Abstract: It is becoming more and more complicated for glasshouse growers to steer the purchasing and selling of energy. It is time-consuming and complex due to the enormous amount of factors and data that have to be taken into account when making a bid.
To help greenhouse growers optimize their energy management and lower energy costs, AgroEnergy has developed BidOptimal. An innovation that automates and optimizes the daily bidding process of growers: at what price do I trade what amount of gas and electricity?
BidOptimal calculates APX bids for four days in advance, whereby the heat cost price for the grower is minimized within all boundary conditions. In this way, the grower gets a good return and valuable energy is used as sustainably as possible.
The greenhouse grower obviously has no knowledge of the exact mathematics that BidOptimal has inside, but without the intelligence of CQM the tool would not work. Every day, BidOptimal does only have two hours to carry out all the calculations for more than 100 growers.
With statistical models and deep learning, we predict the relevant data ('predictive analytics'). To achieve good predictions, many different data sources are used in real-time. The optimization model then determines the best solution for almost 30,000 variables within 15,000 boundary conditions, resulting in APX bids ('prescriptive analytics'). A typically data science project at CQM.