Solar forecasting for the Eclipse X Solar Car
This week, we have a guest blog post from the Eclipse X solar car team, who are accessing the Solcast API to support their race strategy
** ABOVE: The Eclipse X.1 at an unveiling earlier this year **
In rally type competitions, such as American Solar Challenge and World Solar Challenge, you need to have a plan. And not just any plan, a well thought out plan! After all, your solar car will be travelling more than 3000km - powered solely by the sun. This is no small feat, and proper energy management is crucial if you plan to cross that finish line. Success requires adjusting your optimal cruising speed for every stage of the race. Going too fast means you deplete your stored energy before arriving. Going too slow means you’ve wasted available energy – and worst of all, allowing other teams to catchup to you (we’re in it to win it, after all!).
The Role of High Quality Solar Irradiance Data
In order to properly manage our energy reserves and find our optimal cruising speed, we at Eclipse run elaborate simulation models taking many different parameters into account. Weather, wind speed, air density, air pressure, road elevations and our own vehicle’s efficiency are all taken into consideration when calculating optimal speed. However, most importantly, Eclipse X is a solar car. It has four square meters of highly efficient solar panels incorporated into its hull, continuously charging the battery storage system, on the go. Thus, the most important factor to consider when calculating optimal driving speed is the available solar irradiance (the amount of sunlight that reaches our solar panels)!
Using Solcast for Solar Irradiance Forecasting
This prediction of the available solar irradiance (and thereby our available energy for moving the Eclipse X forward!) is where Solcast comes in. Solcast is a global solar irradiance and power forecasting company, who delivers rapid-updating information about the available sunshine via an API service. In order for our race strategy to succeed, Eclipse X requires calculations to include the most precise estimates of the available solar radiation along the race route as possible. Our team trusts Solcast to provide us this information and through extensive testing, we’ve found the data received to be very accurate. Furthermore, the API is well documented and easy to use with Matlab, which allows us to easily pipe-in the data into our racing strategy software. In short, Solcast is a great match for our needs in solar irradiance information, and a key part of our race day strategy.
How We use Solcast Solar Forecasting Data
In solar car racing, the goal is to cover as much distance in a day as possible, without compromising your following days of racing. Solar car races typically last between five to 10 days in length, depending on weather and team performance. In order to go the distance, the team must consider the amount of energy gained during your morning charge (pre-race), the amount you are gaining as you are travelling along you route, and the amount of energy you hope to collect during your allotted evening recharge time, and compare these values with your energy expenditure. And don’t forget - the faster you go, the more energy you consume, but going faster is also how you win!
In order to optimise these many different factors, the Eclipse team has constructed an elaborate Matlab simulation model that calculates optimal cruising speeds for our solar car. This model allows us to calculate our projected energy expenditure for a given race speed. We can then compare our expenditures with the real-time (live) and forecast solar irradiance available along our route (thanks to the Solcast API!). This allows us to then predict the amount of energy left in our battery as a function of the distance covered. The script repeats the process for various race speeds, allowing us to compare different scenarios and make an informed decision. The following graph is an example of the output of our simulation model.
• Caption 1: On the left, we have a graph showing, on the x-axis (front to back), the distance covered, from 0 to 250 km, on the y-axis (vertical), the amount of solar energy available at ground level (W/m2) and on the z-axis (horizontal), the time (in half hours) for the three next days after the starting time. The red line represents the position of the solar car in distance and in time according to the simulated speed. Solar energy data are graciously given by Solcast. • Caption 2: On the right, we have the projected state of charge of our battery (Wh) according to distance covered (km).
Our next race: The World Solar Challenge
We will next put our strategy (and entire team!) to the test during the 2019 Bridgestone World Solar Challenge (BWSC): a 3000 km rally to cross the Australian desert from Darwin to Adelaide whilst relying solely on solar energy for power. Starting on the 13th of October, 53 teams from 24 different countries will compete for the rights to call themselves World Champions. We at Eclipse have been preparing ourselves, our solar car and our racing strategy for over a year for this challenge and with less than a month to go before race day, we have never been more confident in ourselves. Of course, our participation in such an event is largely due to our commercial partners, such as Solcast, that help and support us in our endeavour. We are immensely grateful.
Follow our progress during the 2019 BWSC on Facebook and Instagram!
You can also visit our website at: https://eclipseets.ca/wordpress/en/