WO2024227767A1 - Automatic configuration for system controllers - Google Patents

Abstract

The invention relates to a computer system for processing control data of a system, wherein the system has technical components which represent a data point with respect to the control data, comprising a computer and an interface for a system controller. The interface is connected to the computer and to the system controller, which exchanges control data with the technical components of the system via a network, wherein the interface detects the control data and transmit same to the computer, and the computer system is designed to classify a data point using natural language processing, assign said data point to a technical component, and generate a digital controller image of the system controller. In this manner, the disadvantages of the prior art are overcome.

Description

Automated configuration for plant controls

The invention relates to a computer system and an associated method for automated creation of a configuration for a system control, in particular for monitoring or consumption management of buildings, whereby data from an existing system control or automation is recorded, classified and interpreted and assigned using language models and functions such as semantic matching (SM). In particular, the invention relates to an innovative method for mapping semantically heterogeneous described control data or system information, in particular in the form of data points, of an existing technical system to a control model standard without manual configuration and thereby forming the basis for self-X monitoring applications, in particular for energy or consumption management.

In particular, 35% of Germany's final energy consumption is used for the use and operation of apartments and non-residential buildings. This operation causes around 28% of greenhouse gas emissions in Germany. International agreements such as the Paris Climate Agreement call for a significant reduction in greenhouse gas emissions in order to prevent major climate change. To this end, the aim is to increase the renovation rate in Germany to 2%. In terms of control technology, technical systems generally face the same challenges and tasks as buildings.

It is often unclear which modernization measure is the most effective if the current operating status of a system, especially a building, cannot be displayed and thus assessed. This results in the need to first comprehensively record the energy consumption status of a system. Often, significant savings can be achieved by optimizing the existing technical components without having to carry out major modernizations. Therefore, it is first necessary to record the status of the existing system and set up a system control system, especially in the case of building automation and energy consumption monitoring. Various obstacles, such as the high complexity of setting up such systems, high costs and a lack of data, stand in the way of a broad availability of energy consumption monitoring applications. The reasons for the high complexity are usually heterogeneous communication protocols at a so-called field level and the heterogeneous semantics of the information transmitted. In particular, a uniform nomenclature was often not used to designate the same or equivalent or type of technical components in the existing system. In order to integrate the information into the energy consumption monitoring, also referred to as energy monitoring, different protocols typically have to be processed, their information analyzed and integrated into the energy consumption monitoring. The analysis of the data points and the integration at the right point in the energy consumption monitoring usually require a high level of manual effort.

Due to a semantically heterogeneous description of technical components of a system in a system control system, often in the form of so-called data points, and/or due to inconsistent control standards, it is often not possible to search for the power consumption of certain consumers such as all pumps used and to integrate these automatically into a new system control system or user application. The data points typically have to be analyzed manually and integrated into a new system control system or user application.

There are currently no known solutions that automatically process the semantic information of a system control system. For example, a semantic matching service has been specified for building automation data, which uses the BACnet protocol in particular. Each BACnet data point has a name (object name) and a description (description). However, this depends on the manufacturer or the building automation company that implements the system. If the data points are to be integrated into a monitoring application or application at the management level, it must be manually analyzed which data point is involved. Only then can it be integrated into the correct place in a new user application or a new control program. Current solutions for the semi-automated processing of data are based on a behavioral analysis of the control data in combination with metadata. For example, there are solutions that record the course of a data point and assign it to a system together with the unit of the data point. A data point from a building that is around 65 °C can be assigned to a temperature of a heating system, for example. However, these approaches are not yet widely used. At the moment, data from buildings in particular is integrated into monitoring applications by engineers manually assigning it: The data from a building is recorded, analyzed and added to the right places in the monitoring application. This is time-consuming and expensive. As a result, the majority of buildings do not have energy consumption monitoring, so optimization potential remains undiscovered and buildings are not operated optimally.

Based on this, the object of the invention is to improve the recording and/or analysis and/or optimization of existing system controls. In particular, the recording of technical components for software-based optimization of existing systems such as buildings is to be improved.

This object is achieved by the subject matter of patent claim 1 and the independent method claim. Preferred further developments can be found in the subclaims.

