Patent Application
20200160193
The following relates to a method and an apparatus for the computer-aided creation of digital rules for monitoring a technical system. Further, the following relates to a method and an apparatus for monitoring a technical system and to a computer program product and a computer program.
The prior art discloses rule-based monitoring methods for technical systems in order to diagnose the operation of the technical system and to identify or predict malfunctions in good time. Such rule-based methods involve digital rules being executed by a rules engine, wherein the rules take specific criteria as a basis for identifying irregularities in the operation of the technical system and for example outputting appropriate warnings.
Rule-based methods for monitoring technical systems are specifically matched to the components of the system to be monitored, i.e. device identifications specific to the technical system are used to denote the components in the digital rules. Accordingly, rule-based methods are always individually designed for the applicable technical system and cannot readily be transferred to another technical system, even if the latter has the same or a similar design.
An aspect relates to provide a method for the computer-aided creation of digital rules for monitoring a technical system that allows rules created by a user to be easily used for multiple technical systems.
The method according to embodiments of the invention is used to create (actual) digital rules in computer-aided fashion for monitoring a technical system. The term monitoring should be understood broadly in this instance and relates to the analysis of a technical system on the basis of operating data. According to embodiments of the invention, an ontology is provided that contains multiple classes comprising classes of components of the technical system and classes of operating state characteristics of the technical system and also semantic relations between the classes. The components are applicable parts of the technical system in this instance. A component can further also relate to the technical system as a whole. The classes of components comprise classes of parts of the technical system without a sensor function and also classes of sensors. Classes of parts without a sensor function are in this instance functional units relating to an operating function that is not directly linked to sensed measured values. Nonetheless, such classes of parts can also contain sensors. The classes of operating state characteristics of the technical system are dependent on state variables of the technical system. Examples of such classes are cited later on.
The semantic relations used in the ontology can be relations known per se, such as e.g. a relation specifying that one class is a subclass of another class, or a relation specifying that one class is part of another class. The definition of further relations on the basis of the classes present in the ontology is within the scope of action of a person skilled in the art in this instance. If the ontology contains e.g. the classes of operating processes that are described later on, a relation can also state that a class of a component is involved in a class of an operating process.
The method according to embodiments of the invention involves a user interface that comprises a visual display being provided, wherein a user can use the user interface to create abstract rule formulations on the basis of the classes and the semantic relations of the ontology by using the visual display. The rule formulations created by the user are stored as digital abstract rules in this instance. In one variant, the digital abstract rules are stored in XML format.
Finally, the digital abstract rules are instantiated for the technical system in automated fashion as digital actual rules by virtue of the classes of components in the digital abstract rules being mapped onto specific device identifications of the technical system by accessing a database containing data about the technical system, and the classes of operating state characteristics being converted into operating state characteristics specific to the technical system that comprise the specific device identifications. A generic rule language is therefore mapped in automated fashion onto a rule language specific to the system to be monitored. This results in the advantage that abstract rules for identical or similar technical systems do not need to be recreated for the specific technical system by the user each time, but rather can be used repeatedly for identical or similar technical systems by means of suitable instantiation.
In one exemplary embodiment, the mapping of the classes of components in the digital abstract rules onto the specific device identifications of the technical system is based on the R2RML language (RDB to RDF Mapping Language). This is a language standard of the W3C consortium in order to map information from relational databases into RDF data records of an ontology.
In a further refinement of the method according to embodiments of the invention, the classes of components contain not only individual components but also classes of functional groups comprising multiple subcomponents. A class of a functional group can relate e.g. to an oil system or a gas path in the technical system.
In one variant, the classes of operating state characteristics contain the following classes:
These classes are therefore used to characterize state variables and values dependent thereon. The classes for describing a change over time can indicate different types of changes in this instance, e.g. a severe change, a moderate change or a slow change. Such a class can also be used to characterize longer-term changes in the form of trends.
In a further exemplary embodiment, the classes of the ontology further comprise classes of abstract rules that the user can take into consideration for creating the abstract rule formulations and that are instantiated as applicable actual rules. In this manner, rule blocks for creating the abstract rules are already prescribed for the user, which means that rule creation is simplified.
In a further refinement, the classes of the ontology further comprise classes of operating processes that a user can take into consideration for creating an abstract rule formulation, wherein the instantiation of the digital actual rules involves classes of operating processes in the digital abstract rules being mapped onto specific operating process identifications of the technical system by accessing the above database. These operating process identifications are then contained in the operating state characteristics specific to the technical system. Such operating processes can relate to a starting process or a switch-off process of the technical system, for example. These classes of operating processes extend the options for creating applicable rules.
In a further exemplary embodiment, the method according to embodiments of the invention further involves a query interface being provided that a user can use to query operating data of the technical system on the basis of the classes and the semantic relations of the ontology. This is possible if the above database contains such operating data. If need be, this variant can also be used online during operation of the technical system to access up-to-date operating data.
The method according to embodiments of the invention can be used to create rules for any technical systems. In particular, the technical system to be monitored can be an electrical power generation installation or an electrical power grid or a turbine, such as e.g. a gas turbine, or an automation installation or a means of transport, such as e.g. a train, or a medical device, such as e.g. a computed tomography scanner or a magnetic resonance imaging scanner. An automation installation should be understood in this instance to mean an industrial installation for the automated performance of processes and in particular for the automated production or manufacture of products.
Embodiments of the invention furthermore relates to a method for monitoring a technical system, wherein digital actual rules are produced using the method according to embodiments of the invention that have just been described or one or more variants of said method and are subsequently executed using a rules engine on the technical system for the purpose of monitoring the latter.
