Processing system and method for controlling and / or monitoring an electrical power system

Abstract

In order to process alarms generated based on monitoring data (68) for an electrical power system, a processing system (60) includes at least one processing circuit (70) operable to: identify an alarm flooding, identify at least one alarm not included in the alarm flooding, and transmit the at least one alarm to the processing circuit (70). And generating and providing an output (69) based on the identified alarm flooding and the identified at least one alarm not included in the alarm flooding.

Description

Technical Field

[0001] Embodiments of the present invention relate to processing systems and methods for controlling and / or monitoring power systems. Specifically, embodiments of the present invention relate to processing systems and methods operable to process alarms generated in response to monitoring data obtained for a power system. Background Technology

[0002] Power system operators handle a large volume of alarms generated by corresponding network monitoring and control components. Alarms for transmission and distribution systems allow for the detection of various conditions within the grid and the implementation of appropriate actions to prevent, correct, and / or control undesirable grid states. Inefficient alarm systems can pose significant challenges to efficient alarm handling. The situation worsens when at least one alarm flood occurs. In an alarm flood, the alarm rate exceeds what can be reasonably managed according to the definition of alarm flood (e.g., more than 10 alarms per 10 minutes). This can jeopardize the normal operation of individual grid components and even compromise the stability of the power system. Therefore, intelligent alarm handling is crucial for ensuring reliable grid operation.

[0003] However, compared to process industries, intelligent alarm handling is an underdeveloped topic for power systems. Existing methods and tools developed for process industries cannot be easily applied to power systems. The fundamental reason behind this limitation lies in the significant dynamic differences between industrial processes and power systems. Power systems exhibit significantly faster and more complex dynamics, rendering direct application of methods ineffective.

[0004] US 10 928 815 B2, US 2021 / 0150877 A1 and US 8 554 714 B2 disclose systems or methods that can be used in conjunction with alarms.

[0005] Further technological improvements are still needed to allow for appropriate actions related to power system control and / or management in response to alarms. Summary of the Invention

[0006] The object of this invention is to provide methods and / or processing systems that provide enhanced techniques for controlling and / or managing power systems based on alarms generated in response to monitoring data. In particular, such methods and / or processing systems are needed that can provide alarm processing associated with the complex dynamics and behavior of modern power systems, such as transmission and / or distribution networks.

[0007] According to exemplary embodiments, the method and processing system described in the independent claims are provided. Dependent claims define preferred or advantageous embodiments.

[0008] According to one aspect of the present invention, a processing system for controlling and / or monitoring a power system is provided. The processing system includes at least one interface operable to receive monitoring data during operation of the power system. The processing system includes at least one processing circuit operable to: identify alarm flooding among alarms generated based on the received monitoring data; identify at least one alarm not belonging to the alarm flooding; and generate an output based on the identified alarm flooding and the identified at least one alarm not belonging to the alarm flooding.

[0009] This processing system achieves various effects and advantages. It is operable to detect not only alarm flooding but also at least one alarm that does not belong to the alarm flooding. It is operable to generate an output that depends on both the identified alarm flooding and the identified at least one alarm that does not belong to the alarm flooding. Therefore, the processing system is operable to take appropriate action when alarm flooding exists and at least one alarm that does not belong to the alarm flooding exists, even when the at least one alarm that does not belong to the alarm flooding is issued during the time period when the alarm flooding alarm is being generated.

[0010] The at least one processing circuit can be operated to identify at least one alarm that does not belong to the alarm flood, wherein the at least one alarm that does not belong to the alarm flood is issued in a manner that overlaps with the alarm flood in time (i.e., within the same time interval as the alarm flood).

[0011] Therefore, the processing system is operable such that appropriate action can be taken even when at least one alarm not belonging to the alarm flood is issued within the same time interval as the alarm flood. In the above situation, alarms not belonging to the alarm flood are particularly prone to being missed. The processing system reduces or eliminates the risk of missing such alarms(s) that are not belonging to the alarm flood but are issued simultaneously with the alarm flood.

[0012] The at least one processing circuit can be operated such that processing these alarms includes: processing the alarm type and the time when the alarms were issued, in order to identify at least one alarm that is not part of the alarm flood.

[0013] Therefore, the dynamics of the power system can be considered to identify the alarm flood and at least one alarm that does not belong to the alarm flood.

[0014] The at least one processing circuitry is operable such that processing the alarms includes: determining at least one alarm occurrence rate of the generated alarms; and evaluating the at least one alarm occurrence rate based on an alarm flooding definition to identify the alarm flooding and at least one alarm that does not belong to the alarm flooding.

[0015] Therefore, alarm flooding and at least one alarm that does not belong to the alarm flooding can be identified in an efficient manner, thus making the technology suitable for online alarm processing (e.g., for processing alarms in real time when they are issued).

[0016] The definition of alarm flooding can include alarm patterns of one or more alarm types, where the alarm pattern is based on the occurrence rate of the corresponding alarm type in the alarm flooding cluster.

[0017] This allows for the efficient identification of alarm overload, making the technology suitable for online alarm processing (e.g., for real-time processing of alarms as they are issued).

[0018] An alarm flooding definition can include an alarm flooding pattern for each of several different alarm flooding clusters. An alarm flooding pattern can include a definition of the occurrence rate (e.g., a range of occurrence rates) of one or more different alarm types within the corresponding cluster. An alarm flooding pattern can be independent of the temporal relationship between individual alarms of different alarm types (e.g., the time delay between alarms of the first alarm type and alarms of the second alarm type in an alarm flood).

[0019] Therefore, alarm flooding can be efficiently identified using an occurrence-based definition. The independence of the temporal relationship between alarms of different alarm types (e.g., the time delay between alarms of the first alarm type and alarms of the second alarm type in alarm flooding) not only facilitates the efficient identification of alarm flooding and alarms (multiple) that do not belong to any alarm flooding, but is also particularly applicable to power systems that exhibit more complex dynamics compared to, for example, industrial process systems.

[0020] The output may include output provided via a human-machine interface (HMI).

[0021] Therefore, the processing system can provide information about generated alarms via the HMI in a manner dependent on both alarm flooding and at least one alarm not belonging to the alarm flooding, even when at least one alarm not belonging to the alarm flooding is issued within the same time period as the alarm flooding alarms are being generated. The processing system is operable to mitigate the risk of at least one alarm being missed when at least one alarm not belonging to the alarm flooding is issued simultaneously with the alarm flooding alarms. The processing system is operable to enable appropriate action to be taken when at least one alarm not belonging to the alarm flooding is issued within the same interval as the alarms belonging to the alarm flooding.

[0022] The at least one processing circuit can be operated to control the HMI such that alarms that do not belong to the alarm flood and alarms that belong to the identified alarm flood are output in a visually different manner.

[0023] Therefore, this processing system can mitigate the risk of at least one alarm that is not part of an alarm flood being missed when it is issued simultaneously with an alarm that is part of an alarm flood. Consequently, this processing system can mitigate the risk of potential hazards or instability in the power system.

[0024] The at least one processing circuit can be operated to control the HMI such that alarms that do not belong to the alarm flood and alarms that belong to the identified alarm flood are output at different scales.

[0025] Therefore, this processing system can mitigate the risk of at least one alarm that is not part of an alarm flood being missed when it is issued simultaneously with an alarm that is part of an alarm flood. Consequently, this processing system can mitigate the risk of potential hazards or instability in the power system.

[0026] The at least one processing circuit can be operated to control the HMI so that alarms belonging to the identified alarm flood are output in a manner with low cognitive load.

[0027] Therefore, this processing system can mitigate the risk of at least one alarm that is not part of an alarm flood being missed when it is issued simultaneously with an alarm that is part of an alarm flood. Consequently, this processing system can mitigate the risk of potential hazards or instability in the power system.

[0028] The at least one processing circuit can be operated to perform alarm classification in response to the detection of alarm flooding, and to identify at least one alarm that does not belong to the alarm flooding based on the alarm classification.

[0029] Therefore, the processing system can automatically identify at least one alarm that does not belong to the alarm flood, even when the at least one alarm is issued within the same time interval as the alarms in the alarm flood.

[0030] The at least one processing circuit can be operated such that alarm classification may include cluster analysis.

[0031] Therefore, the processing system can identify at least one alarm that does not belong to the alarm flooding, even if that at least one alarm is issued within the same time interval as the alarms(s) in the alarm flooding. Identification can be performed efficiently, enabling the processing system to operate and perform online alarm processing functions.

[0032] The at least one processing circuit can be operated to classify alarms based on the alarm occurrence rate.

[0033] Therefore, the processing system can efficiently identify at least one alarm that does not belong to alarm flooding, thereby enabling the processing system to operate and perform online alarm processing functions.

[0034] The at least one processing circuit can be operated to enable alarm classification based on unsupervised clustering techniques.

[0035] Therefore, the processing system can efficiently identify at least one alarm that does not belong to alarm flooding.

[0036] Unsupervised clustering techniques can include k-means clustering.

[0037] Therefore, the processing system can efficiently identify at least one alarm that does not belong to alarm flooding.

[0038] The at least one processing circuit can be operated such that alarm classification may include cluster analysis using cluster definitions determined in a data-driven manner.

[0039] Thus, the processing system can use criteria (especially cluster characteristics) to identify alarm flooding and at least one alarm that does not belong to the alarm flooding. These criteria are determined in a data-driven manner and are therefore based on objective standards, rather than relying (entirely or partially) on human expert knowledge.

[0040] The at least one processing circuit may further operate to identify at least one additional alarm flood among the generated alarms and generate additional output, the additional output including information aggregated by the at least one processing circuit from the identified alarm flood and the at least one additional alarm flood.

[0041] Therefore, information suitable for asynchronous review of alarm flooding and at least one additional alarm flooding (e.g., for performing root cause analysis) can be provided under the control of this processing system.

[0042] The aggregated information may include counts of alarms with different priorities (e.g., low, medium, or high priority) and / or information on the timing of alarm occurrences within an alarm flood or at least one additional alarm flood.

[0043] Therefore, information suitable for asynchronous review of alarm flooding and at least one additional alarm flooding (e.g., for performing root cause analysis) can be provided under the control of this processing system.

