Help with Simulink Model of Solar-charged battery charge & discharge curves

[Paniolo]

I have built a small, lightweight pontoon boat that uses an electric trolling motor as the sole means of locomotion drawing power from a single, solar-charged marine battery. The electrical system is as follows: two 135W solar modules are wired in parallel then wired (input) into a charge controller; the charge controller is also wired (output) to a 12V 75Ah marine battery; the 12V DC motor is also wired to the battery.

I need to first create simplified Simulink models of the battery, the PV solar array, the charge controller, and the motor. Then, I would like to create a Simulink circuit that first models these components as integrated in real life. The Matlab Simulink circuit I am trying to emulate is found here: https://www.mathworks.com/help/physmod/sps/powersys/ref/battery.html This is much more sophisticated than I need--I simply want to build a circuit in which I can insert or use the empirical data I have gathered from experiments. Ultimately, I'd like this Simulink model to reasonably predict how long the battery will last given a certain amp-draw from the motor while accounting for the input from the PV array. I would like to use data that I have gathered in the Simulink model to validate it.
I have all the data and specs on the components, so I can provide some more of that if it helps clarify my query. Here is some of the data I have gathered:

-I used a digitally controlled battery analyzer to discharge the fully charged battery at a constant 3.75A. I considered 10.5V as fully discharged per manufacturer specifications. I then fully charged this battery using an automatic battery charger (not the charge controller) with approximately 40A. Both of these graphs are uploaded. I would like to build a circuit that recreates these two curves--the discharge and recharge.

-I also submerged the trolling motor in a large bin of water and ran it non-stop for over 8 hours. The solar-array was able to fully charge the battery while the motor simultaneously discharged the battery. Per the manufacturer, the motor draws 55A per hour at full speed. The experiment was run with the motor at a constant half speed, but the actual amp-draw at this speed is unclear.
Again, I am trying to emulate the generic power_battery exampled cited above from the Mathworks.

Any suggestions?

 

[BradtheRad]

Since you made a real-world setup, I suppose you're interested in real-world scenarios...

* Such as how much charge do your PV panels provide on a cloudy day?

* What is the net charge/draw on the battery in various conditions?

* Can you travel 4 hours distance, and return the same distance, before dark, and not run down the battery?

-------------------------------

It is important to test PV panel output while connected to the battery. It will be different from results using a simple resistive load.

------------------------------

A PV panel puts out (in terms of extremes), either low Amperes at high V, or high Amperes at low V. You get maximum power output at some range in the middle.

I have not seen, nor have I been able to figure out, a simple circuit to model these characteristics. Your simulation program may or may not contain a model of a PV panel.

The experiment was run with the motor at a constant half speed, but the actual amp-draw at this speed is unclear.

You can estimate this as half power, which prospectively is 70-75 percent voltage and 70-75 percent Amperage. Then you use the formula V x A = W.

 

[Paniolo]

BradTheRad:

Thank you for your input.

I overbuilt the PV Array so that I would not have to worry too much about getting back to shore (or the MPP of the PV Array). In November here in AZ (PHX area), I ran the motor at half speed with the PV array attached to the battery via the multi-stage charge controller. The array quickly charged to full the battery and maintained at that level until afternoon. In the end, the battery was still ~75% full--a comfortable margin. Whay I'd like to do is build the Simulink circuit that models this actual data, and then use this now-validated model to input NREL's available monthly avg. solar radiation data for the rest of the year and predict battery performance (at least better than a guess).

I've derated the PV array accounting for heat-related losses, parasitic losses in the wire, etc. and my calculations were within less than a 1% of NREL's 5% power derating, so for my purposes I can use my derated numbers as input.

My problem is I have all this data, but I don't know Simulink well enough to build the circuit that will allow me to simulate the performance graphs I recorded.


*

 

[wizpic]

paniolo
The first graph in your first post of the charge curve from full discharge to fully charged, Is that simulated in software or real time hard wear, on looking at it It's totally wrong type of cuve
How often was the samples taken of the battery voltage and the voltage seemed to have gone fully high at top end for my liking. the picture attached is what it should look like to me depending on what type of charger you used. I would think even a MPPT charger would be the same or close not so many spikes in it.

If your battery is is 75amp hour and your load is say 25amps continuous then the my calculations says that it would only last 1.576 hours to reach 10.5 volts and at 55amps then 0.612 hours at 3.25amps then it would last 18.231 hours depending on how good the battery is of course.

I think you need a watt/ampere hour meter what is exactly you want to do build a data logger type of thing to SD card ??
I've done an ampere/watt meter and recorded all the results to an SD card then used a littler program to plot and view the data.