Embodiments of the invention furthermore relates to an apparatus for the computer-aided creation of digital rules for monitoring a technical system, wherein the apparatus is designed such that during operation it performs the method according to embodiments of the invention for creating digital rules or one or more variants of said method.
Furthermore, embodiments of the invention relates to an apparatus for monitoring a technical system that is configured to perform the method described above for monitoring a technical system.
Embodiments of the invention furthermore relates to a computer program product (non-transitory computer readable storage medium having instructions, which when executed by a processor, perform actions) having a program code stored on a machine-readable carrier for performing the method according to embodiments of the invention for creating digital rules or for performing the method according to embodiments of the invention for monitoring a technical system when the program code is executed on a computer.
Further, embodiments of the invention comprises a computer program having a program code for performing the method according to embodiments of the invention for creating digital rules or for performing the method according to embodiments of the invention for monitoring a technical system when the program code is executed on a computer.
Some of the embodiments will be described in detail, with reference to the following figures, wherein like designations denote like members, wherein:
The method described with reference to
In accordance with
The ontology ON contains a multiplicity of classes in a manner known per se. In the example of
In a manner known per se, there are semantic relations SR between the classes, only one relation being indicated in exemplary fashion in
Examples of classes KO of components in the above ontology ON are parts of a gas turbine, such as e.g. specific rotor blades or specific sensors capturing measurement data, such as e.g. sensors for measuring combustion chamber temperatures or sensors for measuring rotor speeds.
Examples of classes BZ of operating state characteristics are state variables measured in the gas turbine, such as e.g. temperatures or speeds, or minimum values, maximum values and average values of such state variables. A further example of a class BZ of an operating state characteristic is what is known as a temporal trend, which specifies a longer-term trend in a state variable over time. Further, the operating state characteristics can specify increases and decreases in state variables, the increase and the decrease possibly also being able to be classified in regard to whether there is a severe, moderate or slight increase or fall.
Examples of classes BP of operating processes are starting up or shutting down the gas turbine under consideration. A class RK of an abstract rule can prescribe in the manner of a template that there is a specific fault in the gas turbine if measured variables come from a specific range of values.
A user can now use the user interface UI, by accessing the cited classes and semantic relations of the ontology ON, to create different abstract rules AR, which are stored in XML format in the exemplary embodiment under consideration in the present case. An example of an abstract rule of this kind can be that if the average speed of a rotor, as measured by means of a speed sensor, exceeds a predetermined limit over a predetermined period of time and at the same time there is a severe increase in a combustion chamber temperature in the gas turbine over this period of time, a warning is output during the operation of the technical system.
The abstract rules AR created by the user are formulated generically for components of the technical system and then need to be related to actual subassemblies of the system under consideration, i.e. a specific gas turbine in the present case. To achieve this, the abstract rules AR are instantiated in an instantiation step IN, i.e. they are converted into actual rules KR. This involves the classes KO of components and classes BP of operating processes that the abstract rules contain being converted into actual identifications. In particular, the classes KO are assigned to device identifications GID and the classes BP are assigned to process identifications PID, these identifications being specific to the actual gas turbine to be monitored. Mapping MP of generic classes onto actual identifications is therefore performed. Furthermore, the classes BZ of operating state characteristics are converted into actual operating state characteristics in the course of a translation step TR, these characteristics now containing applicable device identifications GID and process identifications PID.
To achieve the mapping MP just described, information from a database DB that specifies actual data of the technical system to be monitored and in particular contains the device identifications and process identifications is accessed. In the variant under consideration in the present case, the mapping is performed on the basis of the language standard R2RML known per se. The result obtained after the instantiation IN is a set of actual rules KR. These rules can then be stored in the technical system to be monitored and can be executed therein using a rules engine, so that the fulfilment of applicable conditions in the rules results in warnings being output or other countermeasures, such as e.g. shutdown of the gas turbine, being initiated.
The architecture of
The embodiment of the invention that is described above has a series of advantages. In particular, the semantic model of an ontology is taken as a basis for providing the opportunity to formulate, in a simple manner, abstract rules for monitoring a technical system, these rules automatically being translated into actual rules for the technical system under consideration. The complexity of creating rules is substantially reduced thereby. This has been able to be demonstrated by experiments. During these experiments, users have created applicable rules directly in the specific rule language of the technical system and by using the ontology according to embodiments of the invention. Creation of the abstract rules on the basis of the ontology has been able to be performed much more quickly by the user than on the basis of the actual rule language.
Furthermore, an interface for querying information from the technical system is provided, the generic terms from the ontology being able to be used for the query. This allows a simple search for heterogeneous data sources by using semantic terms from the ontology.
Although the present invention has been disclosed in the form of preferred embodiments and variations thereon, it will be understood that numerous additional modifications and variations could be made thereto without departing from the scope of the invention.
For the sake of clarity, it is to be understood that the use of “a” or “an” throughout this application does not exclude a plurality, and “comprising” does not exclude other steps or elements.
| Number | Date | Country | Kind |
|---|---|---|---|
| 17171786.1 | May 2017 | EP | regional |
This application claims priority to PCT Application No. PCT/EP2018/061814, having a filing date of May 8, 2018, which is based off of EP Application No. 17171786.1, having a filing date of May 18, 2017, the entire contents both of which are hereby incorporated by reference.
| Filing Document | Filing Date | Country | Kind |
|---|---|---|---|
| PCT/EP2018/061814 | 5/8/2018 | WO | 00 |
