one of the data processing systems.

[0308] Figure 9 is a functional block diagram of modules executed by the processing circuit 30 .

[0309] The processing circuit 30 has one or more digital twins 80. The one or more digital twins 80 may each include an electrical model 81. The one or more digital twins 80 may include optional additional models, such as thermal, mechanical and / or other models 82. As mentioned above, techniques for modeling individual components of the primary system or the primary system as a whole are available to dispatchers and need not be discussed in more detail herein. The one or more digital twins 80 may be used to implement the component-level digital twins 61, 62, 63. The one or more digital twins 80 may be used to implement the digital twin 64 of the primary system. The digital twin 69 for the secondary system generally does not need to include thermal and / or mechanical models.

[0310] The processing circuitry 30 executes a model validation function 71. The model validation function 71 may provide an output 78 that may be used by the supervisory and / or assistance function 70 or may be otherwise used to determine whether to take action. The model validation 71 may be performed based on discrepancies between the data 72 from the secondary system and the expected behavior according to the digital twin(s) 80.

[0311] The data 72 from the secondary system may include measurement values ​​and / or event data. The data 72 from the secondary system may include current and / or voltage at all relevant nodes of the primary system. The data 72 may optionally include information about the phase angle (relative to a reference) of an electrical characteristic (such as current, voltage, and / or power). The data 72 may include both the magnitude and phase angle of the corresponding electrical characteristic. The data 72 from the secondary system may include phasor measurements.

[0312] The data processing system may use the output 78 to perform output and / or control function 33. Actions may be selectively taken depending on whether the model is deemed valid, as reflected by the model validation output 78. Additionally or alternatively, decisions about what actions to take may depend on whether the model is deemed valid, as reflected by the model validation output 78.

[0313] The processing circuit 30 performs supervisory and / or assistance functions 70. The supervisory and / or assistance functions may be component level functions 63, 64, 65 or substation level functions 68. Figure 5 、 Figure 6 、 Figure 7 、 Figure 8 As explained, the supervisory and / or assistance function 70 receives a condition assessment based on at least one digital twin 80. The supervisory and / or assistance function 70 may check the validity of the condition assessment and / or may perform other processing that assists in model-based protection or monitoring performed by the data processing system 20.

[0314] The supervision and / or assistance function 70 receives a state assessment based on at least one digital twin 80. The supervision and / or assistance function may also receive results of a model validation 71.

[0315] The supervisory and / or assistance function 70 uses the additional information to check the condition assessment and / or otherwise assist the protection or monitoring functions performed by the data processing system 20. This additional information may include redundant measurements 73. The redundant measurements may be provided in the form of measurements or as event data. The redundant measurements 73 may be related to the same electrical or other physical parameters used in performing the condition assessment. The use of redundant measurements 73 is particularly preferred when the supervisory and / or assistance function 70 is a substation-level function.

[0316] Alternatively or additionally, the additional information may include domain knowledge data 74. The domain knowledge data 74 may be based on previous events in the substation or in other substations having a similar or identical configuration to the substation in question.

[0317] Alternatively or additionally, the additional information may include other data-driven inputs 75. For illustration, previous events may be used to train the ML model. Thus, historical data is encoded in the parameters of the ML model. The ML model may include an input layer that receives measurements. The ML model may have outputs that provide signatures related to the model of the primary system components of the entire primary system of the substation. Thus, the data-driven inputs 75 may include parameters of the trained ML model or the ML model as learned during training.

[0318] Alternatively or additionally, the additional information may include decision logic outputs 76 of other decision logic deployed in the substation.The decision logic outputs 76 may be decisions made by the IED, such as tripping or not tripping (inhibiting) of a switchgear.

[0319] The supervisory and / or assistance function 70 may utilize any or any combination of these additional information segments to review the state assessment or otherwise support the protection or monitoring functions performed by the data processing system 20. The data processing system may use the output 79 of the supervisory and / or assistance function 70 to perform the output and / or control function 33. Action may be selectively taken depending on whether the supervisory and / or assistance function 70 deems the state assessment consistent with any or any combination of the additional information segments 73, 74, 75, and / or 76. Additionally or alternatively, the decision as to which action to take may depend on whether the supervisory and / or assistance function 70 deems the state assessment consistent with any or any combination of the additional information segments 73, 74, 75, and / or 76. Additionally or alternatively, the decision as to which action to take may depend on the root cause of the difference between the measured value and the digital twin modeling result determined by the supervisory and / or assistance function 70 using any or any combination of the additional information segments 73, 74, 75, and / or 76.

[0320] Will refer to Figure 10 (component level functionality) and Figure 11 (Substation Level Functions) is used to describe specific implementations of function blocks executed by the processing circuitry to implement component level functions and substation level functions.

[0321] Figure 10 1 shows a functional block diagram of modules implemented by processing circuitry 30. Processing circuitry 30 may execute verification function 100 to verify the digital twin of a component (wherein verification function 100 is an illustrative implementation of verification function 71 described previously). Processing circuitry 30 may execute assistance function 90 (wherein assistance function 90 is an illustrative implementation of supervision and / or assistance function 70 described previously).

