How to predict one step ahead using Narxnet and need clarification about how they neural network is predicting

Hi, I am not from computer science background. I am new to neural networks.I trained my data using ID = 1:5 FD = 1:5 with input x(t) and targets y(t) my equation is y(t) = f(x(t-1), ... ,x(t-5),y(t-1), ... ,y(t-5)) ..

so is original input x(t) and target y(t) aren't they used for training? Is backpropagation means my x(t) last term is trained first ?

For prediction I want to give current input value and predict future output...

I am using the code below it has Id = 0:4 FD = 1:5 (fd I dont know how to change it)...now my equation looks like this y(t+1) = f(x(t), ... ,x(t-4),y(t-1), ... ,y(t-5))... I trained with different equation and predicting for different equation is this going to give me future output prediction for my current input.Am I predicting in the right or is there any mismatch ? Will my neural network work well?

Here is my prediction code nets = removedelay(net); nets.name = [net.name ' - Predict One Step Ahead']; view(nets) [xs,xis,ais] = preparets(nets,X,{},T); [netp,xip,aip] = closeloop(nets,xis,ais); [xs,xiss,aiss] = preparets(netp,X,{},T); [ys,xs,as] = netp(toPredict,xiss,aiss); view(netp)