% Tutorial: generat cmprhnsv_Data using function gen_cmprhnsv_Data.p

>> sysc=[tf(1,[1 1]), tf(2,[10 1]);tf(1,[5 1],'inputdelay',2), tf(3,[8 1],'inputdelay',1)];  % generate continuous TF
>> sysd=c2d(sysc,1); % try discrete TF
>> sysK=dcgain(sysc); % try gain matrix. Note LMIPA can take continuous model, discrete mode, or simply gain matrix
>> y=randn(100,2); % generate output data randomly
>> u=randn(100,2); % generate input data randomly (ignore actual input-output relation for illustration purpose)
>> LL=-10*ones(4,1); % low constraint limit
>> UU=10*ones(4,1); % high constrain limit
>> Lc=randn(4,1); % linear coefficient for the objective function
>> Qc=randn(4,1); % Quadratic coefficient for the objective function
>> Tc=zeros(4,1); % Target coefficient for the objective function (see manual for explanation)
>> gen_cmprhnsv_Data  % generate data and model for LMIPA
 
Consider a process with m inputs and p outputs
 
1. Plant model (continuous/discrete): p x m tf matrix or static gain matrix
1. Output data (y): N x p matrix
2. Input data (u): N x m matrix
3. Sampling time
4. Low limit data: (p + m) x 1 vector
5. High limit data: (p + m) x 1 vector
6. Linear coefficients: (p + m) x 1 vector
7. Quadratic coefficients: (p + m) x 1 vector
8. Target coefficients: (p + m) x 1 vector
 
Continuous/discrete transfer function matrix: sysc  % try continuous TF
Output data matrix y
Input data matrix u
Sampling time: 1
Low limit data vector: LL
High limit data vector: UU
Linear coefficients data vector: Lc
Quadratic coefficients data vector: Qc
Target coefficients data vector: Tc
The file name to save comprehensive data (e.g. test_cmprhnsv_data): test_cmprhnsv_data_con
The file name to save plant model (e.g. test_model): test_model_con
>> % Data/model generation has completed. Let's try to generate alternative data/model using discrete TF
>> gen_cmprhnsv_Data
 
Consider a process with m inputs and p outputs
 
1. Plant model (continuous/discrete): p x m tf matrix or static gain matrix
1. Output data (y): N x p matrix
2. Input data (u): N x m matrix
3. Sampling time
4. Low limit data: (p + m) x 1 vector
5. High limit data: (p + m) x 1 vector
6. Linear coefficients: (p + m) x 1 vector
7. Quadratic coefficients: (p + m) x 1 vector
8. Target coefficients: (p + m) x 1 vector
 
Continuous/discrete transfer function matrix: sysd
Output data matrix y
Input data matrix u
Sampling time: 1
Low limit data vector: LL
High limit data vector: UU
Linear coefficients data vector: Lc
Quadratic coefficients data vector: Qc
Target coefficients data vector: Tc
The file name to save comprehensive data (e.g. test_cmprhnsv_data): test_cmprhnsv_data_dis
The file name to save plant model (e.g. test_model): test_model_dis
>> % Data/model generation using discrete TF has completed. Let's try to use gain directly next
>> gen_cmprhnsv_Data
 
Consider a process with m inputs and p outputs
 
1. Plant model (continuous/discrete): p x m tf matrix or static gain matrix
1. Output data (y): N x p matrix
2. Input data (u): N x m matrix
3. Sampling time
4. Low limit data: (p + m) x 1 vector
5. High limit data: (p + m) x 1 vector
6. Linear coefficients: (p + m) x 1 vector
7. Quadratic coefficients: (p + m) x 1 vector
8. Target coefficients: (p + m) x 1 vector
 
Continuous/discrete transfer function matrix: sysK
Output data matrix y
Input data matrix u
Sampling time: 1
Low limit data vector: LL
High limit data vector: UU
Linear coefficients data vector: Lc
Quadratic coefficients data vector: Qc
Target coefficients data vector: Tc
The file name to save comprehensive data (e.g. test_cmprhnsv_data): test_cmprhnsv_data_K
The file name to save plant model (e.g. test_model): test_model_K
>> % We have generate 3 sets of data/models. They are equivalent. We can use any one for LMIPA and the results should be same
>> % Next we shall start main_lmipa in the directory contaving LMIPA and then input any of the three generated data/models for economic performance assessment