[0322] The verification function 100 can be implemented using a model-driven solution to leverage a digital twin of the component and its terminal measurements. This can be used to perform any or any combination of the following:

[0323] Component-level digital twin validation 101: The digital twin can be validated continuously (i.e., in an ongoing, optionally periodically repeated manner). This validation can be performed in real time during field operations. Steady-state and / or event data can be used to perform component-level digital twin validation 101.

[0324] - determining an internal anomaly 102 in a component to trigger an action (such as an alarm, warning or other output) 109. This may include power lines connecting substations by seamlessly exchanging measurements and other internal states via inter-substation communications (e.g. in accordance with IEC TR 61850-90-1, -2 and -3 as in force at the priority or filing date of the present application, or in a manner compatible therewith).

[0325] - Determine primary system parameters 103 that cannot be obtained from the secondary system by direct measurement. Examples of such primary system parameters include, but are not limited to, any one or any combination of the following: flux linkage of the corresponding component, magnetizing current, delta branch current, rotor angle / position, renewable energy control and operating parameters, etc. The corresponding parameters can be estimated based on data received from the secondary system. For example, the output 109 can be based on the determined parameters and can be used for monitoring, control and other network level applications, such as by converting the reference Figure 11 The network status estimation, network source equivalence estimation, etc. performed by substation-level functions described in more detail are described.

[0326] - Assess 104 component health and incipient failure modes. The health assessment may be based on electrical modeling and measurements. The health assessment may additionally be based on thermal and / or mechanical condition assessment.

[0327] The output 109 of the verification module can be obtained by referring to Figure 11 Substation level functions are described in more detail and / or used by the output and / or control functions 33 .

[0328] The assistance function 90 may be operable to assist the verification function 100 in correctly evaluating anomalies given model errors and measurement uncertainties, such as systematic errors in the instrument, noise, etc. For example, the assistance function 90 may assist the verification function 100 by supplying an event flag indicating whether the event is external or internal to a component of the primary system.

[0329] Alternatively or additionally, the assistance function 90 may be operable to supplement the verification function 100 with IED flags (such as CT saturation, etc.) so that operation of the verification function 100 may be adjusted by discarding unreliable measurements based on, for example, corresponding IED flags.

[0330] Alternatively or additionally, the assistance function 90 may be operable to use the sampled values ​​to detect scenarios where the incoming data may be unreliable, thereby leading to faster detection of measurement / model problems. This information may help avoid misjudgments made in the presence of an incorrect model of a component of the primary system.

[0331] The assistance function 90 may be operable even when substation level processing is not available to provide an additional level of safety.

[0332] The assistance function 90 may implement solutions based on domain expertise 91 and / or use an ML model 92 to judge events based on patterns learned a priori from previous events (such as data collected from IEDs, fault recorders, historian databases, etc.). The trained ML model may have an input layer that receives measurements from secondary systems and an output layer that indicates whether the measurements are reliable and / or should be discarded.

[0333] The output 109 of the verification function 100 and the assistance output 99 of the assistance function 90 may be logically combined to provide a component level protection / monitoring alarm.

[0334] Figure 11 A functional block diagram of modules implemented by processing circuitry 30 is shown. Processing circuitry 30 may execute verification function 120 to verify the digital twin of the primary system of the substation (where verification function 120 is an illustrative implementation of verification function 71 described previously). Processing circuitry 30 may execute supervisory function 110 (where supervisory function 110 is an illustrative implementation of supervisory and / or assistance function 70 described previously). Processing circuitry 30 may execute assistance function 130 (where assistance function 130 is an illustrative implementation of supervisory and / or assistance function 70 described previously).

[0335] The verification function 120 may be implemented using a model-driven solution to leverage a digital twin of the primary system (and optionally the secondary system) of the substation system 140, substation-wide measurements, and component-level processed output(s) 99, 109. This may be done to perform any one or any combination of the following:

[0336] Substation-level digital twin validation 121: The digital twin can be validated continuously (i.e., in an ongoing, optionally periodically repeated, manner). This validation can be performed in real time during field operations. Steady-state and / or event data can be used to perform component-level digital twin validation 121. Substation-level digital twin validation 121 can include oversight of component-level model validation 101.

[0337] Criticality monitoring 122. Monitoring 122 may include continuously monitoring and / or tracking criticality (measured or unmeasured parameters). Output 129 based on criticality monitoring 122 may be used to facilitate smooth substation operation. Alternatively or additionally, data processing system 20 may be operable to provide output 129 based on criticality monitoring to enable inter-substation data exchange for network-level applications and / or protection of interconnected power lines.

[0338] Tracking of component-level determinations 123, which may be based on the output(s) 99, 109 of the component-level processing. Tracking 123 may include checking that the outputs 99, 109 obtained for the various components of the primary system are consistent with each other. Tracking 123 may include conflict resolution in the event of conflicts or contentions between the results provided by different component-level functions 90, 100.

[0339] - Detecting 124 anomalies within the substation. Output 129 based on the anomaly detection 124 can be used for various actions, such as triggering protection / monitoring alarms.

[0340] The supervisory function 110 may be triggered by the substation level verification function 120. The supervisory function 110 may be operable to evaluate the legitimacy of anomalies, faults, alarms and / or warnings that may be issued by the substation level verification function 120 or the component level functions 90,100.

