Genetic Algorithms in Labview

I'd like to implement some diagnostic applications in Labview, using
genetic algorithms and fuzzy logic. Anyone can give me some guidelines
to do this?
Best Regards,
Sérgio Cruz

wrote:
> I'd like to implement some diagnostic applications in Labview, using
> genetic algorithms and fuzzy logic. Anyone can give me some guidelines
> to do this?
>
I have implemented a genetic algorithm based control of a mobile robot
using Mathematica Link for LabVIEW http://www.e-notebooks.com/products/
link/. LabVIEW is in charge of communication with the robot, Mathematica
was more suitable to implement the GA. I can provide with more details if
you wish.
Jean.
e-NoteBooks Ltd
http://www.e-notebooks.com
http://www.sstreams.com
Sent via Deja.com http://www.deja.com/
Before you buy.

Similar Messages

Interfacing a C++ Console Application Genetic Algorithm with LabVIEW

Hi,
I am have recently modified the GENESIS genetic algorithm using C++ so that it runs on a Windows machine (before it only ran under UNIX). In my experimental work the aim is to control everything using LabVIEW, including this Genetic algorithm.
My problem comes from the fact that the Genetic Algorithm is set-up as a console application, a library file is created containing the input parameters of the VI, then this is joined with an evaluation function that is dependant on the library file an an executable is created.
I need to take some information collected from LabVIEW and give it to the genetic algorithm, but I don't know how to do this.
Any Ideas?
Thanks
Alan Homer

You can run most applications from a command line using the location/application.exe (where the location is where the executable is stored), all c++ applications run using: int main (int argc, char **argv); where the argc is the number of arguments and argv is essentially a string table with the parameters passed to the program (will be used as code) the output if a command line application is the std out which can be piped to file or wired out from the standard out terminal. To pass parameters from your LV code to the command line you should know the parameter list (sometimes empty).
Paul
Paul Falkenstein
Coleman Technologies Inc.
CLA, CPI, AIA-Vision
Labview 4.0- 2013, RT, Vision, FPGA

Genetic algorithm for use within LabVIEW

I want to use a genetic algorithm to optimise a control system. I was going to use MatLab because I can write the GA with that, but I cannot get it running fast enough with my C-DAQ. The alternative is to use LabView, but I could not write a GA VI. Do you know of a GA for use in LabVIEW?

Hi Jack@Newcastle,
Take a look at the following example of Genetic Algorithm implemented in LabVIEW (https://decibel.ni.com/content/docs/DOC-13699).It should give you an idea how to write your own code.
Regards,
Edvin

Dijkstra's Algorithm in LabVIEW

So I've attempted to undergo the painful task of trying to implement dijkstra's algorithm in LabVIEW. The problem is that I think I am going about doing this wrong, or at least not in the most efficient manner. I've attached my VI so people can take a look at it. I guess I'll also give a brief description of how I'm attempting to implement this. So the input from the user ( at this point) is three 1D arrays as well as a start and end vertex value. The three arrays are the weight values of the edges, V1, and V2. The way these arrays are filled out is so that in each array the coordinating values are all stored at the same index. So if there is a path from vertex 0 to 1, with a weight of 2 then 2 will be stored in the edge array, 0 in the V1 array, and 1 in the V2 array, and all at the same index. These values get fed into functions and the shortest path is found, yada yada yada. I'd explain it more, but I think it'll be easier if you look at the VIs. The problem is that the way I have it setup, creating a finished product to sort through even 20 vertexes would result in massive code. The VI i have is only about 25% complete I'd say, but i dont want to go any further as it is getting exponentially more complicated with each iteration and I feel there has to be a better way. So if anyone has any ideas I would greatly appreciate it!
P.S. I realize that i can write the code in C and call it as a function in the LabVIEW program, but I'm kind of stubborn and want to make this work written solely in LabVIEW
Attachments:
Dijkstras.vi ‏23 KB
Short path.vi ‏21 KB
Short path 2.vi ‏24 KB