[0044] The aggregated information may include counts of alarms with different priorities (e.g., low, medium, or high priority), which are represented in the form of a histogram (where different histogram bars are provided in association with the time intervals at which the associated alarm flooding occurs).

[0045] Therefore, information suitable for asynchronous review (e.g., for performing root cause analysis) can be provided under the control of this processing system.

[0046] The aggregated information can include details about the total number of alerts for each event. This information can be overlaid on a histogram representation.

[0047] Therefore, information suitable for asynchronous review (e.g., for performing root cause analysis) can be provided under the control of this processing system.

[0048] The processing system can be operated to: control the human-machine interface (HMI) to output aggregated information, control the HMI to enable selection of alarm flooding or at least one additional alarm flooding, and provide information on alarm priority and / or occurrence time based on the selection.

[0049] Therefore, information suitable for asynchronous review (e.g., for performing root cause analysis) can be provided under the control of this processing system to facilitate root cause analysis and / or control logic modification.

[0050] The at least one processing circuit can be further operated to perform root cause determination on the identified alarm flood, and further generate an output based on the result of the root cause determination.

[0051] Therefore, alarm messages can be used by the processing system to automatically determine the root cause of alarm overload.

[0052] The at least one processing circuit can be further operated to generate an output based on a root cause determination.

[0053] Therefore, alarm messages can be used by the processing system to automatically determine the root cause of alarm proliferation, and use the root cause(s) to generate output (e.g., triggering actions of the main system or auxiliary system affecting the power system).

[0054] The at least one processing circuit can be operated to perform root cause determination based on natural language processing (NLP) of the alarm message associated with the generated alarm.

[0055] Therefore, alarm messages can be used by the processing system to automatically determine the root cause of alarm proliferation, and use the root cause(s) to generate output (e.g., triggering actions of the main system or auxiliary system affecting the power system).

[0056] The at least one processing circuit can be operated such that the output can include at least one control command, which causes at least one action to be performed on the main system equipment and / or auxiliary system devices of the power system.

[0057] Therefore, the processing system can trigger at least one action based on the identified alarm flood and at least one alarm that does not belong to the alarm flood. This allows for the triggering of multiple corrective actions, such as multiple mitigation actions, based on the existence of alarm flood and based on (multiple) alarms that do not belong to the alarm flood.

[0058] The at least one processing circuit can be further operated to generate an alarm based on the received monitoring data.

[0059] Therefore, this processing system can serve as an integrated system for alarm generation and alarm management.

[0060] The processing system can be operated to perform alarm processing on power generation systems, transmission systems, and / or distribution systems.

[0061] Therefore, the processing system can be operated in conjunction with power systems that are deficient when using conventional alarm processing techniques.

[0062] The processing system may be or may include an alarm processing system capable of operating to perform alarm processing on the transmission network and / or distribution network, the alarm processing system including the at least one processing circuit.

[0063] Therefore, the processing system can be operated in conjunction with power systems that are deficient when using conventional alarm processing techniques.

[0064] The processing system may be or may include a control subsystem capable of operating to control the power system based on the output.

[0065] Thus, the processing system is able to operate to control the power system in a manner that takes appropriate action in response to at least one alarm that is not part of an alarm flood, even when the at least one alarm is issued within the same time interval as the alarm flood (e.g., simultaneously).

[0066] According to another aspect of the present invention, an electric power system is provided. The electric power system may include a main system device, an auxiliary system device of a monitoring data acquisition and control (SCADA) system associated with the main system device, and a processing system for controlling and / or monitoring the electric power system, as described in any aspect or embodiment, the processing system being operable to receive monitoring data from the auxiliary system device.

[0067] This power system achieves various effects and advantages. The processing system is operable to detect not only alarm flooding but also at least one alarm that does not belong to the alarm flooding. The processing system is operable to generate an output that depends on both the identified alarm flooding and the identified at least one alarm that does not belong to the alarm flooding. Therefore, the processing system is operable to take appropriate action when alarm flooding exists and at least one alarm that does not belong to the alarm flooding exists, even when the at least one alarm that does not belong to the alarm flooding is issued during the period when the alarm flooding alarm is being generated.

[0068] The processing system can be operated to trigger at least one action based on alarm flooding and at least one alarm that does not belong to the alarm flooding.

[0069] Therefore, the processing system can trigger at least one action based on the identified alarm flood and at least one alarm that does not belong to the alarm flood. This allows for the triggering of multiple corrective actions, such as multiple mitigation actions, based on the existence of alarm flood and based on (multiple) alarms that do not belong to the alarm flood.

[0070] The SCADA system may include a control subsystem, wherein the processing system is operable such that the control subsystem performs at least one action in response to an output generated by the processing system.

[0071] Therefore, the processing system can trigger at least one action via the SCADA control subsystem based on the identified alarm flood and at least one alarm that does not belong to the alarm flood. This allows for the triggering of multiple corrective actions, such as multiple mitigation actions, based on the presence of alarm flood and based on (multiple) alarms that do not belong to the alarm flood.

[0072] The power system may include at least one human-machine interface (HMI). The processing system may operate to provide outputs to control at least one HMI.

[0073] Thus, the processing system helps operators perform their complex tasks, for example, by reducing the risk that at least one alarm not included in the alarm flood will be ignored.

[0074] HMIs can be installed in control centers, such as in the control rooms of power transmission and / or distribution networks.

[0075] Thus, the processing system helps grid operators perform their complex tasks, for example, by reducing the risk that at least one alarm not included in the alarm flood will be ignored.

[0076] The power system may include a power generation system, a transmission system, and / or a distribution system. The processing system can be operated to perform alarm processing on the transmission network and / or the distribution network.

[0077] Therefore, the processing system can be operated in conjunction with power systems that are deficient when using conventional alarm processing techniques.

[0078] The power system may include a transmission network and / or a distribution network.

[0079] Therefore, the processing system can be operated in conjunction with power systems that are deficient when using conventional alarm processing techniques.

[0080] The power system may include a high-voltage direct current (HVDC) system. The power system may include inverters / converters that incorporate valves based on insulated-gate bipolar transistors (IGBTs) or thyristors.

[0081] Therefore, the processing system can be operated in conjunction with power systems where alarm processing is often particularly challenging, as HVDC systems typically include redundant implementations of both main system equipment and auxiliary system devices to ensure continued operation in the event of, for example, maintenance or asset failure.

[0082] According to another aspect of the present invention, a method for controlling and / or monitoring a power system is provided. The method includes receiving monitoring data by a processing system during operation of the power system. The method includes performing alarm processing by the processing system, the alarm processing comprising: identifying alarm flooding among alarms generated based on the received monitoring data; identifying at least one alarm that does not belong to the alarm flooding; and generating an output based on the identified alarm flooding and the identified at least one alarm that does not belong to the alarm flooding.

[0083] This method achieves various effects and advantages. The processing system is capable of detecting not only alarm flooding but also at least one alarm that does not belong to the alarm flooding. The processing system generates output based on both the identified alarm flooding and the identified at least one alarm that does not belong to the alarm flooding. Thus, when alarm flooding exists and at least one alarm that does not belong to the alarm flooding exists, appropriate actions are triggered, even when at least one alarm that does not belong to the alarm flooding is issued during the time period when the alarm flooding is being generated.

[0084] The optional features of the methods that can be used in embodiments of the invention, and the effects thereto, correspond to features that are interpreted in relation to the processing system for controlling and / or monitoring the power system and in relation to the power system according to the embodiments.

[0085] This method can be performed by or using a processing system for controlling and / or monitoring a power system, or by or using a power system according to any aspect or embodiment disclosed herein. The processing system for controlling and / or monitoring a power system can operate to perform the methods of any aspect or embodiment disclosed herein.

[0086] The power system may include a power generation system, a power transmission system, and / or a power distribution system.

[0087] Therefore, the results of alarm processing performed by the processing system are used in conjunction with power systems that have deficiencies when using traditional alarm processing techniques.

[0088] The power system may include a transmission network and / or a distribution network.

[0089] Therefore, the results of alarm processing performed by the processing system are used in conjunction with power systems that have deficiencies when using traditional alarm processing techniques.

[0090] The power system may include a high-voltage direct current (HVDC) system. The power system may include inverters / converters that incorporate valves based on insulated-gate bipolar transistors (IGBTs) or thyristors.

[0091] Therefore, the results of alarm processing performed by the processing system are used in connection with power systems where alarm processing is often particularly challenging, as HVDC systems typically include redundant implementations of both main system equipment and auxiliary system devices to ensure continued operation in the event of, for example, maintenance or asset failure.

[0092] The method can be a method for controlling a power system, wherein the output includes at least one control command that causes at least one action to be performed on the main system equipment and / or auxiliary system devices of the power system.

[0093] Therefore, control operations can be automatically performed based on whether alarm flooding exists and based on (multiple) alarms that do not belong to the alarm flooding category.

[0094] Identifying alarm flooding can include assessing alarms and their occurrence times based on the definition of alarm flooding.

[0095] This allows for the efficient identification of alarm overload, enabling these alarms to be processed online as they are issued (i.e., near real-time or real-time).

[0096] Alarm flooding is defined as an alarm flooding characteristic that can include several types of alarm flooding. Alarm flooding characteristics can include the occurrence rate of alarms of at least one alarm type, for example, the occurrence rate of alarms of at least two different alarm types (e.g., alarms associated with different primary or secondary power system assets).

[0097] This allows for the efficient identification of the types of alarms that are flooding the system.

[0098] Alarm flooding definitions can include alarm flooding definitions obtained from historical alarms in a data-driven manner.

[0099] Therefore, alarm overload can be defined objectively, thereby mitigating the risk of human error.

[0100] Alarm flooding definitions can include those obtained by performing cluster analysis on historical alarms.

[0101] Therefore, alarm proliferation can be defined in an objective, data-driven manner, thereby mitigating the risk of human error.

[0102] Alarm flooding definitions can include those obtained by performing cluster analysis on historical alarms.

[0103] Therefore, alarm overload can be defined objectively, thereby mitigating the risk of human error.

[0104] The definition of alarm flooding can include an alarm flooding definition obtained by performing an unsupervised clustering analysis on historical alarms, which can include k-means clustering.

