Tracking the behaviour of a system enables to operate it at peak level.
Improvement can have different flavours: reduced consumption, increased velocity, diminished noise, etc.
The aim here is to predict the behaviour of engineering systems. This allows us to simulate different operating scenarios and adjust the control variables to improve efficiency.
But the output of a system arises from complex interactions between all the variables.
In such a dynamic environment, ordinary forecasting techniques are not sufficient, and more sophisticated methods are needed.
The objective is to effectively untangle all the factors that influence the performance of a system.
But analysing multiple factors is complicated, to say the least. We need both rich sensor data, along with complex predictive models to analyse it.
Performance optimization can consider different targets:
Neural Designer uses machine learning to model the behaviour of systems and fine tune the control variables to optimize their performance.
The next image shows a representation of a neural network that could be used for this case.
The inputs to the neural network include initial states, environmental variables, control actions, etc. The output from the neural network is the predicted performance of the system for that scenario.
To conclude, machine learning is a very powerful technique to predict the performance of engineering systems. These predictive models help us to improve efficiency or reduce energy consumption.