Neural network training Help

 have to write a forecast for : Irradiation, Windspeed

I cut the xls Data, because the original one is too big. Or Take the CSV. Data

After this i want to calculate the Load, but that's ok, i can do this, i just need help with fitting my data into this.

can someone help me for training a neural net for this ?

%%%%%%%%%%% THAT IS WHAT I HAVE RIGHT NOW %%%%%%%%%%%%%%%%%%%

clc clear clear all

%% Import Data

% Wind and Irradiation

file='Wetterdaten.xls'; Wetterdaten=readtable(file);%Import Data as a table

DateWI=datenum(Wetterdaten.Datum);%Calling the coloumn,Convert String into num TimeWI=Wetterdaten.Zeit;%Calling the coloumn DateAndTime=datetime(datestr(DateWI+TimeWI+1462));%Combining Date and Time, formating into datetime Irr=Wetterdaten.Globalstrahlung;%Calling the coloumn Wind=Wetterdaten.Windgeschwindigkeit;%Calling the coloumn

%Temp %Druck %rel.Feuchte

%Cp-Wert Enercon Anlage

%% adjustment of the imported Data

%time t=(datetime(2018,1,1):hours(1):datetime(2018,10,1))';% missingTime=ismember(t,DateAndTime,'rows');%

t(end)=[]

%plot(DateAndTime,Irr,'.') %hold on

%time Format t.Format='dd-MMM-yyyy';%change Format into Date Date=datetime(cellstr(t)); t.Format='HH:mm:ss';%change Format into Time Time=duration(cellstr(t));

%interpolation of wind velocity v=interp1(DateAndTime,Wind,t,'pchip'); v(~missingTime)=NaN;

%interpolation of solar irradiation i=interp1(DateAndTime,Irr,t,'pchip'); i(~missingTime)=NaN;

%plot(t,i,'.') %hold off

%interpolation of temperature

%% 3D plotting x=unique(Date); y=unique(Time); z=reshape(i,length(y),length(x));

surf(x,y,z,'Facecolor','interp');

%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

Answer 1

[Stephan]

Hi,

i guess this question is related to the other 2 questions you asked. The basic procedure is retained as shown in this answer. Just start with the basics, read in the documentation and work through the examples given. The following documentation links may be helpful for your purposes:

1. Getting Started with Deep Learning Toolbox
2. Webinars at the Tutorial Section
3. Sequence-to-Sequence Regression Using Deep Learning
4. Time Series Forecasting Using Deep Learning

Best regards

Stephan