[0341] The supervisory function 110 may be operable to perform monitoring 111 on alarms from the component level functions 90 , 100 and assess whether the alarms are due to an anomaly (e.g., an internal fault) in a component of the primary system or due to any problems with the secondary data acquisition system and / or secondary system settings.

[0342] The supervisory function 110 may be operable to identify 112 faulty measurement channels (such as faulty instrument channels) of the secondary system, if any. The supervisory function 110 may be operable to provide estimates of the faulty channels, which may be used in the validation modules 100, 120. Various missing data replacement functions (such as hard interpolation, correlation, etc.) may be used to provide estimates of the faulty channels.

[0343] The supervisory function 110 may be operable to perform monitoring 111 and / or identification 112 on a protection time scale in order to maintain a reliable, trustworthy, and safe protection system for the entire substation.

[0344] The assistance function 130 may be operable to utilize a digital twin of the secondary system (i.e., the measuring instruments) of the substation system 140. The assistance function 130 may be operable to provide correct settings and / or calibrating meters and indicate whether there are any physical problems (e.g., open circuit, shorted connection) with the faulty instrument channel. To this end, the assistance function 130 may use redundant data obtained from various sources, such as any one or any combination of the following: process bus, IED, fault recorder, SCADA, etc. The identification 131 of incorrect instrument settings and / or the determination 132 of correct instrument settings may be performed on a slower time scale than the processing of modules 110, 120 (e.g., in seconds or minutes) because it involves calibration of the substation measuring instruments.

[0345] The component-level processing and / or substation-level processing disclosed in detail herein can be advantageously used to supplement existing (e.g., threshold-based) decision logic. The component-level processing and / or substation-level processing disclosed in detail herein can be advantageously used in power systems that include high penetrations of renewable energy and / or include AC / DC, DC / AC, or DC / DC converters; conventional protection and / or monitoring systems may not be suitable for such power system substations.

[0346] Figure 12 1 is a schematic representation of a power system substation 140. Substation 140 includes a primary system 141, a secondary system 142 including measuring instruments, and one or more protection and / or monitoring devices 143. One or more protection and / or monitoring devices 143 may each employ conventional threshold-based decision logic.

[0347] The data processing system 20 according to the present invention is provided in addition to and interfaces with conventional protection and / or monitoring devices 143. Thus, the data processing system 20 adds another layer of safety in identifying anomalies or faults, linking anomalies or faults in the primary system 141 to incorrect measuring instruments or incorrect measuring instrument settings, and preventing unnecessary downtime.

[0348] like Figure 12 As shown in , the power system may be a power system comprising a renewable energy generator 144 and one or several AC / DC, DC / AC and / or DC / DC converters 145 .

[0349] The various processing components of data processing system 20 may receive data to be processed in various ways (eg, from a data bus, a process bus, a SCADA system, or otherwise).

[0350] Figure 131 is a schematic block diagram of a data processing system 20. The data processing system 20 includes a substation-level processing module 151 that uses substation-wide measurements. The data processing system 20 also includes a component-level processing module 152 that processes measurements from terminals of corresponding components. Data can be exchanged within the substation via a data bus 153 or process bus. For example, data can be exchanged between data processing systems 20 in different substations via a data bus 154. Communication between different substations can be implemented in accordance with IEC TR 61850-90 (e.g., in accordance with IEC TR 61850-90-1, -2, and -3 as valid on the filing or priority date of this application).

[0351] Figure 14 FIG2 is a schematic block diagram of a data processing system 20 for a first substation, including a substation-level processing module 151 and a component-level processing module 152, which can exchange data via a data bus 153 or process bus. The system includes a further substation-level processing module 161 and a further component-level processing module 162 for another substation, which can exchange data via a further data bus 163 or process bus. The different substation-level processing modules 151, 161 can be operable to communicate with each other via, for example, data buses 154, 164. Communication between the different substation-level processing modules 151, 161 can be implemented in accordance with IEC TR 61850-90 (e.g., in accordance with IEC TR 61850-90-1, -2, and -3 as in effect on the filing or priority date of this application). Communication between the different substation-level processing modules 151, 161 can allow for coordination for performing protection functions (such as distance protection functions) on lines extending between the different substations.

[0352] The data processing systems and methods disclosed herein provide an enhanced approach for monitoring, protecting, and controlling substations in power systems, such as power systems that include integrated renewable energy systems. The data processing systems and methods are based on a model-driven concept. The data processing systems and methods can operate to support and supplement conventional decision logic (such as conventional IEDs) or as a standalone solution.

[0353] Various benefits are provided by the data processing system and method. The data processing system and method can provide additional intelligence to the protection / monitoring system of a substation. The data processing system and method can protect the system from unseen errors in instrumentation that could trigger spurious relay operations. The data processing system and method can also maintain validated models of substation components and regularly calibrate instrumentation, while providing monitoring and protection functions as its primary functions.

[0354] The data processing system functionality can be operable to utilize centralized (e.g., process bus and station bus) substation-wide (phasor and time domain) data at various frame rates. Illustrative frame rates are 25 Hz, 1 kHz, 4.8 kHz, and 14.4 kHz, but are not limited thereto. The data processing system and method can utilize data redundancy to implement the disclosed functionality. Current communications in modern substations, combined with processing performed by the data processing system, provide the desired functionality.