The invention provides a computer system for processing control data of a plant with a plant control, with a computer and an interface to a plant control, wherein the plant has at least one technical component which represents at least one data point with respect to the control data, wherein the interface is connected to the computer and the plant control, which exchanges control data with the technical components of the plant via a network, wherein the interface records the control data and transmits it to the computer, and wherein the computer system is set up to classify a data point by means of natural language processing, to assign it to a technical component and/or create a digital control image of the system control.

The invention particularly comprises a semantic processing and mapping of system and/or operating data of a technical system automation, particularly of a building, to a uniform information model. Part of the invention also comprises a gateway that can be integrated into a system that has automation. The gateway collects control data from the technical components of the system. The focus of the invention is in particular on the automated analysis and processing of control data of an existing system control that is not uniformly named or labeled and is therefore heterogeneous using an algorithm. This is based in particular on methods of natural language processing (NLP), a sub-area of artificial intelligence for language processing. Different data point names are processed, interpreted and presented in a structured manner in a uniform, understandable information model. This makes it possible to provide a so-called semantic digital twin of an existing system automation or system control, essentially independent of the nomenclature that was originally chosen when the system in question was created.

In order to be able to optimize the energy consumption of a system, for example a building, the system or operating data, such as power consumption, energy consumption or switching cycles of the technical components, must be systematically recorded, analyzed and optimized holistically. Various systems of an operator usually have a large number of operating data, in particular on relevant consumers, which can be represented or characterized by a data point. Such a data point, also referred to as a data vector, can in particular contain the power consumption of a consumer and/or its name and/or function and/or an energy consumption profile and can also be in the form of a multi-dimensional matrix with data entries. A data vector contains in particular the data of a technical system or an individual system component, which is made available to the outside by the system control system in question. Data points are also used in particular by the control system in question for the connection. system operation, which may be physical or virtual information, which is recorded or generated in particular by sensors, actuators or other components of the technical system.

The computer can also be cloud-based, so that applications in particular can be carried out remotely. The interface is in particular a gateway, which is typically located between the computer and the technical system and is in particular connected to an existing control system, in particular via its network such as a BUS system. A digital control image of the system control comprises in particular a control program-related and/or emulating and/or virtual computer program, which can replace and/or control and/or intervene in the system control of an existing system. The digital control image can comprise a control and/or information model and can in particular provide the control data of the existing system within the digital image.

Through the semantic processing and representation as a digital twin of the system information, the invention can enable automated generation of dashboards for monitoring applications, especially for buildings. By implementing executable rules, different dashboards can be created based on the building data found. If, for example, data from a heat pump is found in one building and data from a gas boiler is found in another building, different applications can be created. Dashboards are therefore generated automatically, tailored to the user's data. The basis for this is a process that maps heterogeneous data to a uniform model.

The result in the form of a digital control image, or a control and/or information model, can - as previously described - also be referred to as a digital twin of a system with technical components. Preferably, all data points of a system are processed and classified by the NLP models, whereby the data is assigned to categories or classes that were defined in a previous training session as part of training runs and learned by the models. This enables patterns and relationships in the control data to be identified. can be learned, which enable a classification of the operating data. On the basis of the classifying data points, an entire digital control image of a system can be created, which can look different for each system. The basic structure of the digital control image in relation to different user applications is usually the same, so that the digital control image forms the basis for automated monitoring applications. Executable rules can be implemented for the applications, which search a digital control image for certain data points. This is made possible in particular by the semantic processing and uniform representation. This makes it possible, for example, to implement the rule: search for every flow temperature of the heating systems in a building and create a line diagram for them. These rules then only need to be implemented once and can then be executed each time the system is started up, so that new user applications can be created automatically. Trends in the system data are then registered. These can form the basis for optimization suggestions or anomaly detection within the system.

