Sørensen, Kresten Kjær1; Nielsen, Jens Frederik Dalsgaard4; Stoustrup, Jakob3
1 Department of Electronic Systems, The Faculty of Engineering and Science, Aalborg University, VBN2 The Faculty of Engineering and Science, Aalborg University, VBN3 Automation & Control, The Faculty of Engineering and Science, Aalborg University, VBN4 Aalborg University Space Center, The Faculty of Engineering and Science, Aalborg University, VBN5 Aalborg U Robotics, The Faculty of Humanities, Aalborg University, VBN6 Strategic Research Centre on Zero Energy Buildings, The Faculty of Engineering and Science, Aalborg University, VBN
The amount of food transported long distances in reefer containers is constantly increasing and so is the cost per mile because of rising fuel prices. One way to reduce the cost is to minimize the energy consumed by reefer containers through a better controller but in order to achieve this a fast and flexible simulation model is needed for controller development. The simulation model may also be used for developing fault diagnosis methods for the reefer container and thereby further lowering costs by reducing the amount of functioning spare parts that is replaced and by providing early warning for faults enabling preventive maintenance. In this paper the feasibility of using different simulation methods is assessed with the goal of identifying a fast but accurate method that works well in a multi-rate environment. A modular multi-rate simulation environment for a dynamical system consisting of components with different dynamical speeds is presented with an improvement of previous results. The simulation speed is improved by 350% with no reduction in accuracy of the solution, by substituting the MATLAB ode15s solver with an explicit first order solver with a step size calculation algorithm that ensures numerical stability and that the error is bounded using a minimum of calculations. The reefer container model is simulated using both ode15s and the proposed method both in multi-rate and monolithic configurations. The results are analyzed and compared with respect to speed and accuracy.