[0355] Although the data processing system 20 can be implemented as a substation-specific system, it does not need to be physically installed at or in the substation. For illustration purposes, the data processing system 20 can be provided remotely from the substation to implement the processes disclosed in detail herein.

[0356] Various specific illustrative operations of the data processing system and method according to the embodiments will be described next.

[0357] Figure 15 is a flow chart of method 170. Method 170 may be automatically performed by data processing system 20.

[0358] At process block 171, data from a secondary system of a substation is processed. Processing the measurements may include using a digital twin to perform a condition assessment. Processing the measurements may include using redundant information included in the received data to perform a check on the condition assessment and / or otherwise support the condition assessment.

[0359] At process block 172, fault or anomaly detection is performed. Fault or anomaly detection may be performed to supplement processing by conventional protection and / or monitoring devices (such as conventional IEDs). Alternatively, fault or anomaly detection may be performed as a standalone solution that does not require the presence of conventional protection and / or monitoring devices. Fault or anomaly detection may include using redundant information included in the received data to perform checks on condition assessments and / or otherwise support condition assessments. Fault or anomaly detection may include identifying the root cause of discrepancies between measured values ​​and modeled results for components of the substation and the station's primary system and / or the substation's primary system.

[0360] At process block 173, an action is performed in response to the detected fault or anomaly. The action may include an output action (such as an alarm outputting a warning) or a control action. The control action may include a mitigation and / or corrective action that is operable to counteract the detected fault or anomaly.

[0361] Figure 16 is a flow chart of method 175. Method 175 may be automatically performed by data processing system 20. Method 175 may be performed to implement fault or anomaly detection at process block 172.

[0362] At process block 176, the measured values ​​are compared with results obtained from at least one digital twin. The comparison may include determining a fit error between the electrical characteristics expected from the at least one digital twin and the received measured values. The fit error may be detected as a function of time.

[0363] At process block 177, a difference is detected. The difference can be detected in a time-resolved manner based on the fitting error.

[0364] Figure 17 is a flow chart of method 180. Method 180 may be automatically performed by data processing system 20. Method 180 may be performed to implement discrepancy detection at process block 277. Method 180 may be performed to examine condition assessments based on measurement data.

[0365] At process block 181, measurement data is received. The measurement data may be or may include time domain data (e.g., a time series of measurements) and / or frequency domain data. The measurement data may include samples of electrical characteristics (such as voltage, current, and / or power) in the time domain and / or frequency domain.

[0366] At process block 182, a fitting error is determined. The fitting error quantifies the difference between the measured data and the expected behavior based on, for example, the digital twin(s). The fitting error can be determined as a function of time (e.g., as a function of sampling time) and / or as a function of frequency. The fitting error can be determined based on a metric (e.g., L2 or other L metric) that quantifies the deviation of the measured value from the characteristics based on the digital twin.

[0367] At process block 183, a fitting error is detected. A threshold comparison may be performed on the fitting error to detect anomalies. Anomaly detection may trigger the output of an anomaly flag, which the data processing system 20 (e.g., supervisory and / or assist functions) may use to perform or trigger specific processing operations (such as a residual analysis or several residual analyses). Figure 18 1 is a flow chart of method 185. Method 185 may be automatically executed by data processing system 20. Method 185 may be executed to perform root cause analysis. Method 185 may be selectively executed once an anomaly has been detected.

[0368] At process block 186, a residual analysis may be performed. The residual analysis may provide a residual error for each of a set of equation tags each associated with a measurement value. Each of the set of equation tags may be associated with a specific different measurement instrument in the secondary system.

[0369] The measurement(s) with the highest residual error may be identified at process block 187. From this, a possible root cause for the identified analysis may be identified.

[0370] Process blocks 186 and 187 can be repeated for different scenarios (e.g., using different data sources). For illustration purposes, when repeating the residual analysis and its evaluation, data from a specific measuring instrument (e.g., a specific current transformer in a substation) may be discarded. Thus, it can be determined whether the specific measuring instrument (e.g., a specific current transformer in a substation) or its settings is the root cause of the detected anomaly. The changes in the residual errors obtained in different repetitions of process blocks 186 and 187 can be used to check whether the residual errors meet acceptability criteria (e.g., all residual errors fall below an acceptability threshold) when the erroneous measurement channels are discarded.

[0371] At process block 188, supervisory function results may be provided based on the results of the residual analysis. For illustration, based on method 185 detecting that the identified anomaly is caused by an incorrect measuring instrument or its setting(s) rather than a true problem in the primary system, tripping of the switch or CB may be prevented. For further illustration, the HMI may be controlled to notify the operator of the root cause of the anomaly, where the root cause is identified in method 185.

[0372] Figure 19 is a block diagram of (at least a portion of) the processing system 30. Component-level digital twin verification 101 may be performed using, for example Figure 17 The first fitting error is determined by the method 180. Based on the determined first fitting error, the component-level digital twin verification 101 may provide an anomaly flag to the supervisory function 110.

