Re earlier: "my load as u can see in the schematic, is a fixed load, 5.6 ohms/ 20 W. since im using solar cells, my input voltage and current will change sometimes, therefore i have to regulate to always extract the maximum power from the cells and protect the cells too."
Firstly, re protecting the cells, that should not be a problem. Cells are pretty robust things - you can short them, you can pump current in reverse through them etc.
Re MPPT, the simplest answer here is not to bother. Just oversize the panel.
But if you want to build MPPT, for the sheer technical fun of it, then you need to get a few basics working. There is no point trying to extract 10% more efficiency if your MPPT is only running at 30% efficiency.
So you need a very good switching regulator design, and I don't think you have that yet.
If you did, it might be useful to still think through the flowchart above. What you want are maxiumum watts, not maximum volts nor amps. So measuring volts is only part of the story. In thinking about MPPT I think it helps to use two really big capacitors. You can put them there if you like in a real circuit and it doesn't alter anything. So put 10,000uF across the solar cell, and put 10,000uF across your resistor. When this is working you might have 18V on the solar panel and whatever (5V say) across the resistor.
The job of the switching reg is to translate 18V to 5V without losing power. So you grab a bit of energy from the first big capacitor, put it into an inductor, then take it out of the inductor at a lower voltage and put it into the second capacitor. The first cap goes down a bit, and the second one goes up a bit.
From a picaxe perspective, you want to measure watts, so you need to measure volts on both capacitors, and current into both capacitors. You will need to smooth those values, so you will need dropping resistors, op amp amplifiers and RC filters before you feed your 4 values into a picaxe. Then you will need to multiply volts by amps to get watts.
But you don't need to build anything to think through how this needs to work. Consider a solar panel running at midday in full sun. Your MPPT circuit might have determined that you can get the most power out of that by loading it at 17V. You don't need to know the power curves or anything, all you need to have worked out is that at 17V you get x watts, but if you load it at 17.1V or 16.9V you get less watts.
Let's say a cloud goes over. The volts on that first capacitor starts falling, and if you didn't do anything you would have no volts at all pretty soon because you are drawing too much current out of that first capacitor. So you need to take less power out. How?
Well, your algorithm was happily sitting at 17V, and every now and then it was experimenting by letting the volts go to 17.1 or 16.9, and measuring the watts. Now the cloud has gone over, you find that you get more power at 16.9V. So you let the volts fall to that level by drawing less energy out of the first capacitor (ie by making the pulse width narrower). If you sample too infrequently, you might overshoot, so the cloud goes over and the volts fall to 14V, but you keep sampling, and eventually you find that the new optimum volts is 16.4V.
Technically, you only need to measure watts in one place, not in two, but it is helpful to measure watts in and watts out so you can work out the efficiency of your switcher at various current levels.
And if you do that, you might actually find that your switcher was losing more power than just connecting the panel up to the resistor.
But MPPT is still a fun thing to play with. I built one once out of discrete op amps, by converting the volts and amps into an op amp logarithm circuit, then adding the logs, then doing an antilog. So I have a real appreciation of the beauty and simplicity of picaxe - one multiply instruction replacing about 10 chips!