A semantic matching service is based on methods from the research area of NLP. NLP is a branch of artificial intelligence that enables computers to understand natural language in order to carry out actions based on it. State-of-the-art NLP models are based in particular on the transformer architecture. It is characterized by an encoder-decoder structure. The encoder is used for language analysis and classification and the decoder for language generation tasks. In order to cover more specific requirements in the areas of classification or generation, today's models are usually characterized by either an encoder or decoder structure and are referred to as pretrained language models (PLM). These are first trained on large amounts of general text (pretraining) and then refined (fine-tuning) or further trained for specific NLP tasks (e.g. text classification). The combination of general pretraining followed by fine-tuning for a specific task is called transfer learning. Based on these PLMs and the specific NLP tasks Text Classification (TC) and Natural Language Inference (NLI), heterogeneous semantics can be processed. 1

Preferably, the computer system is set up to assign and/or interpret different data point names from the building automation using natural language processing based on language models, preferably based on word embeddings, which are generated from transformer models and/or long short-term memory-based models. This makes it possible to achieve a uniform assignment in particular with regard to the name of the technical components and the control system as a whole.

In a particularly preferred embodiment, the computer system is set up to train and/or refine the language models, in particular the transformer models and/or long short-term memory-based models, with texts so that they have a basic understanding of language and/or appropriate links. These models are typically pre-trained on large amounts of text and are refined in a further step on labeled data points from the building automation so that they are able to map heterogeneous data points from the building automation onto the developed information model.

In a likewise preferred embodiment, the computer system is set up to train and/or refine and/or improve the transformer models and/or long short-term memory-based models on preselected data points in a further step, whereby the accuracy of the assignment results and the creation of a control image can be significantly improved. During operation, the data point in particular goes through a multi-stage process which classifies the data point ever more finely (text classification). For example, in the case of building management, the basic function (e.g. supplying air) to which the data point is assigned is first classified. In a next - possibly second - step, the more precise function is classified (here, for example, providing air) and in a possible third step the component (supply air fan). The last step classifies the data point, for example, as a switching command (e.g. supply air fan). A possible multi-stage process of increasingly fine classification can achieve a very high mapping accuracy of sometimes over 95%. The computer system is preferably set up to create a monitoring application from the control and/or information model. A monitoring application comprises in particular an individually adapted software-based application which can control the technical components or with which the user can influence the controls, in particular by means of a user interface. The monitoring application preferably relates to an energy monitoring application.

In particular, the computer system is designed to create a visualization of the control and/or information model. This can in particular also involve one or more dashboards that are created automatically and that the user can use to view data or make further configurations. This can achieve a high level of user-friendliness and/or operability.

The invention also relates to a method for processing control data of a system with technical components, wherein a technical component represents a data point with respect to the control data, comprising the following steps:

Acquisition of control data via an interface,

Sending the control data to a computer,

Classifying a data point using Natural Language Processing, assigning the data point to a technical component and

Creating a digital control and/or information model, which represents a digital control image of the plant control.

In particular, the method comprises the interpretation of different data point designations from the building automation system using natural language processing, in particular based on word embeddings, which are generated from transformer models and/or long short-term memory based models; training the aforementioned models with texts, in particular so that they have a basic understanding of language, and refining the aforementioned models on preselected data points.

The method may also include creating a monitoring application from the control and/or information model. It is also conceivable to visualize the control and/or information model and/or the creation of one or more dashboards.

The invention is explained in more detail below using preferred embodiments with reference to the drawings.

In the drawings show

Fig. 1 shows the computer system according to the invention in a preferred embodiment, Fig. 2 shows the method steps according to the invention in a preferred embodiment,

Fig. 3 an overview of data points and labels for a preferred embodiment of the invention,

Fig. 4 Labeling using the example of a data point,

Fig. 5 an illustration related to Natural Language Inference (NLI),

Fig. 6 a tabular overview of training results of the different models and

Fig. 7 a general structure for building management in companies.

Figure 1 schematically shows the computer system 1 according to the invention in a preferred embodiment, with a computer 5 and an interface 6, wherein the interface 6 connects the computer system 1 to a system control or system 4, which has several technical components 3 or is also networked in a network 7. Control data 2 are exchanged with the computer 5 via the interface 6.

Figure 2 shows schematically the most important steps according to the invention, which steps include:

Sl : Acquisition of the control data 2 via an interface 6,

S2: Sending the control data 2 to a computer 5,

S3: Classifying a data point using Natural Language Processing,

S4: Assignment of the data point to a technical component 3 and S5: Creation of a digital control and/or information model 8, which represents a digital control image of the plant control.