[0373] Substation-level digital twin verification121 can be done using e.g. Figure 17 The method 180 of is used to determine the second fitting error. Based on the determined second fitting error, the substation-level digital twin verification 121 may provide one or several additional anomaly flags to the supervisory function 110. If an anomaly is detected, several additional anomaly flags may be provided based on: (i) all available measurements and (ii) only some of the available measurements. For illustration purposes, a modification instruction from the supervisory function 110 may cause the substation-level digital twin verification 121 to repeat the fitting error determination while discarding some of the available measurements. Since the measurements include redundant measurements (from, for example, different current transformers measuring the same current), analysis can still be performed in a meaningful way.

[0374] The supervisory function 110, which may be operable at the substation level, may cause the substation-level digital twin verification 121 to repeat the fitting error determination while discarding some of the available measurements. This may be done in response to (a plurality of) anomaly flags received from at least one of the digital twin verifications 101, 121.

[0375] Figure 20 and Figure 21 The operation of the data processing system when an incipient fault is detected is illustrated. Figure 20 The processing illustrated in (which shows the fitting error 191 as determined by the component level processing module) does not require substation level processing.

[0376] As discussed in the present disclosure, component-level processing may include component-level model validation functionality (such as component-level model validation functionality 100) and component-level assistance functionality (such as component-level assistance functionality 90), which may be provided in addition to IEDs or other conventional protection and / or monitoring devices. A model-driven component-level model validation functionality with access to all terminal measurements of a protected component / zone provides a protection / monitoring technique that is complementary to conventional IED functionality. For example, the component-level model validation functionality may be able to detect incipient faults (e.g., by maintaining a sensitivity threshold on the model-measurement fit error 191). Figure 20 ), such as slowly developing insulation failure in transformer windings. Even when the signal 192 in the IED does not reach the triggering threshold of the IED ( Figure 21 ), which also provides detection of faults or anomalies.

[0377] Model-based component-level model verification capabilities can also track other anomalies (which are not faults) such as magnetic saturation, core overexcitation, and magnetizing inrush current. Figure 22 Illustrated is a fitting error 193 as experienced by a component-level model validation function that allows for detection of anomalies based on deviations in the expected current (digital twin) and measured current (data from the secondary system) drawn by the transformer during energization. Figure 23 Illustrated are fitting errors 194 as experienced by the data processing system 20 , which are indicative of an anomaly or fault.

[0378] The data processing system 20 and data processing methods may similarly be used to detect other sporadic faults not detected by conventional IEDs (such as sporadic failures of capacitor cells in the upper and lower legs of a capacitor bank).

[0379] While component level model verification functionality may be operable to supplement bay level IED functionality, the proposed processing at component level and substation level may be operable to provide further enhanced protection, monitoring and / or supervision in a substation.

[0380] Figure 24 is a flow chart of method 195. Method 195 may be automatically performed by data processing system 20.

[0381] At process block 196, the substation-level processing module receives a comparison of component-level measurements (e.g., electrical measurements at the terminals of the component) with expected characteristics determined based on the component-level digital twin. The comparison may indicate an alarm, such as that issued by the component-level processing.

[0382] At process block 197, substation-wide measurements are used in conjunction with the substation digital twin to check for component-level processed alarms. The substation-wide measurements may include redundant measurements that are utilized to determine whether any alarms issued by component-level processing are considered valid.

[0383] A root cause analysis may optionally be performed at process block 198. The root cause analysis may determine whether the anomaly or fault identified by the component level process corresponds to an anomaly or fault in the primary system, is due to an instrument error, or has another cause.

[0384] Action is taken at process block 199. The action may include, but is not limited to, an output action, a corrective action, and / or a mitigation action.

[0385] Will refer to Figures 25 to 34 The following diagram illustrates the operation of a data processing system employing both component-level and substation-level processing. For purposes of illustration and understanding, two scenarios will be considered: in one scenario, an internal component fault occurs during phase 'a', and in the other, an ammeter measuring current at the device terminals experiences measurement errors during phase 'a' due to a disconnected CT secondary (open CT condition). Such current measurements (e.g., those of current transformer CT1) can serve as a source for component-level model validation functionality at both the IED and component levels. At the substation level, an additional measurement is available for the component current, which will be referred to as CT2.

[0386] Case 1: Failure in stage 'a'

[0387] In this case, the component-level model validation function showed high model-measurement fit errors after failure 201 ( Figure 25 ). An anomaly flag indicating an anomaly at the component level may be issued. The IED may also observe the fault based on its own processing (not shown) and may not issue an auxiliary IED flag (not shown). For the analysis presented here, dynamic state estimates and normalized residuals may be used to implement verification and supervision functions. However, the disclosed technology can be based on other processing techniques that can utilize digital twins and measurements for the same purpose.

[0388] If the component level assistance function only uses intelligence from the IED, no further steps will be implemented because no auxiliary IED flag is issued and the IED picks up the fault due to its own logic.

[0389] At the substation level, the substation level model verification function also experienced a failure due to a high fitting error for the substation level model verification function 202 ( Figure 26 ). The substation level supervisory function will then further perform an alarm legitimacy check. It is found that CT1a (ie, the current measurement from CT1 for phase a) is the measurement with the highest residual error. Figure 27 An exemplary result of the residual analysis is shown in . The substation-level supervisory function then requests the substation-level model verification function to perform its processing as before, but without using the CT1a measurements. The data processing system observes that the substation-level model verification function fitting error 203 continues to remain high even without CT1a ( Figure 28 The residual analysis is done by the substation level supervisory function at this stage ( Figure 29 ) reveals that CT2a has a high residual error, which indicates a fault in phase 'a'. Therefore, in this case, the substation level module concludes that there is indeed a fault in phase 'a' in the primary system.

