Dear all,
I am applying MPPT techniques for a 400 KW grid connected system. I use optimization technique to tune PID controller parameters embedded in the MPPT controller. I get the response of the power as shown in the picture. the problem is with the settling and rise time of the power signal. when i use stepinfo command i get too large numbers for both of them. for example i get the rise time equals 75600 s ! and get a rise time that is larger than the settling time ? what may cause such numbers ? and how to get reasonable values ?
Please elaborate on the objective(s) and constraints you used in optimising the MPPT performance.
You may refer to the reduced order models presented in the following article; they will give satisfactory performances.
https://www.researchgate.net/publication/322022340_Reduced-order_modelling_of_solar-PV_generators_for_small-signal_stability_assessment_of_power_systems_and_estimation_of_maximum_penetration_levels
Suggestions:
1. Check for how long you are running the simulations for. Zoom in the settled version of the curve and see whether there are small oscillations in the power curve.
2. If there are oscillations, you will realize why you are getting a large settling time.
3. Since I don't know how and where you are using this MPPT algorithm, there might be an issue with the power electronics design also.
4. Another suggestion will be to check your optimization algorithm also.
Mohamed EL-Shimy Dear professor, I am using the incremental conductance MPPT technique. there is only one objective function which is the error signal produced from the following relation : I/v+dI/dv = error. the lower and upper bounds for KP KI and KD are zero and 5.
Syed Ahmed Raza I tried to run the model for 15 sec and it seems it reaches to steady state with very small oscillations. I use step info command and i define the error signal so that it ignores this small oscillations however i get too large numbers
Until or unless I see your simulations, any speculations that we make will not be fruitful. Try to run the simulation for at least 1-2 mins to check if the system is really stable or not.
- Badges
- Science topic