Model adjustment
Continuous optimisation of the model quality
When using NeuroModel as a modelling tool, models can be automatically adapted to new process states with NeuroModel Script Online. The models learn process changes and are constantly tracked to the process. This is particularly necessary for ageing processes (such as catalytic converters) or wear. Post-training is script-controlled and simulates the behaviour of the engineer who performed the initial modelling. The training is triggered either cyclically or when the model quality decreases

Post-training – NeuroModel
- Perform post-training automatically with NeuroModel Script
- Automatic retraining ensures continuous model quality with NeuroModel.
- Data extraction from different data sources and automatic adaptation to new data sets
- Exchange of models at runtime
- Bumpless switching between different models possible in APC Professional
- Start of training by NeuroModel Script takes place automatically or via trigger
- PreSetups and templates of pre-trained networks guarantee fast generation of adapted models
NeuroModel
NeuroModel creates models of processes or sub-processes through automated machine learning. It is specially adapted to the needs of power plants and the manufacturing industry. NeuroModel provides in-depth knowledge of the interrelationships in your processes in a very short space of time, ensuring you find optimal parameters for process control. Models are automatically generated and analysed from historical data.
Intelligent algorithms enable modelling even with just a few data sets. An expert system does the work for you and intelligently decides which solution path is optimal for your problem. Throughout all of this, the configuration time is only a few minutes.

NeuroModel Script Online
Each project in NeuroModel generates a model and an editable ASCII script (XML-like) with precise rules on how to retrieve data and how to handle new data. The script can be executed automatically by NeuroModel Script, extracts data from data sources, cleans them and automatically adapts a model to new data sets. It thus contains a kind of “recipe” that precisely describes the way the model is created. NeuroModel Script can be regularly started automatically or via a trigger – for example, event-controlled from APC Professional. If NeuroModel is used as a modelling tool, models can be automatically adapted online to changes in the process with NeuroModel Script.

APC Professional
APC Professional is a block-oriented and universal control and simulation programme. With it, more than 800 user-configurable function blocks, as well as user-defined blockes, can be interconnected fully graphically to form any dynamic system. APC Professional(R) is used in projects where complex control strategies are to be implemented in process plants or interfaces to a heterogeneous system environment are required.
In addition to 200 different interface blocks, APC Professional also contains blocks for any white-box models, NeuroModel models, GenOpt optimiser blocks, extensions for the use of statistical models and an open programming interface to connect almost any external calculation module. Operation is as easy and intuitive as a Lego. The blocks are simply placed on the surface and connected graphically whith the mouse.

HAZOP+
Databased Process Safety and Operational Excellence

Digitalisation and Industry 4.0 enable continuous, data-based and automated investigation and optimisation using AI-based methods of processes of all kinds. This opens up new potentials for HAZOP studies and Operational Excellence (OPEX). However, HAZOP and OPEX are often treated separately in companies. This means the great synergy potential that results from linking AI-based HAZOP and OPEX studies remains untapped. Synergised data utilisation and analysis can save time, effort and costs and increase plant safety and operational excellence simultaneously.
YOUR ADVANTAGES
- When preparing the HAZOP study and data-driven OPEX project, you leverage synergies that save time and effort
- The HAZOP study benefits from the systematic You use the in-depth analysis of your process using modern big data analysis methods and automated machine learning for data-based plant and process optimisation.
- You use the in-depth analysis of your process using modern big data analysis methods and automated machine learning for data-based plant and process optimisation.
- You eliminate the risk of exceeding safe limits in plant and process optimisation and gain optimisation leeway
- Your plant is operated with increased stability at the optimum operating point, which means savings and increased profitability for you
- Your plant safety is up to date and fulfils the legal obligations
- The end result is that your production plant becomes safer and more economical.