%
读取数据
data=xlsread(
'data.xls'
);
%
训练预测数据
data_train=data(1:113,:);
data_test=data(118:123,:);
input_train=data_train(:,1:9)';
output_train=data_train(:,10)';
input_test=data_test(:,1:9)';
output_test=data_test(:,10)';
%
数据归一化
[inputn,mininput,maxinput,outputn,minoutput,maxoutput]=premnmx(input_tr
ain,output_train);
%
对
p
和
t
进行字标准化预处理
net=newff(minmax(inputn),[10,1],{
'tansig'
,
'purelin'
},
'trainlm'
);
net.trainParam.epochs=100;
net.trainParam.lr=0.1;
net.trainParam.goal=0.00001;
%net.trainParam.show=NaN
%
网络训练
net=train(net,inputn,outputn);
%
数据归一化
inputn_test = tramnmx(input_test,mininput,maxinput);
an=sim(net,inputn);
test_simu=postmnmx(an,minoutput,maxoutput);
error=test_simu-output_train;
plot(error)
k=error./output_train