[0105] Therefore, alarm overload can be defined objectively, thereby mitigating the risk of human error.

[0106] An alarm flooding definition can include an alarm flooding pattern for each of several different alarm flooding clusters. An alarm flooding pattern can include a definition of the occurrence rate (e.g., a range of occurrence rates) of one or more different alarm types within the corresponding cluster. An alarm flooding pattern can be independent of the temporal relationship between individual alarms of different alarm types (e.g., the time delay between alarms of the first alarm type and alarms of the second alarm type in an alarm flood).

[0107] Therefore, alarm flooding can be efficiently identified using an occurrence-based definition. The independence of the temporal relationship between alarms of different alarm types (e.g., the time delay between alarms of the first alarm type and alarms of the second alarm type in alarm flooding) not only facilitates the efficient identification of alarm flooding and alarms (multiple) that do not belong to any alarm flooding, but is also particularly suitable for use with power systems that exhibit more complex dynamics compared to, for example, industrial process systems.

[0108] The definition of alarm flooding can include alarm message-based data of alarm messages that are included in the corresponding alarm flooding.

[0109] This spurred the development of root cause analysis.

[0110] Data based on alert messages may include the results of processing the alert messages using natural language processing (NLP).

[0111] This prompted an analysis of the root causes of the alarm overload.

[0112] This method may include generating an alarm flooding definition by processing historical alarms (which may include alarm flooding patterns and, optionally, NLP results of alarm messages).

[0113] Therefore, an objective definition of alarm proliferation can be generated.

[0114] This method may include performing control and / or monitoring operations by a processing system or by a power system using the generated alarm flooding definition.

[0115] According to another aspect of the invention, a method is provided for generating alarm flood definitions for use in the above-described method or by the processing system to control and / or monitor a power system. The method includes retrieving historical alarm data by a computing system (e.g., by a server or distributed server system) and processing the historical alarm data to generate alarm flood definitions. Processing the historical alarm data may include: identifying alarm floods in the historical data (using, for example, a criterion based on occurrence rate), determining the temporal patterns of alarms in the identified alarm floods, establishing alarm flood definitions based on the determined time patterns, and providing these alarm flood definitions to a processing device for controlling and / or monitoring the power system.

[0116] Therefore, the characteristics of alarm flooding can be defined in an objective way, thereby mitigating the risk of human error. Alarm flooding is defined to identify alarm flooding and at least one alarm not included in the alarm flooding, thereby allowing appropriate actions to be taken to mitigate the risk of at least one alarm not included in the alarm flooding being ignored.

[0117] Alarm flooding definitions can include those obtained by performing cluster analysis on historical alarms.

[0118] Therefore, computing systems can define alarm proliferation in an objective, data-driven manner, thereby mitigating the risk of human error.

[0119] Alarm flooding definitions can include those obtained by performing cluster analysis on historical alarms.

[0120] Therefore, the computing system can define alarm proliferation in an objective way, thereby reducing the risk of human error.

[0121] The definition of alarm flooding can include an alarm flooding definition obtained by performing an unsupervised clustering analysis on historical alarms, which can include k-means clustering.

[0122] Therefore, the computing system can define alarm proliferation in an objective way, thereby reducing the risk of human error.

[0123] An alarm flooding definition can include an alarm flooding pattern for each of several different alarm flooding clusters. An alarm flooding pattern can include a definition of the occurrence rate (e.g., a range of occurrence rates) of one or more different alarm types within the corresponding cluster. An alarm flooding pattern can be independent of the temporal relationship between individual alarms of different alarm types (e.g., the time delay between alarms of the first alarm type and alarms of the second alarm type in an alarm flood).

[0124] Therefore, alarm flooding can be efficiently identified using an occurrence-based definition. The independence of the temporal relationship between alarms of different alarm types (e.g., the time delay between alarms of the first alarm type and alarms of the second alarm type in alarm flooding) not only facilitates the efficient identification of alarm flooding and alarms (multiple) that do not belong to any alarm flooding, but is also particularly suitable for use with power systems that exhibit more complex dynamics compared to, for example, industrial process systems.

[0125] The definition of alarm flooding can include alarm message-based data of alarm messages that are included in the corresponding alarm flooding.

[0126] This spurred the development of root cause analysis.

[0127] Data based on alert messages may include the results of processing the alert messages using natural language processing (NLP).

[0128] This prompted an analysis of the root causes of the alarm overload.

[0129] The computing system can be included in the processing system or communicatively coupled to the processing system.

[0130] This allows alarm flooding definitions to be made available to the processing system.

[0131] According to another aspect of the invention, there is provided the use of alarm flooding definitions obtained by a method for determining alarm flooding definitions, wherein these alarm flooding definitions are used by a processing system or processing method according to an aspect or embodiment to identify alarm flooding and at least one alarm that does not belong to the alarm flooding.

[0132] Therefore, the processing system and method can use an alarm flood definition obtained in a data-driven manner to process alarms.

[0133] According to another embodiment, machine-readable instruction code is provided that, when executed by at least one programmable circuit, causes the at least one programmable circuit to perform a processing method, control method, or operation method according to any aspect or embodiment disclosed herein.

[0134] According to another embodiment, a non-transitory storage medium is provided that stores machine-readable instruction code, which, when executed by at least one programmable circuit, causes the at least one programmable circuit to perform a processing method, control method, or operation method according to any aspect or embodiment disclosed herein.

[0135] The following list describes embodiments of the present invention:

[0136] Example 1: A processing system for controlling and / or monitoring a power system, the processing system comprising: at least one interface operable to receive monitoring data during operation of the power system; and at least one processing circuit operable to: identify alarm flooding among alarms generated based on the received monitoring data, identify at least one alarm not belonging to the alarm flooding, and generate an output based on the identified alarm flooding and the identified at least one alarm not belonging to the alarm flooding.

[0137] Example 2: The processing system as described in Example 1, wherein the output includes output provided via a human-machine interface (HMI), and the at least one processing circuit is operable to control the HMI such that the at least one alarm not belonging to the alarm flood and the alarm belonging to the identified alarm flood are output in a visually different manner.

[0138] Example 3: The processing system as described in Example 2, wherein the at least one processing circuit is operable to control the HMI such that the at least one alarm not belonging to the alarm flood and the alarm belonging to the identified alarm flood are output at different scales.

[0139] Example 4: A processing system as described in Example 2 or Example 3, wherein the at least one processing circuit is operable to control the HMI such that alarms belonging to the identified alarm flood are output in a manner with low cognitive load compared to at least one alarm not belonging to the identified alarm flood.

[0140] Example 5: A processing system as described in any of the preceding embodiments, wherein the at least one processing circuit is operable to perform alarm classification in response to the detection of alarm flooding, and to identify at least one alarm that does not belong to the alarm flooding based on the alarm classification.

[0141] Example 6: The processing system as described in Example 5, wherein the at least one processing circuit is operable such that the alarm classification includes cluster analysis.

[0142] Example 7: A processing system as described in Example 5 or Example 6, wherein the at least one processing circuit is operable such that the alarm classification is based on the alarm occurrence rate.

[0143] Example 8: A processing system as described in any of the preceding embodiments, wherein the at least one processing circuit is further operable to identify at least one additional alarm flood among the generated alarms and to generate additional output, the additional output including information aggregated by the at least one processing circuit from the identified alarm flood and the at least one additional alarm flood.

[0144] Example 9: A processing system as described in any of the preceding embodiments, wherein the at least one processing circuit is further operable to perform root cause determination on the identified alarm flood, and further generates the output based on the result of the root cause determination.

[0145] Example 10: The processing system as described in Example 9, wherein the at least one processing circuit is operable to perform the root cause determination based on natural language processing (NLP) of the alarm message associated with the generated alarm.

[0146] Example 11: A processing system as described in any of the preceding embodiments, wherein the at least one processing circuit is operable such that the output includes at least one control command, the at least one control command causing at least one action to be performed on the main system equipment and / or the auxiliary system equipment of the power system.

[0147] Example 12: A processing system as described in any of the preceding embodiments, wherein the at least one processing circuit is further operable to generate the alarm based on the received monitoring data.

[0148] Example 13: A power system comprising: a main system device, an auxiliary system device of an auxiliary system associated with the main system device, and a processing system as described in any of the preceding embodiments, the processing system being operable to receive the monitoring data from the auxiliary system device.

[0149] Example 14: A method for controlling and / or monitoring a power system, the method comprising: receiving monitoring data by a processing system during operation of the power system, performing alarm processing by the processing system, the alarm processing comprising: identifying alarm flooding among alarms generated based on the received monitoring data, identifying at least one alarm that does not belong to the alarm flooding, and generating an output based on the identified alarm flooding and the identified at least one alarm that does not belong to the alarm flooding.

[0150] Example 15: The method as described in Example 14, wherein the method is performed by a processing system as described in any one of Examples 1 to 12 or by a power system as described in Example 13.

[0151] Embodiments of the present invention achieve various effects and advantages. For illustration, the processing systems and methods according to the embodiments provide enhanced techniques for processing alarms that mitigate the risk of alarms that do not belong to an alarm flood occurring within the same time interval (e.g., simultaneously) as alarms belonging to the alarm flood.

[0152] The processing system and method are capable of compressing a large number of alarms during alarm overload to facilitate efficient alarm processing.

[0153] The processing system and method can be operated to identify and output the root cause of the alarm proliferation.

[0154] The processing system and method are capable of distinguishing alarms that do not belong to alarm flooding and generating output based on these alarms.

[0155] This processing system and method provide intelligent alarm management for the power system domain.

[0156] The processing system and method can be used in conjunction with a power grid or its subsystems, but are not limited thereto. Attached Figure Description

[0157] Embodiments of the present invention will be described with reference to the accompanying drawings, in which similar or identical reference numerals denote elements having similar or identical configurations and / or functions.

[0158] Figure 1 This is a block diagram of the processing system.

[0159] Figure 2It is a schematic diagram of a power system including a processing system.

[0160] Figure 3 This is a flowchart of the method.

[0161] Figure 4 This is a block diagram of the processing system.

[0162] Figure 5 This is a block diagram of the processing system.