One way to avoid the need to manually enter and assign control data 2, especially building automation data, is to implement a Semantic Matching (SM) service, which automatically processes heterogeneous semantics and maps them to a semantic standard. The prerequisite for such applications is a uniform structure of the information, i.e. a uniform information model on which the components are based. In order to be able to automatically integrate the components or their automation stations into, for example, energy monitoring, the information must first be integrated into a digital twin (e.g. the administration shell (VWS)). The result is a VWS per automation station with the submodels "Asset Interface Description" and "Datapoint Information". The submodel "Datapoint Information" is then processed by the SM service. The heterogeneous information is mapped to functions of the TGA using artificial intelligence methods, specifically Natural Language Processing (NLP). The result is in turn mapped to the submodels "Operation Information" and "Monitoring", which are referred to as structural information. The structural information obtained through processing in the submodels is integrated into the energy monitoring through rule-based automation, thus enabling so-called Self-X energy monitoring applications for existing systems.

In the area of Industry 4.0 (14.0), the concept of the VWS has been established as the basis for interoperability. The VWS is the digital representative of an asset in the digital world. The information model of the VWS defines the structure of how information from different assets must be structured. The central building block of the VWS are submodels that map the properties and functionalities of the assets. Submodels represent, for example, the topics of identification, design or configuration. Access to the information of a VWS is realized through interfaces that are provided by the VWS. In order to enable automated access to the interfaces of different VWS, these interfaces are standardized. In the literature, four levels are often described that are passed through when standardizing interfaces. For example, levels 1 and 2 of different interfaces are described, e.g. “GetSubmo- del”. This interface can be called to access a specific submodel of a VWS. The general interfaces can be provided in different services.

Figure 3 gives an exemplary overview of the data points and the assigned labels. The initial data set consists of 8,809 data points from the cities of Cologne and Hamburg. These were labeled and assigned to the different functions and components of the ordering method.

In the first step, all data points were assigned to one of the basic functions of the TGA (heat, air, cold, media, power supply, security, transport). If a data point could not be assigned to any of the functions due to missing or incomprehensible names or descriptions, it was assigned to the category "Other systems". The basic functions of transport, power and cold supply were underrepresented in the data set, so that the data points associated with them were only labeled up to the first level. The basic functions of air, heat and media supply were divided into further functions and the data points were each assigned to one of these functions (dashed line). In the third step, the data points were assigned to components or higher-level areas of the TGA (dotted line). In the last step, all data points are given a label that expresses the actual information of the data point. An example labeling process for a data point is shown in Figure 4. The individual data sets were used to train NLP models.

Current state-of-the-art models are typically trained using transfer learning. First, a PLM must be selected for this. Since the data points of the dataset consist of German names and German descriptions, only German PLMs were considered. Several German models are described in [Chan, B., Schweter, S. and Möller, T.: German's Next Language Model. Proceedings of the 28th International Conference on Computational Linguistics. Barcelona,: International Committee on Computational Linguistics 2020]. Of these, the "gBERT-base" model was selected as the basis because it delivered the best results in initial tests on the created dataset. The model is based on the architecture of the BERT-base model. and is an encoder model. Encoder models are particularly suitable for classification tasks such as TC. The models of the first three levels (basic function, second function, component) were trained using TC, the last level, data point level, was trained using NLI. The last level was not trained using TC because, on the one hand, the number of resulting models is too large and, on the other hand, the amount of training data for the individual components 3 is too small.

Figure 3 shows 35 technical components 3. If fine-tuning on TC were also carried out for the fourth stage, 35 models would have to be trained. Since an even larger number of components 3 is to be expected in the future due to the expansion of the training data set, the number of models to be trained would continue to rise, which would lead to increasing costs for training and the models in operation. The data points of the components 3 listed are also unevenly distributed, for example 807 data points are assigned to the pump, while only 16 data points are assigned to the exhaust air flap. This means that underrepresented components 3 do not have enough data to train high-performance models. By using the NLI method, it is possible to train just one model that can be used for each component, as shown in Figure 5 as an example.

