基于PCA的TE过程故障检测

PCA作为最为基础的降维方法，已得到广泛应用，这里使用MATLAB代码实现对TE过程数据的降维处理，旨在用较少的主元成分解释尽可能多是原数据信息，并通过计算T2/SPE统计量来进行故障检测。

数据

链接：https://pan.baidu.com/s/1yHgZRqy6yyk9wsm6igqUSw 
提取码：odsz

主程序

clear all;
close all;
clc;
%%%%%%%%%%%%%%%%%样本数据导入%%%%%%%%%%%%%%%%%%%%

X_Y_train = matfile('datatexy.mat'); 
X_train = X_Y_train.X;
load('d01_te.mat')
X_test =  d01_te(:,[1:22,42:45,47:49,51:52]);
%%%%%%%%%%%%%%%%%样本数据标准化%%%%%%%%%%%%%%%%%%%%

X_train_zscore = zscore(X_train);

X_mean = mean(X_train);X_std = std(X_train);
[X_train_zscore_row,X_train_zscore_col] = size(X_train_zscore);

X_test=(X_test-repmat(X_mean,X_train_zscore_row,1))./repmat(X_std,X_train_zscore_row,1);
%X_test_zscore = zscore(X_test);

%%%%%%%%%%%%%%%%%训练集数据size%%%%%%%%%%%%%%%%%%%%

[X_train_zscore_row,X_train_zscore_col] = size(X_train_zscore);

%%%%%%%%%%%%%%%%%测试集数据size%%%%%%%%%%