[0163] Figure 6 This is a simplified diagram illustrating the occurrence of an alarm.

[0164] Figure 7 This is a flowchart of the method.

[0165] Figure 8 This is a simplified diagram illustrating the clustering of alarm overload.

[0166] Figure 9 This is a flowchart of the method.

[0167] Figure 10 This is a flowchart of the method.

[0168] Figure 11 It is a schematic representation of the output provided through the human-machine interface under the control of the processing system.

[0169] Figure 12 This is a flowchart of the method.

[0170] Figure 13 It is a schematic representation of the output provided through the human-machine interface under the control of the processing system.

[0171] Figure 14 It is a schematic representation of the output provided through the human-machine interface under the control of the processing system.

[0172] Figure 15 This is a flowchart of the method.

[0173] Figure 16 This is a flowchart of the method.

[0174] Figure 17 This is a flowchart of the method.

[0175] Figure 18 It is a schematic representation of a power system.

[0176] Figure 19 It is a schematic representation of a power system.

[0177] Figure 20 This is a flowchart of the method. Detailed Implementation

[0178] Embodiments of the invention will be described with reference to the accompanying drawings. In the drawings, similar or identical reference numerals denote elements having similar or identical configurations and / or functions.

[0179] The embodiments relate to processing systems and methods capable of processing alarms generated based on monitoring data for power systems.

[0180] Although embodiments will be described in detail primarily in relation to power systems, including transmission and / or distribution systems, the embodiments are not limited thereto.

[0181] The processing method and system disclosed herein are operable to process alarms generated based on monitoring data obtained for a power system. The processing system is operable to: identify alarm flooding among alarms generated based on the received monitoring data, identify at least one alarm that does not belong to the alarm flooding, and generate output based on the identified alarm flooding and the identified at least one alarm that does not belong to the alarm flooding.

[0182] As used herein, the term "electric system" encompasses generation systems, transmission systems, and / or distribution systems, such as a power grid or a portion thereof.

[0183] As used herein, the term "power grid" encompasses both transmission and / or distribution networks. A power grid may include both transmission and distribution networks.

[0184] As used herein, the term "monitoring data" encompasses both measurement results and event-based data. Monitoring data may include measurement results and / or event-based data (e.g., GOOSE messages). Measurement results may include any one or any combination of measurements relating to electrical variables of the power grid or its components (e.g., voltage, current, phasor measurement unit (PMU) measurements, and / or frequency measurements), switchgear status (e.g., open / closed status of isolators and / or circuit breakers), thermal variables (e.g., temperature measurements), and insulation parameters (e.g., dissolved gas analysis (DGA) concentration of insulating oil or other insulating fluids, moisture content of insulation, etc.). Monitoring data may be received by a processing system from auxiliary system components of the power system (e.g., intelligent electronic devices (IEDs)).

[0185] As used herein, the term "alarm" can encompass alarms with various priorities (e.g., priorities assigned to a range of ordinal values, such as low / medium / high priority). A technician may obtain, for example, the technology for generating alarms from the applicant's commercial products. Technology for generating alarms based on monitoring data may include comparing the monitoring data with nominal operating conditions defined for the power system. A variety of technologies are available to the technician for determining whether an alarm should be issued. More advanced technologies, such as those performing power flow calculations to predict potential critical situations, may be used. Where such technologies are available to the technician, the technician may use any of these known technologies to implement alarm generation based on monitoring data.

[0186] As used herein, the term "alarm flooding" refers to a sequence of alarms that meets at least one of the criteria for defining alarm flooding. Alarm flooding can be identified based on the number of alarms generated within a given time period reaching or exceeding an occurrence rate threshold. This number of alarms per time period is also referred to herein as the "occurrence rate," and it should be understood that this number may involve averaging over that time period (i.e., the total number of alarms divided by the length of the time period).

[0187] Processing systems and methods can perform alarm flooding identification based on at least one alarm flooding definition generated in a data-driven manner. As used herein, the terms "alarm flooding definition" and / or "alarm flooding characteristic" can include one or more alarm occurrence-based characteristics for identifying the presence of alarm flooding and optionally for assigning alarm flooding to the cluster associated with the "alarm flooding definition" and / or "alarm flooding characteristic".

[0188] According to the method and processing system of the present invention, the processing system is operable to process alarms generated based on monitoring data. Processing these alarms may include processing at least one time in which alarms are issued (optionally processing in association with the alarm type of the alarm) to determine whether alarm flooding exists and to identify at least one alarm not included in the alarm flooding. The occurrence rate of alarms (e.g., alarms of one or more alarm types) may be used in the processing, wherein the time when the alarm is triggered can be used to determine the occurrence rate.

[0189] The techniques disclosed herein are operable to mitigate the risk that an alarm may remain ignored and / or fail to take appropriate action in response to an alarm when it occurs within a time interval where alarm flooding exists. The techniques disclosed herein can be used for online processing of alarms while they are being issued. The techniques disclosed herein are not limited to this, but can also be used, for example, for offline processing of alarm sequences, which can be used for root cause analysis and / or to determine what modifications are needed to the decision logic of the power system or at least one of its auxiliary devices.

[0190] The processing system can use alarm flooding definitions during the inference phase. Alarm flooding definitions can be based on alarm flooding clusters established in a data-driven manner during the learning phase. During the learning phase, historical alarm flooding data is used to create alarm flooding clusters. This may include assigning each alarm flooding to clusters with similar alarm flooding characteristics (e.g., protection events prior to circuit breaker (CB) opening / closing). Once flooding clusters have been formed, an overall alarm flooding pattern can be extracted for each cluster. As an optional further step, natural language processing (NLP) techniques can be used to analyze the alarm text messages for each cluster to identify possible root causes or common characteristics (topics). Both the alarm flooding clusters and their corresponding cluster flooding patterns, along with (if determined) topics, are stored for use during the inference phase. The learning and inference phases can be performed by the same processing system or by different systems, wherein the processing system is operable to obtain alarm flooding clusters and their corresponding cluster flooding patterns, along with (if determined) topics, for use in inference.

[0191] During the inference phase, alarm flooding definitions (e.g., information about identified alarm flooding clusters) can be used to detect whether an alarm belongs to a specific flooding cluster. To this end, the processing system can operate to assign an incoming set of alarms to one of the previously identified flooding clusters and then compare it with the corresponding alarm flooding pattern. Alarms in the incoming set that are not part of the extracted cluster pattern are classified as non-flooded alarms and are used to trigger actions different from those applied in response to flooding. The remaining alarms (alarms that are part of the flooding) can be processed differently from the (multiple) alarms that do not belong to the alarm flooding. For illustration, alarms belonging to the alarm flooding can be compressed in the alarm list to reduce cognitive load. Optionally, the subject of the compressed alarms, as obtained from the assigned flooding clusters, can be used to perform actions, such as presenting them to an operator via a human-machine interface (HMI).

[0192] Figure 1 A block diagram representation of processing system 60 is shown. Processing system 60 includes at least one interface 61. At least one interface 61 is operable to receive monitoring data 68 acquired for the power system. Monitoring data may include measurement results. Measurement results may include any one or any combination of measurements related to electrical variables of the power grid or its components (e.g., voltage, current, phasor measurement unit (PMU) measurements, and / or frequency measurements), thermal variables (e.g., temperature measurements), insulation parameters (e.g., dissolved gas analysis (DGA) concentration of insulating oil or other insulating fluids, insulation moisture content, etc.). Monitoring data may be received by the processing system from auxiliary system components of the power system (e.g., intelligent electronic devices (IEDs)). As a supplement to or alternative to the measurement results, monitoring data 68 may include event-based data (e.g., GOOSE messages).

[0193] The processing system 60 may include an HMI 62. The HMI 62 may include an HMI installed in a transmission system control room and / or a distribution system control room. The HMI 62 may operate to output alarms generated during field operations of the power system in response to received monitoring data.

[0194] Processing system 60 may include storage system 63 or be communicatively interfaced with storage system 63 (e.g., when storage system 63 provides definitions of several alarm flood clusters accessible to grid operators). Storage system 63 may store definitions of alarm flood clusters. The definitions of alarm flood clusters may be obtained using a data-driven approach (e.g., by performing cluster analysis on historical alarm patterns). Storage system 63 may also store data defining the desired operating range of the power system, which can be used to determine when received monitoring data triggers an alarm.

[0195] The processing system 60 includes at least one processing circuit 70. In the illustrated embodiment, at least one processing circuit 70 is operable to perform alarm generation 71. Alarm generation 71 can determine what type of alarm to issue (e.g., single-phase ground fault, two-phase ground fault, overvoltage condition, overcurrent condition, other fault types) and optionally determine the severity of the alarm. Alarm generation 71 can process received monitoring data 68 to determine whether and when to issue an alarm. To this end, alarm generation 71 can compare the power system state reflected by the monitoring data 68 with the nominal operating range of power system components and / or infer trends from the monitoring data 68 to predict potential future emergencies. Technicians can obtain alarm generation techniques from, for example, commercially available products from the applicant. A wide variety of alarm generation techniques can be used by technicians to implement alarm generation.

[0196] Although Figure 1 The processing system 60 is shown to be operable to perform online processing of alarms when alarm generator 71 issues an alarm; however, the processing system 60 may also be alternatively or additionally operable to process alarms offline. In the latter case, alarm generator 71 is not required (but may still be) as part of the processing system 60.

[0197] At least one processing circuit 70 is operable to perform alarm processing 72. Alarm processing 72 may include processing alarms in association with the time they are triggered. Alarm processing 72 may include comparing alarm emission patterns as a function of time with the definition of alarm flooding clusters. Thus, alarm processing 72 is operable to determine which alarms in a series of alarms generated by alarm generator 71 belong to alarm flooding, and which alarm(s) do not belong to any alarm flooding (as defined by various alarm flooding clusters).

