Skip to content

06.0_Batman_Power

Jump to bottom
Rohak edited this page Feb 10, 2017 · 9 revisions 
Team
Joshua Updyke <josh.updyke>
Peter Keen <pwkeen>
Logan Williams <logan>

Goal

The power system for the Protei 006 will be more complicated then the previous versions. If the goal is to achieve some level of autonomous system, a reliable power structure will be needed. One possible solution is solar power and batteries. For this to work several details will need to be defined and specific components will need to be selected that match the needs of this new Protei.

Power Storage

Their are a wide range of battery options available, each with their own pros and cons. Cost is one of the major limiting factors, but weight will also be a concern. It will be vital to keep the weight down for this entire project. The protei will be more powerful if we keep it light. Additionally, some batteries have special charging requirements that make them more complicated to work with.

Lead acid NiMh Lithium Ion
Energy density 30 Wh/kg 200 Wh/kg 300 Wh/kg
Power density 180 W/kg 300 W/kg 300 W/kg
Cost for 120 Wh battery pack (~12V/10Ah) $18 $100
Max. discharge rate of 120Wh battery pack 200A 30A 10A

The table above summarizes some of these pros and cons and the tradeoffs that will have to be made. Important design parameters to determine now are: What is the typical, and maximum current draw of Protei? and What is our energy storage requirement?

Additional factors that will influence our decision: Lead acid and NiMH batteries are much simpler to charge than Lithium Ion, and especially with Lead acid, you can get away with being much less "careful" about how they are charged. For a prototype, weight might be less of an issue than in the final version.

NiMH seems to be a good compromise between weight, cost, and ease of use.

Power Generation:

Solar Cells: In order for the Protei to continue to run for extended lengths of time (days, weeks, maybe months) it will need a way to recharge the power. Solar cells seem to be the simplest solution.

Thin Film Photovoltaics may be another option, offering the opportunity to provide a shape conforming skin to the finished vehicle that generates a charging current.

http://www.nrel.gov/pv/thin_film/about.html

What is the average power consumption? This will be a key determining factor when deciding how many solar cells we need. We will need to calculate an estimate of the power that will used. The motors, control system, communications, guidance and any other electronics will need to me measured. It is also a good idea to build in a Factor of Safety (FOS), this is a margin for errors. If we estimate that we need 80 watts a day, we should plan to generate 100 watts. This would have a FOS of 1.25.

How many solar cells do we need? With a good estimate of the power consumptions, we can decide on how much power we need to generate. This is limited by the physical space available. The solar cells can not be too large or we will not have anywhere to put them and they will add weight. How many hours of sunlight are available will need to be investigated? Cloudy days and adverse weather will play a part as well.

Wind Turbine: many conventional sailboat are equipped with wind turbines. In addition of the energy efficiency and lower price, other factors to take into account are : local resources where the machine will be operating. deck space (can we fit the turbine, does it change the balance, does it affect the dynamic behavior of the vessel?), durability, waterproofness, cost/watt, long-term maintenance, legal aspect (is it legal to have a wind turbine on an unmanned surface vessel?). What is specially interesting with wind power : in heavy winds we'll need more electricity to power our motors, the production of a wind turbine would match that change in the demand and guarantee control in worst times. Durability of a wind turbine is probably the biggest concern.

Wind turbines allow the same amount of power generated in a much smaller area. However, moving parts and difficulty to waterproof will make installing wind turbines much more difficult than solar. Perhaps this could be mitigated to some extent with Vertical Axis Wind Turbines, but even then, moving parts = failure points. With these considerations, my opinion is that solar power is the best choice for Protei. - Logan

Hybrid power: for many reasons, it makes great sense to have both solar and wind power generation on board. Redundancy. Redundancy. The weather conditions are changing constantly. Just think about a storm at night - you need your wind turbine to fill up. Think about long sunny days with a weak wind : you need your solar panel. Some functions are more vital than others (GPS locating, distress signal handler, self-inflation...) and may require a separate electric system. We may need hybrid power.

Conserve Power: Another key part of the power system will be conserving power whenever possible. Most electronics have a sleep mode. The more power we can save the cheaper and lighter the system will become. Will the Protei sail during the night? If not can it power down and conserve energy. What will happen if the system is low on power? A reserve mode can be programmed to limit power usage to only the essential systems. Perhaps only to communications, the Protei can be found. This could happen if there are several days without much sunlight, or if part of the solar cells become covered with debris or oil. The system can save power, wait until better conditions, and then continue working.

Wave power? See for example the Salter´s duck - invented at the University of Edinburgh by prof Steven Salters.

Hydroturbine? See for example DUOGEN

Power System:

The overall power system should be designed to be flexible and scalable, so that new Protei versions do not need new power systems. Here is my initial design for a power system. Note that thick arrows are power transfer, and thin dashed arrows are data transfer.

1

The Arduino based MPPT controller would be based off of Tim Nolan's design, with several modifications: components chosen so that adequate current can flow; sensors that allow the power consumption of the load to be measured; modification to the firmware so that it can charge NiMH batteries with the ∆V method.

Since the battery will be connected to a load while it is charging, the modification is slightly more complicated. The current software has four states: ON, BULK, FLOAT, OFF. To modify this, there will be a second state parameter, CHARGE, and DRAIN. We are in CHARGE when I_load < I_total. When I_load > I_total, we are in DRAIN (draining the battery). In CHARGE mode, we can be in ON, BULK, FLOAT, or OFF, depending on load levels and on power output from the solar panel. in DRAIN mode, we can be in ON, BULK, or OFF, depending on power output from solar panel. Only in CHARGE mode do we do ∆V checking — otherwise, the battery is not actually being charged.

I_load is the total power going to onboard electrical devices. I_total is the total power coming out of the buck converter. The power drawn from the battery is not measured, but can be calculated by I_load - I_total.

Since most of this would be software, it provides many opportunities for increasing efficiency -- for example, when the MCU heard from the power controller that the Solar panel power was dropping, it could disable certain electronics on the Protei to reduce its power demand.

Here is a first guess at what components would work:

Basic system total cost (just the battery): $100

  • With 10W solar panel: $201
  • With 50W solar panel: $376

At this point, the most important consideration are the points that are mentioned above, in the battery section, essentially, how much power do we need? I think that the electronics onboard (communication, MCUs, etc) will consume <3W, with radio and GPS and everything running at full power. Therefore, the most important consideration must be the power consumption of the motors. Most 12V winches available (this is what is on the main Protei v6 page consume far more than the amount of current (10A), assumed to be maximum in the above. The smallest winches I could find in a Google search consumed 12-20A. I think we should determine exactly how powerful a motor is necessary to accomplish the mechanical tasks we need.

  • Logan

Home

Clone this wiki locally