In the NLI approach, the name and description of the data point are used as a premise. This is compared with a hypothesis consisting of a defined introduction and a label (e.g.: The data point describes: alarm message). Since a total of 241 labels are possible, each premise is compared with 241 hypotheses. The model learns to predict which of the hypotheses fits the data point (entailment) or does not describe the data point (contradiction). During training, all labels (241) are compared with all data points (5,485). The four-stage classification process makes it possible to reduce the "candidate labels" during operation to the number that are possible for the predicted component. The results of the various models with training parameters are shown in tabular form in Figure 6.

The Fl score was used as an evaluation metric: Fl Score = 2 * (Precision * Recall) / (Precision + Recall). The results show that the different models can be used successfully to classify data points. With the exception of the Media model, all models achieve a value of over 95%. The trained models were combined in a pipeline and are provided in operation as a 14.0 service.

For automated monitoring, a general structure was first designed to show how this can be placed in the context of a company's building management (see Fig. 7). For this purpose, a hierarchical structure was created that consists of several levels with VWS. The top level is the company's VWS including the associated submodel, which contains the information about the company. A company can have several properties, on which several buildings are constructed. Up to this level, the VWS must be created manually. The other levels and submodels "Building Operational Information" and "Monitoring AMEV Building" are created automatically using the information from the submodel "NLP Classification Result". All automation stations in a building are read out and processed by the 14.0 NLP service. Depending on the result of the classification, different VWS are created. If all data points in a building are assigned to the basic functions of heat and air supply, for example, two VWS are created (one per basic function). This is done for the next levels: system (e.g. two separate heating systems within a building), sub-system (three heating circuits and one generator per system) and component (e.g. pump). The appropriate data points are then assigned to the respective levels within the sub-models "operating information" and "monitoring". In this way, a hierarchical level structure (structural information) is created automatically (Self-X) from VWS. Monitoring applications can then be created using automatically executable rules. List of reference symbols

1 computer system

2 Control data 3 technical components

4 Annex

5 computers

6 Interface/Gateway

7 Network 8 Control and/or information model

Claims

patent claims 1. Computer system (1) for processing control data (2) of a system (4) with a system control, with a computer (5) and an interface (6) to a system control, wherein the system (4) has at least one technical component (3) which represents at least one data point in relation to the control data (2), wherein the interface (6) is connected to the computer (5) and the system control, which exchanges control data (2) with the technical components (3) of the system (4) via a network (7), wherein the interface (6) records the control data (2) and transmits it to the computer (5), and wherein the computer system (1) is set up to classify a data point by means of natural language processing, to assign it to a technical component (3) and/or to create a digital control image of the system control. 2. Computer system according to claim 1, wherein the computer system (1) is configured to assign different designations for a technical component (3) by means of natural language processing on the basis of language models. 3. Computer system according to the preceding claim, wherein the computer system (1) is arranged to train the language models with example data. 4. Computer system according to the preceding claim, wherein the computer system (1) is configured to train and/or refine the language models with preselected data points. 5. Computer system according to one of the preceding claims, wherein the computer system (1) is configured to create a monitoring application from the control and/or information model (8). 6. Computer system according to one of the preceding claims, wherein the computer system (1) is configured to create a visualization of the control and/or information model (8). 7. Computer system according to the preceding claim, wherein the computer system (1) is configured to create one or more dashboards. 8. Method for processing control data (2) of a system (4) with technical components (3), wherein a technical component (3) represents a data point with respect to the control data (2), comprising the following steps: - Acquiring the control data (2) by means of an interface (6), sending the control data (2) to a computer (5), - Classifying a data point using Natural Language Processing, - Assignment of the data point to a technical component (3) and - Creating a digital control and/or information model (8), which represents a digital control image of the plant control. 9. Method according to the preceding claim, further comprising the steps - Interpretation of different names for technical components (3) using natural language processing based on language models, - Training the aforementioned models with texts and - Training and/or refining the aforementioned models on preselected data points. 10. Method according to one of the two preceding claims, further comprising creating a monitoring application from the control and/or information model (8). 11. Method according to one of the three preceding claims, further comprising a visualization of the control and/or information model (8) and/or the creation of one or more dashboards.