[0198] At least one processing circuit 70 is operable to perform interface control 79 in response to an identified alarm flood and an identified alarm(s) not belonging to any alarm flood. Interface control 79 can be operable to control at least one interface 61 and / or HMI 62 to provide output based on the identified alarm flood and the identified alarm(s). This can be done in various ways. At least one processing circuit 70 can be operable to control at least one interface 61 to provide output including control command 79 based on the identified alarm flood and the identified alarm(s) not belonging to any alarm flood. Control command 79 can be operable to cause changes in the main system equipment and / or auxiliary system devices of the power system to implement actions not only according to the alarm flood but also according to at least one identified alarm not belonging to any alarm flood. This allows corrective actions or other actions to be triggered by processing system 60. Alternatively or additionally, at least one processing circuit 70 can be operable to control HMI 62 to output alarms belonging to the alarm flood and at least one alarm not belonging to any alarm flood in different ways. For illustration, at least one processing circuit 70 can be operated to control HMI 62 such that alarms belonging to alarm flooding can be provided in a spatially compressed or otherwise manner, thereby reducing cognitive load, compared to alarms that do not belong to any alarm flooding.

[0199] Figure 2 This is a schematic representation of a power system 10 according to an embodiment. The power system 10 includes a main system 20. The main system 20 may be or may include a generation system, a transmission system, and / or a distribution system, such as a transmission network and / or a distribution network, or a portion thereof. The main system 20 may include at least a portion of a power grid.

[0200] The main system 20 may include switching equipment (such as switches or circuit breakers (CB)) 22, transformers 21, and other main system equipment. The power system 10 may have renewable energy penetration, wherein the main system 20 includes renewable energy resources 24, 25, and energy storage systems (ESS) (such as battery storage systems 23). The main system 20 may also include charging infrastructure 26. Such components in non-traditional AC power systems increase the dynamic complexity of the power system. In this context, handling system 60 is particularly useful, but not limited to this use case.

[0201] System 10 includes an auxiliary system 40. The auxiliary system 40 includes multiple devices 41, 42, 43, 44, 45, and 46. The auxiliary system 40 includes measuring instruments 47 and 48. At least some of the devices 41, 42, 43, 44, 45, and 46 of the auxiliary system 40 are operable to execute decision logic to implement control over components of the main system 20 (such as switchgear 22 or transformer 21). At least some of the devices 41, 42, 43, 44, 45, and 46 of the auxiliary system 40 are operable to execute decision logic based on measurement results (such as measurement results received from measuring instruments 47 and 48). The multiple auxiliary system devices 41, 42, 43, 44, 45, and 46 may include a merging unit (MU) operable to provide monitoring data to the processing system 60. Alternatively or additionally, the auxiliary system devices 41, 42, 43, 44, 45, and 46 may be operable to provide event-based messages as part of the monitoring data to the processing system 60. Examples of the auxiliary system devices 41, 42, 43, 44, 45, and 46 include protective relays or other IEDs operable to perform distance protection, time protection, or other protection functions. The measuring instruments of the auxiliary system 40 may include a current transformer 47, a voltage transformer 48, a phasor measurement unit, a frequency measurement unit, or other measuring instruments, such as those operable to measure multiple temperatures, insulation condition-related data (e.g., dissolved gas analysis (DGA) concentration in insulating oil or other insulating fluids, or insulation humidity).

[0202] The power system 10 may optionally include a redundancy system 50 that provides redundant implementation for at least some of the functions performed by devices 41, 42, 43, 44, 45, and 46 of the auxiliary system 40. The redundancy system 50 can be implemented in various ways. For illustration, dedicated redundant devices 51 and 52 may be associated with one of the devices of the auxiliary system 40 in a one-to-one correspondence. Other implementations are also possible. For illustration, a centralized redundancy system may be present, which provides redundancy for functions performed by several devices among the auxiliary system devices 41, 42, 43, 44, 45, and 46.

[0203] System 10 includes a communication system 55. Communication system 55 may include a communication network. Communication system 55 may include multiple communication links through which devices of auxiliary system 40 can communicate with each other and with processing system 60. Communication can be performed using communication devices such as gateway device 58.

[0204] As described in more detail herein, processing system 60 is operable to process alarms generated in response to monitoring data. Processing system 60 may provide output based on identified alarm floods and alarms not belonging to any of the alarm floods. This output may include control commands for controlling, for example, auxiliary system devices 41, 42, 43, 44, 45, 46 and / or redundant system devices 51, 52 and / or HMI 62.

[0205] The alarm flood cluster definition, which can be generated in a data-driven manner, can be used by the processing system 60 to identify alarm floods (multiple) and one or more alarms that do not belong to any alarm flood. The definition of an alarm flood cluster can be generated in a data-driven manner, for example, by processing historical alarm data (e.g., processing historical alarms in association with the time when a series of historical alarms were triggered).

[0206] Figure 3 This is a flowchart of method 90. Method 90 can be executed automatically by processing system 60.

[0207] At process block 91, processing system 60 identifies alarms belonging to alarm flooding. Processing system 60 can use alarm flooding cluster definitions to determine whether an alarm includes alarms belonging to alarm flooding based on alarm patterns generated as a function of time. Process block 91 may also optionally include determining the cluster to which the alarm flooding belongs. The latter can be used for further analysis and establishing appropriate corrective actions.

[0208] At process block 92, processing system 60 identifies one or more alarms that do not belong to any alarm flood. Processing system 60 can use alarm flood cluster definitions in association with a series of alarms generated in response to monitoring data to determine which of the generated alarms do not belong to any alarm flood.

[0209] At process block 93, processing system 60 causes an action to be performed based on the identified alarm flood and at least one alarm not included in any alarm flood. Process block 93 may include an action that mitigates the risk that at least one alarm not included in any alarm flood may be ignored if it occurs within the same time interval as one or more alarm floods.

[0210] The identification at process blocks 91 and 92 can be based at least on the alarm occurrence rate. The identification at process blocks 91 and 92 can use flood cluster definitions, which define alarm flood clusters based on alarm occurrence rates and independently of the temporal relationships (such as time delays) between alarms with different alarm types. Thus, particularly efficient identification is possible.

[0211] The identification at process blocks 91 and 92 can be based at least on the alarm occurrence rate. The identification at process blocks 91 and 92 can use flood cluster definitions, which define alarm flood clusters based on alarm occurrence rates and independently of the temporal relationships (such as time delays) between alarms with different alarm types. Thus, particularly efficient identification is possible.

[0212] Figure 4 This is a block diagram representation of a processing system 60 according to an embodiment. In terms of configuration and / or functionality, it is consistent with previously referenced... Figure 1 The components explained are denoted by the same reference numerals.

[0213] Storage system 63 stores flooding characteristics 64 for various types of alarm flooding. Flooding characteristics 64 can define flooding patterns for various alarm flooding clusters; these patterns can be data-driven. Flooding patterns can define alarm occurrence rates, such as the occurrence rates of several types of alarms for various alarm flooding clusters. As a supplement or alternative to the flooding patterns associated with the corresponding clusters, flooding characteristics 64 can also include other data, such as data based on alarm messages for the corresponding alarm flooding cluster. Data based on alarm messages can include data obtained through NLP processing of the alarm messages.

[0214] At least one processing circuit 70 is operable such that alarm processing 72 includes determining an alarm flood type. Alarm flood type determination 74 may include assigning the identified alarm flood to one of the clusters defined by alarm cluster definition 64.

[0215] At least one processing circuit 70 is operable such that alarm processing 72 includes alarm classification 76 to identify alarms that do not belong to any alarm flooding. Flooding feature 64 can be used to determine which alarms in a generated series of alarms do not belong to any alarm flooding.

[0216] Interface control 79 can operate to control at least one interface 61 and / or HMI interface 62 based on the identified alarm flood, the determined alarm flood type, and alarms that do not belong to any of the identified alarm floods.

[0217] Figure 5 This is a block diagram representation of a processing system 60 according to an embodiment. In terms of configuration and / or functionality, it is consistent with previously referenced... Figure 1 or Figure 4 The components explained are denoted by the same reference numerals.

[0218] For at least one alarm cluster, storage system 63 stores pattern data 65 that defines the alarm pattern (as a function of time) for the corresponding alarm cluster. Optionally, storage system 63 may also store message-based data 66 related to topics identified through NLP processing of alarm messages for the corresponding alarm cluster.

[0219] At least one processing circuit 70 is operable such that alarm processing 72 includes determining alarm flooding 74 using at least pattern data 64. At least one processing circuit 70 is operable such that alarm processing 72 may include identifying alarms not included in any alarm flooding, i.e., performing pattern-based identification 77. For illustration, once an identified alarm flood has been assigned to one of the alarm flooding clusters in storage system 63 that stores its flooding pattern 65, at least one processing circuit 70 can operate to determine which alarms in a series of alarms generated by alarm generator 71 do not conform to the expected flooding pattern 65 for the identified cluster to which the identified alarm flood belongs.

[0220] Figure 1 , Figure 4 and Figure 5 The processing system 60 in any of the systems can operate to apply a similarity metric of alarm occurrence as a function of time to assign identified alarm flooding to one of the alarm flooding clusters. To do this, alarm patterns (specifically, alarm occurrence rates) can be compared with flooding patterns 65 to quantitatively assess similarity or dissimilarity and perform assignment to one of the alarm flooding clusters.

[0221] Figure 6 This is a schematic representation of the operation of the processing system 60 and processing method according to an embodiment. Monitoring data obtained from the power system 10 causes a series of alarms 100 to be generated. Figure 6 The different alarm types are indicated by offsets and different symbols. This series of alarms includes the first alarm flood 101 and the second alarm flood 102.

[0222] At least one processing circuit 70 is operable to assign alarm flooding to different alarm flooding clusters based on the alarm occurrence rate within a time interval 104. The time interval 104 may be a sliding time interval that adjusts as new alarms are generated. The assignment may be performed based at least on alarm patterns stored in the storage system 63.

[0223] This series of alarms may include one or more alarms 103 that do not belong to any of the alarm floods 101 and 102, even if they are issued within the same time interval as the alarms of the alarm floods 101 and 102 (e.g., simultaneously). Therefore, alarms that do not belong to any of the alarm floods 101 and 102 can be identified based on the alarm patterns stored in the storage system 63.

[0224] Alarm flooding can be identified based on the number of alarms and / or the occurrence rate of alarm triggering. This identification can be used to define alarm flooding clusters and their characteristics (such as alarm patterns) and / or during inference. During inference, the identified alarm flooding clusters are further analyzed by assigning them to one of the alarm flooding clusters.