[0390] Case 2: Open CT in Phase 'a'

[0391] In this case, the component-level model validation function shows a high model-measurement fit error after the open CT event and issues a component-level anomaly flag. The IED can also detect this condition, albeit with some delay. The IED will issue the open CT flag in phase 'a'. Therefore, the event remains unresolved at the component-level module.

[0392] At the substation level, after the open CT event, the substation level model validation function also experienced a high fitting error for the substation level model validation function. The substation level supervision function then performed an alarm legitimacy check. It was found that CT1a was the measurement with the highest residual error ( Figure 32 ). The substation-level supervision function then causes the substation-level model verification function to perform its processing without using the CT1a measurement data. It is observed that without CT1a, the substation-level model verification function fitting error 204 recovers to the pre-event level ( Figure 30 The residual analysis is done by the substation level supervisory function at this stage ( Figure 33) reveals that there are no measurements with high residual errors, indicating that there is an error in the measurement of phase 'a' current by CT1. This result can then be used together with the open CT flag from the IED to avoid spurious operation of the protection relay.

[0393] The substation-level model validation function can also provide a reliable estimate for CT1a based on the redundancy of the information (measurements and models) as a byproduct. The substation-level supervisory function can be operable to use the output of the substation-level supervisory function (e.g., a flag indicating that CT1a is faulty) and the output of the substation-level model validation function (e.g., the estimate for CT1a) and an additional measurement of the phase 'a' current (i.e., the redundant measurement CT2a) to determine that the root cause of the event is indeed that CT1 is open in phase 'a'.

[0394] In the above description, we have assumed that the component-level assistance function took no action. However, it is indeed possible to perform local supervision of the component-level model verification function decisions with the help of the component-level assistance function. Assuming that the component-level assistance function uses intelligence from the IED (i.e., in this case, the IED assists in flagging an open CT), it may cause the component-level model verification function to perform a model-measurement fit by removing the CT1a measurement from consideration. In that case, it is observed that the component-level model verification function fit error 205 falls back to its pre-event value ( Figure 31 ) and it can be concluded at the component level itself that this is not a case of a fault within the device, but rather a case of erroneous measurement. This is also confirmed by the residual analysis, which can be performed at the component level or at the substation level ( Figure 34 ). Additionally, the component level decision will be validated at the substation level module as described above.

[0395] Note that this component-level assistance implementation relies solely on IED auxiliary flags. If the IED fails to detect an open CT condition, the component level will conclude that the event is a fault. This illustrates the additional benefit gained by performing checks at the substation level. Furthermore, performing component-level corrections by removing measurements from the verification step may not always be feasible, as this may result in a mathematically underdetermined system. The level of measurement redundancy available at the component level may be limited compared to the substation level.

[0396] As an alternative or in addition to the above embodiments, the component level assistance functionality may be implemented independently of the IED based on, for example, domain expert knowledge or ML or processing techniques (such as residual analysis) on limited component level information (in a manner similar to that described above for the above substation level supervision functionality).

[0397] In an open CT condition (or any other similar condition when a problem with the measurements exists and is identified), the IED is blocked until the condition is reversed, which may take time. During this period, any fault in the component will not be detected by the IED.

[0398] Thus, the data processing system 20 and data processing method provide the added bonus that CT2 can be treated as a source of current measurements until CT1 is fixed. Thus, any faults during the period when the IED is blocked can still be detected by the component-level model verification function. Furthermore, the data processing system can be operable so that if the IED is not hardwired to CT1, it can be configured to CT2. This improves responsiveness to faults during periods when CT1 is unavailable as a source.

[0399] Figure 35 is a flow chart of method 210. Method 210 may be automatically performed by data processing system 20.

[0400] At process block 211 , a fitting error between the measured values ​​and the expected values ​​based on the model is detected.

[0401] At process block 212, the measurement with the largest fitting error is identified. The fitting error can be determined using any metric that quantifies the difference between the time-dependent measurement (e.g., voltage, current, and / or other electrical characteristics) and the expected time-dependent characteristics calculated using the digital twin modeling method(s).

[0402] At process block 213, action is taken based on which measurement has the largest fitting error. The action may include, but is not limited to, output action, corrective action, and / or mitigation action.

[0403] Will refer to Figure 36 、 Figure 37 、 Figure 38 、 Figure 39 and Figure 40 to illustrate the operation of the method.

[0404] To illustrate, assume that an IED (called CT1) configured to measure current is incorrectly set. Incorrect CT ratio settings could be the result of human error. The component-level model verification function is fed from CT1, but the component-level model verification function is correctly set to read from CT1. In this case, the IED picks up when the component is energized to carry current. Because the measurement is correct, no other IED auxiliary flags exist. However, the component-level model verification function shows no fitting error. Therefore, there is a conflict at the component level regarding the exact nature of the event. This illustrates that different root causes can be identified using the techniques disclosed herein.

