Parameter Estimation of non-linear system

Hi!! Iam trying to estimate the parametersorder and coefficientsof non-linear system using neural network.How could I feed the input -output dataset,coefficient and order in training phase of network?Thank you in advance.

[Greg Heath]

Show what you have done so far

Greg

[Rahavi Selvarajan]

We had developes a nonlinear model and want to estimate the parameters using deep learning. During the training phase we don't know how to feed the coefficients and order of the system.

clear all;

close all;

clc;

%Order of the nonlinear system

order = 20;

disp(order);

%Generating coefficients of the polynomial

for i= 1:order+1

coeff(i) = 0.0001 * randn();

end

%Impulse function

h = rand(10,1);

for i = 1:10000

%Generating input

sigma = 5;

X = sigma*randn(100,1);

in(:,i) = X;

%Output of a nonlinear System

z =0;

for k = 1:order+1

sum = coeff(k) * X.^(k-1);

z = z+sum;

end

%Output of a linear system

y = conv(z,h);

out(:,i) = y;

end

x = in;

t = out;

% Choose a Training Function

trainFcn = 'trainbr';

% Create a Fitting Network

hiddenLayerSize = 10;

net = feedforwardnet(hiddenLayerSize);

net = init(net);

%Create a neuralnetwork estimator

net_estimator = neuralnet(net);

% Setup Division of Data for Training, Validation, Testing

net.divideParam.trainRatio = 70/100;

net.divideParam.valRatio = 15/100;

net.divideParam.testRatio = 15/100;

% Train the Network

net.trainParam.epochs = 1;

[net,tr] = train(net,x,t);

% Test the Network

y = net(x);

e = gsubtract(t,y);

performance = perform(net,t,y)

% View the Network

view(net);

Answer 1

[Edoardo Catenaro]

Hi! I'm also stucked on a similar problem. Did you find the solution to your problem ?

Thanks in advance