[0225] Figure 7 This is a flowchart of method 110. When defining an alarm flooding cluster based on historical alarms, method 110 can be executed automatically by processing system 60 or a separate computing system. Method 110 can be executed automatically by processing system 60 during inference.

[0226] At process block 111, it is determined whether the number of alarms meets a first criterion. The first criterion may include a first threshold comparison (e.g., comparing the number of alarms with a first threshold, wherein the first criterion is met when the number of alarms is equal to or greater than the first threshold).

[0227] At process block 112, it is determined whether the occurrence rate of the alarm (which may be the average occurrence rate over time interval 104) meets a second criterion. The second criterion may include a second threshold comparison (e.g., comparing the occurrence rate to a second threshold, wherein the second criterion is met when the occurrence rate is equal to or greater than the second threshold).

[0228] At process block 113, if both the first and second criteria are met, an alarm flood is identified.

[0229] At process block 114, if at least one of the first and second criteria is not met, it is determined that there is no alarm flooding.

[0230] When using method 110 to determine alarm flooding clusters, there is no need to further consider data that does not include any alarm flooding. For data that includes one or more alarm flooding clusters, clustering can be performed. Unsupervised techniques, such as k-means clustering, can be used. Characteristics of the resulting clusters (such as the typical occurrence rate of alarms in various clusters) can be determined for each alarm flooding cluster, and these characteristics can be used during inference.

[0231] When method 110 is used during inference, the existence of an alarm flood (procedure box 113) can trigger a determination of which alarm flood cluster the alarm flood will be assigned to. The alarm pattern associated with the alarm flood cluster can be used to identify alarms(s) not included in the alarm flood.

[0232] Figure 8 This is a schematic representation 120 showing multiple alarm flooding clusters 121, 122. Alarm flooding assigned to different alarm flooding clusters 121, 122 can be distinguished from each other according to the occurrence rate of at least one alarm. Although Figure 8 The diagram only illustrates one incidence rate, but more complex cluster definitions (e.g., incidence rates for different alarm types) can be used.

[0233] Processing system 60 can be operated to use the alarm patterns of alarm flood cluster 121 to determine whether an alarm flood identified in a series of alarms is associated with alarm flood cluster 121. This determination can be made by comparing a similarity metric of the generated time-related alarm sequences with the alarm patterns of alarm flood cluster 121. This assignment is also referred to herein as determining the type of alarm flood (it should be understood that the type of alarm flood can also be determined in ways other than based on cluster analysis). If an alarm flood is assigned to alarm flood cluster 121, processing system 60 can be operated to use the alarm flood patterns of alarm flood cluster 121 to identify which alarm(s) in the series of alarms do not belong to that alarm flood.

[0234] Processing system 60 can be operated to determine whether an identified alarm flood in a series of alarms is associated with another alarm flood cluster 121 using another alarm pattern of another alarm flood cluster 122. This determination can be made by comparing a similarity measure of the generated time-related alarm sequences with another alarm pattern of the other alarm flood cluster 122. If an alarm flood is assigned to another alarm flood cluster 122, processing system 60 can be operated to use another alarm flood pattern of the other alarm flood cluster 122 to identify which alarm(s) in the series of alarms do not belong to that other alarm flood.

[0235] When a series of alarms includes alarm floods belonging to several alarm flood clusters, the processing system 60 can operate to determine which alarms generated by the alarm generator 71 do not belong to any alarm flood by using the alarm patterns of all these alarm flood clusters to which at least one alarm flood is assigned.

[0236] The processing system 60 and the processing method are operable such that the action taken depends on the identified alarm flood and at least one alarm that is not part of any alarm flood in a series of generated alarms. Figure 9 The diagram illustrates this operation.

[0237] Figure 9 This is a flowchart of method 130. Method 130 can be executed automatically through alarm processing 72 of processing system 60 and interface control 79.

[0238] At process frame 131, alarm processing 72 receives alarms. Alarms may include those generated by alarm generator 71, which are received and processed (online processing) upon generation. Alarms may also include those retrieved from a repository (e.g., for root cause analysis or other purposes).

[0239] At procedure block 132, determine whether the alarm includes at least one alarm flood 132. Procedure block 132 can be implemented using procedure blocks 111, 112 of method 110.

[0240] At process block 133, if there is no alarm flooding, interface control 79 enables output to be provided for each of the individual alarms.

[0241] At process block 134, if alarm flooding exists, interface control 79 enables the output to treat alarms not included in any alarm flooding and alarms included in alarm flooding differently.

[0242] The output may accordingly include control commands. The processing system 70 may operate such that alarms not included in any alarm flood are treated differently to ensure that appropriate action is taken in response to these alarms even when they are issued within the time interval of several alarm floods. Control commands may include control commands for the HMI 62 and / or control commands for auxiliary system devices of the power system (e.g., commands to trigger a Supervisory Control and Data Acquisition (SCADA) control subsystem to perform at least one action based on the identified alarm flood(s) and at least one alarm not included in the alarm flood).

[0243] Figure 10 This is a flowchart of method 130. Method 130 can be executed automatically through alarm processing 72 of processing system 60 and interface control 79. (Already referenced...) Figure 9 The process flowcharts are represented using the same reference numerals.

[0244] Process frames 131 and 132 can be used as follows: Figure 9 Implement it as discussed in connection with the above.

[0245] At process block 135, when there is no alarm flooding among the received series of alarms, processing system 60 can perform a first operation. The first operation can be performed individually for each alarm, for example, by providing a control command that causes each alarm to be output individually. Alternatively or additionally, the first operation may include a corrective action in response to the alarm.

[0246] At process block 141, if alarm flooding exists, alarm processing 72 can determine the type of alarm flooding. This may include evaluating a similarity metric for a range of alarms using the alarm patterns of each of several alarm flooding clusters. Other techniques may be used to assign alarm flooding to alarm flooding groups that have been identified in historical data as having common characteristics, such as the frequency of alarm issuance.

[0247] At procedure box 142, select alarms from the received set. These alarms can be selected sequentially based on the time they were issued.

[0248] At procedure block 143, it is determined whether the selected alarm is part of any alarm flood that was identified as existing at procedure block 141. The determination at procedure block 143 can be performed by checking whether the alarm was issued at a time consistent with the alarm flood pattern of any alarm flood cluster to which the alarm flood from this series is assigned.

[0249] At process block 144, if the alarm is not part of any alarm flood, the processing system 60 can perform a first action for the alarm. This first action can be performed individually for the alarm, for example, by providing a control command that causes the alarm to be output separately via HMI 62 and / or triggers corrective actions in response to the alarm.

[0250] At process block 145, if the alarm is part of an alarm flood, the processing system 60 is operable to perform a second action associated with the alarm flood, rather than with individual alarms. The second action differs from the first action. The second action may include providing control commands that cause indicators of the alarm flood (but not all of its constituent alarms) to be output via HMI 62 and / or trigger corrective actions in response to the alarm flood.

[0251] At procedure box 146, determine if any remaining alarms exist. If any remaining alarms exist, the method can return to procedure box 142. Otherwise, at procedure box 147, monitoring for new alarms can continue, and the entire method can be repeated.

[0252] It should be understood that during the repeated iterations of the loop from process blocks 142 to 146, it is not necessary to repeatedly execute the second operation at process block 145. Instead, the processing system 60 can be operated such that the second operation 145 is executed only once collectively for all alarms in an alarm flood.

[0253] Therefore, the processing system 60 and the processing method can operate as follows: - Receive online and / or offline alarm data.

[0254] - Classify incoming alerts as either flooded or not flooded.

[0255] - If there is no flooding, perform a first action for each alarm individually (e.g., by performing HMI control to output a separate alarm and / or performing another control action to correct the situation that triggered the alarm).

[0256] - If an alarm flood exists, the alarm flood will be assigned to one of the alarm flood clusters whose characteristics were previously identified during the learning phase.

[0257] - If alarm flooding exists, classify each alarm as either part of the alarm flooding or not.

[0258] - For alarms that do not belong to any alarm flooding, perform the first action (e.g., by executing HMI control to output a separate alarm or by executing another control action to correct the situation that triggered the alarm).

[0259] - For alarms that fall under the category of alarm flooding, perform a different second action (e.g., by performing HMI control to output a representation of the alarms that fall under alarm flooding with reduced cognitive load (optionally output in association with a topic that helps identify the root cause), and / or by performing another control action to correct the situation that triggered the alarm flooding).

[0260] Processing system 60 can operate to perform HMI control based on identified alarm floods (multiple) and alarms not belonging to any alarm flood. Processing system 60 can operate to perform HMI control to enable power system operators to perform synchronous review (…). Figure 11 (or asynchronous review by the power system operator) Figure 13 and Figure 14 In either case, the processing system 60 can be operated to enable operator input. Figure 12 It also modifies HMI controls in response to operator input to provide additional information, such as identified alarm floods and / or alarms that do not belong to any alarm flood.

[0261] Figure 11This is a schematic diagram of an HMI controlled by at least one processing circuit 70 to provide an alarm output display 150. The processing system 60 can operate to control the HMI 62 such that the alarm output display 150 displays alarms 151, 152, 153, 154, 155 that do not belong to any alarm flooding in a first manner (e.g., such that outputs 151, 152, 153, 154, 155 include additional information about affected main system devices and / or auxiliary system devices). The processing system 60 can operate to control the HMI 62 such that the alarm output display 150 displays an area 156 associated with the alarm flooding. The processing system 60 can operate such that all alarms belonging to the alarm flooding are not displayed separately as individual alarms. The processing system 60 can operate such that the area 156 of the alarm flooding includes information about at least some of the alarms belonging to the alarm flooding, in a manner that is compressed in display space size compared to the area associated with each individual alarm 151, 152, 153, 154, 155. Therefore, alarm flooding, represented by zone 156, reduces cognitive load and mitigates the risk of individual alarms 151, 152, 153, 154, and 155 being ignored.