Well I'm glad that there are others interested in this project. First I'd just like to give a little update. The program currently will give the length of the shortest path to the desired vertex, but it doesn't output what the path it took to get that shortest value yet. So that is what I am working on now, as well as verifying its scalability and trying to implement other features that will actually make this a useful program vs. mere curiosity.
As far as my references go, I've gotten all my info from the internet. Wikipedia, of course, was my starting place. I found a well explained article at http://eoinbailey.com/blog/dijkstras-algorithm-illustrated-explanation that helped me decide how I was going to try and implement the program. I also read pretty much the rest of the links that come up on google when you type in "dijkstras algorithm." Again, I have no formal CS background, Im studying engineering, so I'm working with a lot of trial and error here. My basic methodology behind all this was simply to read as much as I could, try to understand it, and then work on the program until I started getting somewhere. Maybe not the best strategy, but it works for me. Thank you also for the link, I wish I had read that before I had started this project lol
I will post my latest version of the VI later today, This way I can get some feedback on how I can improve it and what works or doesn't work. Thanks again everyone for all ur help!

Genetic Algorithm help please!!!!

Hi people, i really need help badly with a project im doing as im not very good at programming yet,
The project is as follows
Project Allocation
This project will develop a Genetic Algorithm that will evolve project allocations, with the aim of maximising several criteria: giving every student the highest ranked project possible, spreading the project load between lecturers etc. It should work for a single group of students, and for several groups. This project could be coded in any suitable language.
Any help in coding this would be really welcomed!
Thanks

kennalan wrote:
your right warneja but what can i do, Start by learning English: that should be "you're", as in "you are".
What can you do? Do what all students are expected to do - work hard, struggle with difficult material until you achieve some measure of mastery, and accept the result.
i just need some source code Our source code? Something that you'll steal and pass off as your own? In my universe, that's called "cheating". What do you call it?
or anything to help!Here's some advice: Don't come back here unless you have problems with code that you've written, with specific questions about the problems.
Google for something and get started. I recommend "Java genetic algorithms" as a search criterion.
%

Levy flight or firefly algorithm in Labview or matlab

Hello,
Has anyone used Levy flight algorithm with Labview. Is there any toolkit or library to implement this algorithm in Labview or at least a Matlab code that can be coupled with Labview via Mathscript??
Thanks,
Zied

If you search the Web for "Levy Flight", you will find Matlab code.
Bob Schor

Perturbation & Observation algorithm using labview

Hi
i have a problem with the mppt (maximum power point tracking) algorithm in labview, if you know how to do it help me please

السلام عليكم ورحمة الله وبركاته
الرجاء طرح الأسئلة بشكل واضح و كامل وذلك للدفع بهذا المنتدى العربي إلى الأمام وبذلك تعم المنفعة للجميع.
للإجابة على شيئ من السؤال التالي قد يفيدك
Using NI CompactRIO to Design a Maximum Power Point Tracking Controller for Solar Energy Application...

Calling all Genetic Algorithm Freaks : Postage stamp problem

This all started from the other thread " A Silly Trick question (Asked by a mathematican) for a Dummy like me" which grows over 6 pages long and 80+ post.
Here is to restate the problem in Marlin words
the old Frobenius Postage Stamp problem
http://mathworld.wolfram.com/PostageStampProblem.html
Why didn't ya say so. This is where you should have started this thread.
"I have a mathematician who is trying to solve the Frobenius postage stamp problem for the subcase of two stamps. In the language of the Wolfram article this would be n sub 2 (A sub N) meaning the smallest integer amount that can't be repersented by the sum of two stamps that are choosen from a fixed set of N possible values that are presumably provided by the post office."
"He would like to run the search for solutions on the order of N > 25 and wants to divide that process up between multiple machines. Any ideas on how to do this?"
Had you started there, you would have found lots of different answers. For one thing it is fairly easy to write code that will test what is the smallest sum that can be obtained from a given set. This means that all the Genetic Algorithm freaks out there will say, "Hey I can code of a GA to find optimal sets" and they will start right in on the hardest sets up in the 90's cause they want to show off the power of GAs.
Meanwhile the simulated annealing folks will target the numbers in the 50s because they are not quite so cocky as the GA crowd but they still want to show off.
And finally us low brow mathematician sorts will start with some grubby backtracking exhaustive search algorithm and see if we can't get it to replicate the result for N = 4.
========================================================================
Eagerly waiting to witness the power of Genetic Algorithm

