NARX Neuronal Network with feedback target AND non-feedback target?

Hello guys, I'm trying to train a NARX Network by seriell-parallel training to simulate later in parallel mode.

My problem is that I want to have 3 Inputs and 2 Outputs, so that the 3 Inputs are not delayed, while 1 Output is delayed and recurrent. The other Output is just a normal Output.

In my understanding I would define the recurrent Output as feedback-target and the standard Output as a non feedback target. Here is my code:

%%delay
dly1 = 0;
dly2 = 1;
n1= 10;
u = [Input2.VEH_v';Input2.Gradient';Input2.a']; 
y1 = [Input2.ENG_fr'];
y2 = [Input2.SOC'];
u1 = con2seq(u);
y3 = con2seq(y1); 
y4 = con2seq(y2);

narx_net = narxnet(dly1,dly2,n1);
narx_net.divideFcn = '';
narx_net.trainParam.min_grad = 1e-10;
[p,Pi,Ai,t] = preparets(narx_net,u1,y3,y4);

narx_net = train(narx_net,p,t,Pi);

I always get the error: "Error using preparets (line 175) Number of output signals does not match number of network's non-feedback outputs."

Can you please help me solve my problem?

[Kylie Meier]

Hello, I am having a similar issue and would be curious to know the answer to your question. I am wondering if a closed loop or open loop has any affect on this as well.