[0262] The processing system 60 can be operated to control the HMI 62 such that the alarm output display 150 allows the power system operator to activate the display element 159, which, in response to its activation, causes the processing system 60 to expand information related to alarm flooding (e.g., by modifying the HMI controls to make additional alarms compressed into zone 156 appear separately).

[0263] Alarm flooding is often noticeable to synchronous users (such as grid control room operators) because they frequently monitor real-time alarm feeds. However, flooding presents usability challenges, namely, identifying which alarms are triggered due to alarm flooding and which are triggered independently of it. This is because a flood of alarms caused by a single incident can mask independent alarms that may be triggered by unrelated incidents. Therefore, using features such as "acknowledge all" or batch selecting multiple alarms without inadvertently including unrelated alarms in the selection increases the likelihood of false alarm confirmations.

[0264] Processing system 60 integrates alarm flooding clustering analysis into real-time alarm feeds to help synchronous users identify which alarms are caused by alarm flooding and which are not (shown in the figure below). The HMI control also collapses many alarms into a single group in zone 156, saving valuable screen space. When users wish to explore alarms within the alarm flooding in detail, they can expand this group in zone 156.

[0265] exist Figure 11In this section, each of lines 151, 152, 153, 154, and 155 represents an alert that does not belong to the alert flooding category. Line 156 represents the alert flooding category. The constituent alerts are grouped in line 156 and truncated to save screen space. Labels (such as color codes or symbol '159') can be associated with area 156 to clearly identify the alert flooding category. The processing system 60 can be operated to output further topics in association with the alert flooding category to provide further insight into the possible root causes of the alert flooding. Topics are determined by performing NLP on the alert messages.

[0266] The processing system can also be operated to output information in association with information about alarm overload regarding the amount of alarm data that is intentionally hidden or invisible to reduce cognitive load. For illustration, the processing system 60 can control the HMI 62 to provide an output 159'' (which indicates the number of alarms that are part of the alarm overload but are not shown to reduce cognitive load), and optionally also control the HMI 62 to provide an input device 159 that allows a user to instruct the output of suppressed alarms from the alarm overload to be displayed on the HMI 62.

[0267] Figure 12 This is a flowchart of method 160. Method 160 can be executed automatically by processing system 60. Method 160 can be executed to utilize the results of alarm processing in, for example, a power grid control room. Method 160 can be executed automatically by or using at least one processing circuit 70.

[0268] At process block 161, at least one processing circuit 70 performs a control action based on the identified alarm flood(s) and at least one alarm that does not belong to any alarm flood. The control action may include an HMI control action and / or another control action (such as a trigger action to be performed by the SCADA control subsystem).

[0269] At process block 162, at least one processing circuit 70 controls HMI 62 to enable selection of alarm flooding. This may include controlling HMI 62 to output element 159 selectable by the grid operator.

[0270] At process block 163, in response to the selection, at least one processing circuit 70 modifies the control of HMI 62. This may include controlling HMI 62 to provide additional information about alarms included in the selected alarm flood, for example, by expanding previously compressed alarms in the alarm flood.

[0271] Processing system 60 may alternatively or additionally operate to control HMI 62 in a manner particularly suitable for use by asynchronous users. Processing system 60 may operate to provide a more comprehensive understanding of when alarm floods occur and the types of alarms associated with them. Processing system 60 may operate to control HMI 62 to provide an alarm explorer that provides a high-level view of alarms occurring over time associated with each alarm flood. Processing system 60 may operate to control HMI 62 to selectively or in combination output information about alarm occurrence rates. Figure 13 ) and each of the alarms associated with the alarm flood ( Figure 14 The HMI 62 can be operated to switch between various displays and / or enable selection of various alarm flooding.

[0272] Figure 13 This is a schematic diagram of an HMI controlled by at least one processing circuit 70 to provide an alarm output display including an overview representation 170. The overview representation 170 includes histogram bars representing the total alarm rate over various time intervals. Each histogram bar 173 can be indicated by different shaded portions 174, 175, 176 to indicate the portion occupied by alarms of different priorities (e.g., shaded portion 174 indicates low-priority alarms, another shaded portion 175 indicates medium-priority alarms, and yet another shaded portion 176 indicates high-priority alarms). The processing system 60 can also control the HMI 62 such that the overview representation 170 includes information on time-related changes 171 in the alarm occurrence rate. The change curve 171 can be generated by the processing system to provide further insight into the relationships between histogram bins.

[0273] A histogram can represent the frequency of alarms over time. Each interval of the histogram corresponds to a time interval (e.g., one hour). Different graphical representations (e.g., colors, shading lines, or other markers) can be used to show the proportion of high / medium / low priority alarms within that interval. Curves171, using methods such as kernel density estimation, can be overlaid on the histogram to provide further insight into the relationships between intervals. For example, a flood might begin at 12:58, but this might not be visually reflected in the frequency interval between 12:00 and 12:59, whereas curve171 captures this.

[0274] The processing system 60 can be operated to control the HMI 62 to enable the user to select an interval or interval range, and in response to the selection, provide additional information about the selected interval of the histogram. This may include outputting additional information on the screen, such as an alarm about, for example, the selected histogram interval 173. Figure 14 As illustrated in the example.

[0275] Figure 14 A graphical representation of an alert that can be associated with, for example, a selected interval of a histogram is shown. Figure 14 In the diagram, each alarm is represented by a square. The processing system 60 can be operated to control the HMI 62 to optionally output a two-dimensional array 180 in response to user selection, the two-dimensional array being based on the time 181 when the alarm was issued. Figure 14 (X-axis in the middle) and for different alarm types ( Figure 14 The Y-axis in the diagram shows the alarms. Alarms 183, which are not part of the alarm flood or the selected alarm flood, and alarms 182, which are part of the alarm flood or the selected alarm flood, can be output in different ways.

[0276] Although not in Figure 13 and Figure 14 As shown, however, the processing system 60 can also be operated to control the HMI 62 to output to... Figure 14 The alert list is visually filtered using several tags. Each tag can correspond to an alert cluster. Tag selection enhances... Figure 14 The weight of associated alerts in the alert list is adjusted (e.g., by selecting alert 182 in the alert flood), and the weight of unrelated alerts (e.g., alert 181) is reduced.

[0277] Providing alarm-related data to asynchronous users allows them to conduct system reviews of ongoing alarms. This is of great significance to users such as power system operations administrators or engineers.

[0278] As a supplement to or alternative to HMI control operations, processing system 60 may operate (e.g., via SCADA control subsystem) to trigger actions affecting main power system components and / or auxiliary power system components and their operations. Such actions may include corrective actions that can be triggered based on identified alarm flooding and alarms determined not to be part of any alarm flooding.

[0279] Figure 15 This is a flowchart of method 190. Method 190 can be executed by processing system 60 alone, or optionally in association with a separate computing system (process block 191).

[0280] At process block 191, processing system 60 or a separate computing system performs a data-driven determination of the characteristics of alarm flood clusters. This may include: clustering the identified alarm floods (using, for example, k-means clustering or another unsupervised clustering technique), extracting the alarm patterns of the corresponding alarm flood clusters from the alarm floods in each cluster, and providing a definition and characteristics of the alarm flood clusters thus established for use during inference. Therefore, the characteristics of alarm flood clusters can be determined in a data-driven manner.

[0281] At process block 192, processing system 60 utilizes the feature of alarm flooding clusters during inference. Process block 192 can be implemented using a processing method and / or processing system 60 according to any embodiment of the embodiments disclosed herein.

[0282] Figure 16 This is a flowchart of method 200. Method 200 can be executed by processing system 60 or a computing system different from processing system 60. Method 200 can be executed to implement process block 191. Figure 15 ).

[0283] At process frame 201, retrieve historical alarm data. Historical alarm data may include alarms issued during the operation of at least one transmission network and / or distribution network and / or power plant. Historical alarm data may include time information for these alarms.

[0284] At process box 202, identify alarm flooding in the retrieved historical alarm data. For example, you can use... Figure 7 Method 110 is used to identify alarm overload.

[0285] At process block 203, clustering of the identified alarm proliferation can be performed. Clustering can include k-means clustering or another unsupervised clustering technique. Similarity measures (such as frequency- or occurrence-based measures) can be used. Clustering can also be performed without using sequence-based similarity measures (given the complex dynamics of power systems).

[0286] At process block 204, the characteristics of each cluster are determined. Determining the characteristics may include identifying alarm patterns and / or applying NLP to alarm messages to extract common themes of alarm flooding across the corresponding alarm flooding cluster.

[0287] At process block 207, the characteristics of the determined cluster are stored in storage system 63 or otherwise made available for use by processing system 60.

[0288] Figure 17 This is the flowchart for method 208. This method is... Figure 16 Specific implementation of method 200. Method 208 can be executed by processing system 60 or a computing system different from processing system 60. Method 208 can be executed to implement process block 191 ( Figure 15 ).

[0289] Process blocks 201, 202, and 203 can be used as follows: Figure 17The implementation is as described in connection with the preceding description. In process block 202, alarm flooding can be identified based on the definition of alarm flooding (using the minimum total number of alarms and the minimum alarm frequency). In process block 203, alarm flooding can be clustered based on alarm identifiers. To cluster alarm flooding, unsupervised learning techniques (e.g., k-means clustering) can be used. In power system applications, system dynamics are much faster than in process industries, so sequence-based similarity measures are less suitable than frequency- or occurrence-based measures. Therefore, process block 203 can use frequency- or occurrence-based measures to perform clustering.

[0290] At process block 205, the overall alarm flooding pattern can be extracted. This overall alarm flooding pattern can correspond to similar alarms in cluster-wide alarm flooding.

[0291] At process block 206, NLP techniques can be used to process the alert text message to identify the corresponding topic.

[0292] Procedure boxes 205 and 206 can be executed simultaneously, or either procedure box 205 or 206 can be executed in response to user input or selection.

[0293] Alarm processing performed by the processing system 60 and / or in the processing method can be used to control the HMI 62, but can alternatively or additionally be used to trigger actions that affect the operation of the main system and / or auxiliary systems. For this purpose, the processing system 60 can be communicatively interfaced with or integrated with the SCADA control subsystem, or can otherwise communicatively interface with the SCADA system.