[0405] To illustrate further, consider another scenario where both the IED and component-level model verification functions are configured for CT1. The relay engineer incorrectly records the CT ratio settings for CT1, and therefore both the IED and component-level model verification function inputs for CT1 are incorrectly set. In this case, the IED spuriously picks up the component once it is energized to carry current. Since the measurements are correct, there is no IED auxiliary flag. The component-level model verification function also exhibits a high fitting error ( Figure 36 ) to consistently indicate faults in the primary system. Component-level modules draw fault conclusions.

[0406] As mentioned earlier, on the other hand, the substation level module has an additional measuring instrument for the current passing through the component (called CT2). The substation level model verification function implementation is based on both CT1 and CT2. Therefore, the substation level model verification function also shows a high fitting error ( Figure 37 ). Residual error analysis under substation level supervision function ( Figure 39 ) reveals that all current measurements associated with CT1 have high errors. This may then cause the substation-level model validation function to perform its processing in a manner that removes the CT1 measurements. It is observed that when CT1 is removed from consideration, the substation-level model validation function fitting error falls back to its pre-event value ( Figure 38 ), thus indicating that the event was triggered due to an erroneous measurement value in CT1. The residual error dropped to an acceptable level ( Figure 40 ).

[0407] Under the substation-level assistance function, the output of the substation-level supervision function (a flag indicating CT1 as an error), the output of the substation-level model verification function (the estimated value of CT1), and the output of the redundant measurement CT2 can be used to infer that CT1 is indeed incorrectly scaled relative to the actual value of the phase current in all three phases. This indicates that the hidden failure is an incorrect CT ratio setting.

[0408] Various effects and advantages are achieved by the data processing systems and methods according to embodiments. The data processing systems and methods provide enhanced techniques for detecting anomalies and / or faults. The data processing systems and methods can be used to supplement existing protection and / or monitoring systems with devices that can reduce potential errors made by human engineers during protection system configuration and / or commissioning. The data processing systems and methods can be used to provide enhanced fault or anomaly detection, given the higher penetration of renewable energy sources, incorrect instrumentation, or other unseen errors in data acquisition systems.

[0409] Although embodiments have been described with reference to the accompanying drawings, modifications and variations may be implemented in other embodiments.

[0410] For illustration purposes, although embodiments have been described in which the substation is a power system substation, the apparatus, systems, and methods may be used in conjunction with substations of other industrial and / or utility systems, such as as part of an industrial automation and control (IACS) system.

[0411] While embodiments have been described in which substation-level processing based on substation-wide measurements is combined with supervisory and / or assistance functions at the component level, these techniques may also provide benefits when used alone.

[0412] While embodiments have been described in which dynamic state estimates and normalized residuals may be used to implement validation and supervision functions, the disclosed techniques may utilize other processing techniques that may utilize digital twins and measurements to achieve the same purpose.

[0413] Embodiments may be used in conjunction with a substation of an electrical grid, such as a medium voltage or high voltage grid. Embodiments may be used in conjunction with power generation, transmission, and / or distribution systems, including renewable energy systems, such as DERs.

[0414] The description and embodiments herein illustrate aspects and embodiments of the present invention and should not be considered restrictive—the claims define the protected invention. In other words, although the present invention has been illustrated and described in the drawings and the foregoing description, such illustrations and descriptions are to be considered illustrative and not restrictive. Various mechanical, component, structural, electrical, and operational changes may be made without departing from the spirit and scope of the description and claims. In some cases, well-known circuits, structures, and techniques are not shown in detail to avoid obscuring the present invention. Therefore, it will be understood that changes and modifications may be made by one of ordinary skill in the art within the scope and spirit of the following claims. In particular, the present invention encompasses additional embodiments having any combination of features from the different embodiments described above and below.

[0415] The present disclosure also encompasses all additional features individually shown in the figures, although they may not be described in the preceding or following description. Moreover, individual alternatives to the embodiments described in the figures and description, as well as individual alternatives to their features, may be disclaimed from the subject matter of the present invention or from the disclosed subject matter. The present disclosure includes subject matter consisting of features defined in the claims or embodiments, as well as subject matter comprising features.

[0416] The term "comprising" does not exclude other elements or process blocks, and the indefinite article "a" or "an" does not exclude a plurality. A single unit or process block may fulfill the functions of several features recited in the claims. The mere fact that certain measures are recited in mutually different dependent claims does not indicate that a combination of these measures cannot be used to advantage. Components described as coupled or connected may be directly coupled electrically or mechanically, or they may be indirectly coupled via one or more intermediate components. Any reference signs in the claims should not be construed as limiting the scope.

[0417] The machine-readable instruction code may be stored / distributed on a suitable medium, such as an optical storage medium or solid-state medium supplied with or as part of other hardware, but may also be distributed in other forms, such as via a wide area network or other wired or wireless telecommunications system. In addition, the machine-readable instruction code may also be a data structure product or signal for embodying a specific method (such as, the method according to the embodiment).

