Supply me by any documents, examples, and MATLAB codes.
read this book:
Artificial Intelligence in Power System Optimization
Weerakorn Ongsakul, Dieu Ngoc Vo, CRC Press, 2013
If you search through website, you may have matlab examples like matpower 3. If you cant find, please let me know.
Hi dear Prof. Okan, great thanks for your contributions.
I have the matpower examples, but did not find a neural network (NN) application there. Please help me in this regard. May be the one I have is different. As well, I would also like to know how to combine (NN) codes to matpower or generally power system analysis codes.
Thank you again ..
First of all, define your input parameters such as active power, frequency, voltage etc. and output parameter like bus voltage, active/reactive opower consumption. Then train your network and have many simulations using nntool in matlab.
First you have to collect some more data on inputs like Active power, Reactive power, Frequency, Voltage and so on. Because without enough data you can't train the matlab model of Artificial Neural Network.
After collecting data you can build an ANN using this simple code;
setdemorandstream (491218382)
net = patternnet(10);
view(net);
[net,tr] = train(net,Input,Target);
Use the input matrix as Input and desired target matrix as Target. If you have any problem on this codes, please check mathworks website. Then you can test your ANN and give any input and get the output. Collecting more data first is important to have a better result. Check the link bellow to have a better understanding.
http://www.mathworks.com/help/nnet/examples/crab-classification.html
Your points are very helpful dear Mr. Chathura Senanayake,
I will try that and get back to you.
Great thanks Sir.
Dear Mr. Chathura Senanayake,
How to arrange the input and output data is a real challenge. They are many even for a small power network. How to combine different types of data, different sizes of sub matrices in one input matrix and one output matrix.
Which arrangement should be followed?
I feel those are the most obstacles that face who want to deal with ANN for any application. Hope you shed a light on that, give an example, and so.
For instance, the input data should be considered for load flow analysis include active power generated at PV buses (one figure for each), voltage at generator buses (two figure at the reference bus, and one at other generator), load at PQ buses(two figures for each), compensation var at load buses (one figure for each), line data (three figures for each). While the output and target data will be active and reactive power generated at the reference bus (two figures), voltage angle and reactive power generated at generator buses (two figures for each), voltage at load buses (two figures for each).
Great thanks in advance.
Here I have attached a excel file which I used for Demand identification through ANN. The first block of data is the input matrix and the second block is target matrix. If you can arrange your data according to that format, it can easily read by Matlab. First create two simple matrices using less inputs and less outputs and later you can modify it for more inputs and more outputs.
- Badges
- Science topic