[0294] Figure 18 A schematic representation of a power system 210 according to an embodiment is shown. The power system 210 includes a main system device 213 and a SCADA system 211. The SCADA system 211 includes a processing system 60 and a SCADA control subsystem 212 communicatively interfaced with the processing system 60. The SCADA control subsystem 212 is operable to perform control actions in response to outputs provided by the processing system 60. The control actions may depend on both the identified alarm(s) flood and the alarm(s) not included in any of the alarm(s) floods.

[0295] Figure 19A schematic representation of a power system 220 according to an embodiment is shown. The power system 220 includes a main system device 223 and a SCADA system 221. The SCADA system 221 includes a SCADA control subsystem 222. The SCADA system 221 is communicatively interfaced with the processing system 60. The SCADA control subsystem 222 is operable to perform control actions in response to outputs provided by the processing system 60. Control actions may depend on both the identified alarm(s) flood and alarm(s) not included in any of the alarm(s) floods.

[0296] Figure 20 This is a flowchart of method 230. Method 230 can be executed automatically by processing system 60, or optionally in association with SCADA control subsystem or other auxiliary system devices.

[0297] At process frame 231, the processing system 60 identifies at least one alarm flood.

[0298] At process block 232, processing system 60 identifies at least one alarm that is not included in any alarm flood in the multiple alarm floods.

[0299] Process blocks 231 and 232 can be implemented using any of the techniques disclosed in detail herein.

[0300] At process block 233, processing system 60 triggers a control action to affect the main power system device and / or auxiliary system device, depending on both the identified alarm flood and the alarms in any alarm flood not included in the alarm flood.

[0301] The processing system, method, and power system according to the embodiments achieve various effects and advantages. The processing system and method provide enhanced techniques for processing alarms, which are particularly useful when correlated with the complex dynamics of power systems.

[0302] This processing system and method mitigates the risk of alarms not being properly considered when (multiple) alarms not belonging to an alarm flood occur within the same time interval (e.g., simultaneously) as alarms belonging to the alarm flood. The system and method can be operated to compress a large number of alarms during an alarm flood to facilitate efficient alarm processing. The system and method can be operated to identify and output the root cause of the alarm flood. The system and method are operable to distinguish alarms not belonging to the alarm flood and generate outputs based on these alarms. Therefore, the system and method can provide intelligent alarm management for the power system domain. The system and method can be used in conjunction with, but are not limited to, the power grid or its subsystems.

[0303] Therefore, the processing system and method also address the following need: manually processing all alarms within an alarm flood is a tedious and time-consuming task, while some important alarms may be completely ignored. This lack of situational awareness can jeopardize the normal operation of various power grid components and even the stability of the power system. This processing system and method provides support during alarm flooding. This can include compressing a large number of alarms while providing information about the flood (theme). Additionally, alarms not part of the flood are identified and presented separately to the operator.

[0304] The processing system and method can be operated to provide advantages over alarm template matching, which is applicable to a limited set of process alarms in a process system.

[0305] Although embodiments have been described in detail with reference to the accompanying drawings, various modifications may be implemented in other embodiments. This is for illustrative purposes only and not for limitation. • Although embodiments have been described in connection with power grids, the techniques disclosed herein can also be used in connection with high voltage direct current (HVDC) control systems or microgrid control systems.

[0306] • While an embodiment in which the processing system 60 is included in or interfaces with a SCADA system in communication has been described, the processing system 60 may also be used in other use cases, such as for performing offline analysis of alarms.

[0307] The embodiments can be used in conjunction with, but are not limited to, power grids with renewable energy penetration (such as power grids that include renewable energy systems such as DER).

[0308] This specification and accompanying drawings, which illustrate various aspects and embodiments of the invention, should not be considered as limiting the scope of the claims. In other words, while the invention has been illustrated and described in detail in the accompanying drawings and the foregoing specification, such illustrations and descriptions should be considered illustrative rather than restrictive. Various mechanical, compositional, structural, electrical, and operational changes can be made without departing from the spirit and scope of this specification and the claims. In some instances, well-known circuits, structures, and techniques have not been shown in detail to avoid obscuring the invention. Therefore, it will be understood that changes and modifications can be made by those skilled in the art within the scope and spirit of the following claims. In particular, the invention covers other embodiments having any combination of features from the different embodiments described above and below.

[0309] This disclosure also covers all other features shown individually in the drawings, although these features may not be described in the preceding or following description. Furthermore, single alternatives to the embodiments described in the drawings and specification, and single alternatives to their features, may be excluded from the subject matter of the invention or the disclosed subject matter. This disclosure includes subject matter constituted by the features defined in the claims or embodiments, as well as subject matter including said features.

[0310] The term "comprising" does not exclude other elements or process blocks, and the indefinite articles "an" or "a" do not exclude a plural. A single unit or process block can perform the function of multiple features recited in the claims. The simple fact that certain measures are recited in mutually different dependent claims does not imply that a combination of these measures cannot be used advantageously. Components described as coupled or connected can be directly electrically or mechanically coupled, or they can be indirectly coupled via one or more intermediate components. No reference numerals in the claims should be construed as limiting the scope.

[0311] Machine-readable instruction code can be stored / distributed on a suitable medium (such as an optical storage medium or solid-state medium provided with or as part of other hardware), but can also be distributed in other forms (such as via a wide area network or other wired or wireless communication system). Furthermore, machine-readable instruction code can also be a data structure product or signal used to embody a specific method (such as the method according to an embodiment).

Claims

1. A processing system for controlling and / or monitoring a power system (10), the processing system (60) comprising: At least one interface (61) is operable to receive monitoring data (68) during operation of the power system (10). as well as At least one processing circuit (70), said at least one processing circuit being operable to: Identify alarm flooding (101, 102; 182) among alarms generated based on the received monitoring data (68). Identify at least one alarm (103) that does not belong to the alarm flood (101, 102; 182), and Outputs (69; 150; 170; 180) are generated based on the identified alarm flood (101, 102; 182) and at least one alarm (103) that does not belong to the alarm flood (101, 102; 182). The at least one processing circuit (70) is operable to use flood cluster definition to identify the alarm flood (101, 102; 182) and the at least one alarm (103) that does not belong to the alarm flood (101, 102; 182), the flood cluster definition defining alarm flood clusters based on alarm occurrence rate and independent of the temporal relationship between alarms with different alarm types.

2. The processing system as described in claim 1, wherein, The outputs (69; 150; 170; 180) include outputs (150; 170; 180) provided via a human-machine interface (HMI) (62), and the at least one processing circuit (70) is operable to control the HMI (62) such that the at least one alarm (103) that does not belong to the alarm flood (101, 102; 182) and the alarm that belongs to the identified alarm flood (101, 102; 182) are output in a visually different manner.

3. The processing system as described in claim 2, wherein, The at least one processing circuit (70) is operable to control the HMI (62) such that the at least one alarm (103) that does not belong to the alarm flood (101, 102; 182) and the alarm that belongs to the identified alarm flood (101, 102; 182) are output at different scales.

4. The processing system as described in claim 2 or claim 3, wherein, The at least one processing circuit (70) is operable to control the HMI (62) such that the alarms belonging to the identified alarm flood (101, 102; 182) are output in a low cognitive load manner compared to the at least one alarm (103) not belonging to the identified alarm flood (101, 102; 182).

5. The processing system as described in any one of the preceding claims, wherein, The at least one processing circuit (70) is operable to perform alarm classification in response to the detection of the alarm flood (101, 102; 182), and to identify the at least one alarm (103) that does not belong to the alarm flood (101, 102; 182) based on the alarm classification.

6. The processing system as described in claim 5, wherein, The at least one processing circuit (70) is operable to enable the alarm classification to include cluster analysis.

7. The processing system as described in claim 5 or claim 6, wherein, The at least one processing circuit (70) is operable to classify alarms based on alarm occurrence rate.

8. The processing system as described in any one of the preceding claims, wherein, The at least one processing circuit (70) is further operable to identify at least one additional alarm flood (102) among the generated alarms and generate additional outputs (170; 180) including information aggregated by the at least one processing circuit (70) from the identified alarm flood (101) and the at least one additional alarm flood (102).

9. The processing system as described in any one of the preceding claims, wherein, The at least one processing circuit (70) is further operable to perform root cause determination on the identified alarm flood, and further generate the output (69; 150; 170; 180) based on the result of the root cause determination.

10. The processing system of claim 9, wherein, The at least one processing circuit (70) is operable to perform the root cause determination based on natural language processing (NLP) of the alarm message associated with the generated alarm.

11. The processing system as described in any of the preceding claims, wherein, The at least one processing circuit (70) is operable such that the output (69) includes at least one control command that causes at least one action to be performed on the main system equipment (21 to 26) of the power system (10) and / or the auxiliary system equipment (41 to 46) of the power system (10).

12. The processing system as claimed in any of the preceding claims, wherein, The at least one processing circuit (70) is further operable to generate the alarm based on the received monitoring data (68).

13. An electric power system (10), comprising: Main system equipment (21 to 26). Auxiliary system devices (41 to 46) of the auxiliary system (40) associated with the main system devices (21 to 26), and The processing system (60) as described in any of the preceding claims is operable to receive the monitoring data (68) from the auxiliary system devices (41 to 46).

14. A method for controlling and / or monitoring a power system (10), the method comprising: The processing system (60) receives monitoring data (68) during the operation of the power system (10). The alarm processing is performed by the processing system (60), the alarm processing including: Identify alarm flooding (101, 102; 182) among alarms generated based on the received monitoring data (68). Identify at least one alarm (103) that does not belong to the alarm flood (101, 102; 182), and Outputs (69; 150; 170; 180) are generated based on the identified alarm flood (101, 102; 182) and at least one alarm (103) that does not belong to the alarm flood (101, 102; 182). The identification of the alarm flood (101, 102; 182) and the at least one alarm (103) that does not belong to the alarm flood (101, 102; 182) is performed using an alarm flood cluster definition, which defines alarm flood clusters based on alarm occurrence rates and independently of the temporal relationship between alarms with different alarm types.

15. The method of claim 14, wherein, The method is performed by the processing system (60) as described in any one of claims 1 to 12 or by the power system (10) as described in claim 13.

Images