Claims

1. A data processing system (20) for a substation system (10; 140), wherein: The substation system (10; 140) includes a primary system (141) and a secondary system (11, 12; 43, 44; 142), the primary system (141) includes a plurality of primary system components (13), and the data processing system (20) includes: at least one interface (21) operable to receive data acquired or generated by the secondary system (11, 12; 43, 44; 142) and relating to the primary system (141); and At least one processing circuit (30) operable to: continuously and during field operation of the primary system (141), performing a condition assessment on at least one primary system component of the plurality of primary system components or on a primary system (141) of the substation system (140) based on the received data and at least one digital twin (31; 61-64), and Executing at least one function (32; 65-68) that performs a check on or supports the state assessment based on one or more of: redundant information (73) included in the received data, domain knowledge data (74) and / or a trained machine learning (ML) model.

2. The data processing system (20) according to claim 1, in, The at least one digital twin includes a plurality of first digital twins (65-67), each of the plurality of first digital twins (65-67) models an associated one of the plurality of primary system components.

3. The data processing system (20) according to claim 2, in, The at least one function comprises a plurality of first component level functions (65-67), optionally wherein each of the first component level functions (65-67) is operable to: causing, based on the inspection, an adjustment to a process executed based on the first digital twin for performing the condition assessment, and / or The state assessment is supervised based on domain knowledge and / or trained machine learning (ML) models.

4. The data processing system (20) according to claim 3, in, The adjusting includes discarding unreliable portions of the data when performing the state assessment.

5. The data processing system (20) according to any one of the preceding claims, in, The at least one digital twin includes a second digital twin (64) that models at least the primary system (141).

6. A data processing system (20) according to claim 5, when dependent on any one of claims 2 to 4, in, The second digital twin (64) is operable to receive outputs generated by the plurality of first digital twins.

7. The data processing system (20) according to claim 5 or claim 6, in, The second digital twin is operable to receive and process redundant information included in the data.

8. The data processing system (20) according to claim 7, in, The redundant information includes redundant measurement values ​​(73).

9. The data processing system (20) according to claim 7 or claim 8, in, The redundant information includes at least two measured values ​​of the same substation parameter acquired or generated by different measuring instruments of the secondary system (11, 12; 43, 44; 142).

10. The data processing system (20) according to any one of claims 7 to 9, in, The at least one processing circuit (30) is operable to perform a substation status assessment based on the second digital twin (64).

11. The data processing system (20) according to any one of the preceding claims, in, The at least one digital twin comprises a third digital twin (69) modeling at least the secondary system (11, 12; 43, 44; 142).

12. The data processing system (20) according to any one of the preceding claims, in, The at least one processing circuit (30) is operable to identify an instrument error in the secondary system (11, 12; 43, 44; 142) based on the inspection.

13. The data processing system (20) according to any one of the preceding claims, in, The processing circuitry (30) is operable to identify a root cause of a discrepancy between the data and the at least one digital twin (31) based on the inspection.

14. The data processing system (20) according to claim 13, in, The at least one processing circuit (30) is operable to distinguish, based at least on the redundant information (73), a root cause selected from the group consisting of: an anomaly of at least one primary system component of the primary system (141); an inappropriateness of a model or model parameterization of the at least one digital twin; an anomaly of a measuring instrument (11, 12; 43, 44).

15. The data processing system (20) according to any one of the preceding claims, in, The processing circuit (30) is operable such that the at least one function (32; 65-68) performs a check on the status assessment based on one or more of: redundant information (73) included in the received data, domain knowledge data (74) and / or a trained machine learning (ML) model, wherein the processing circuit (30) is operable to identify a root cause of a difference between the data and the at least one digital twin (31) based on the inspection to distinguish anomalies and / or faults in the primary system from anomalies and / or faults in the secondary system.

16. The data processing system (20) according to any one of the preceding claims, in, The at least one interface (21) is operable to receive an IED logic output from at least one IED (143) of the substation system (140), and wherein the at least one processing circuit (30) is operable to check the IED logic output based on the at least one digital twin (31) and based on the received data, the at least one digital twin modeling at least one primary system component or the primary system (141).

17. The data processing system (20) according to claim 16, in, The at least one interface (21) is further operable to receive and process auxiliary data indicative of a problem with the data.

18. The data processing system (20) according to any one of the preceding claims, in, The at least one processing circuit (30) is operable to trigger at least one action based on the inspection, optionally wherein the at least one action comprises one or more of an output action, a corrective action, a mitigation action via a human-machine interface (49) HMI, further optionally wherein the at least one action comprises a corrective or mitigation action affecting a power system protection function and / or a power system monitoring function.

19. A power system substation, comprising: A primary system (141), the primary system comprising a plurality of primary system components, Secondary system (11, 12; 43, 44; 142), said secondary system being operable to acquire or generate data relating to said primary system (141), and A data processing system (20) as claimed in any preceding claim, operable to receive and process acquired or generated data.

20. A data processing method for a substation system (10; 140), wherein: The substation system (10; 140) includes a primary system (141) and a secondary system (11, 12; 43, 44; 142), the primary system (141) includes a plurality of primary system components, and the data processing method includes: receiving data acquired or generated by the secondary system (11, 12; 43, 44; 142) and related to the primary system (141); and Continuously and during field operation of the primary system (141), performing a condition assessment on at least one primary system component of the plurality of primary system components or on the primary system of the substation system (140) based on the received data and at least one digital twin (31); and Executing at least one function that performs a check on the state assessment or supports the state assessment based on at least one of: redundant information (73) included in the received data, domain knowledge (74) and / or a trained machine learning (ML) model.

Images