Try the algoritms forum too. People interested in these sorts of things come there and it's much slower so your question will linger longer.

Iterative signal processing algorithms in Labview

Hi
I am learning to program in Labview. I am trying to program an iterative algorithm in Labview and I am having difficultires. It is a short algorithm and thought that perhaps someone with Labview experience may be able to show me the way. Here I describe the algorithm:
I have data in an I by K matrix A where i are the samples (rows) and k the variables (columns). For n number of runs, n = 1,2,...N need to compute u(subscript n) and r(subscript n) from A(subscript n-1). (Note that u is vector of length i and r is a vector of length k). Results are obatined for each run, n, iteratively by improving on estimates of u and r.
First I need to select start values for u(subscript n) in the first iteration of the algorithm. These start values can be any one of the columns of A. Then I want to repeat the following steps (1 to 5) until convergence:
(Note that ' represents transpositionof vector /matrix; ^ represents power)
1. Improve estimate of r(subscript n) by:
r(subscript n) = [u(subscript n)' u(subscript n)]^-1 u(subscript n)' A(subscript n-1)
2. Scale length of r(subscript n) to 1.0 by:
r(subscript n) = r(subscript n) [r(subscript n)' r(subscript n)]^0.5
3. Improve estimate of u(subscript n) for this run by:
u(subscript n) = A(subscript n-1) r(subscript n) [r(subscript n)' r(subscript n)]^-1
4. Imprve estimate of tee(subscript n) by:
tee(subscript n) = [u(subscript n)' u(subscript n)]
5. check for convergence: if tee(subscript n) minus tee(subscript n) in the previous iteration is smaller than say e.g. 0.00001 times tee(subscript n), the method has converged for this run, if not, go to step 1.
The Substract the effect of this run: A(subscript n) = A(subscript n -1) - u(subscript n) r(subscript n)' and then go to the start for the next run.
I have tried to do this with shift registers but the problems I am having aare to do with the start values for u and the iterative updating...
I hope someone can help. I know that MAth script is probably the best way to do this but I don't yet have it and would like to do it in the block diagram with Labview vi for math.
Cheers,

for anyone trying to help me...i worked it out. Perseverance is certainly good!

Fitness for Shortest path using a genetic algorithm

Hi.
I have this problem:
I'm doing a program that has a map and two points. the map has wall that are not valid positions.
I have used a A* algorithm to calculate the shortest path and is working fine.
now i'm implementing a genetic algorithm to solve the problem but i'm not having sucess.
My genetic operator are directions (left, right, up and down).
I'm not being able to find a goodfitness function to solve the problem.
can someone tell me a god function or here I can found information about that?
I have been searching on google and I have not yet had sucess.
I have tryed two diferent fitness tecniques:
- add 1 for each step and if it finds a wall or gets out of the map add the maximum value (mapsize*mapsize).
- the other one is add the valid steps and if it finds a wall or gets out of the map add the number of necessary steps to gets to destination. the valid steps has a weight of 30% and the necessary steps has a weight of 70% (i have tryed to use diferent weight).
Can someone help me?
thanks

How about adapting your TSP code as follows. Your graph G(V, E) has two special vertices: v_s and v_d (start and destination). Let a candidate solution consist of a permutation of V\{v_s, v_d}u{v_m} where v_m is a new marker vertex whose purpose is to allow you to treat part of the candidate solution as irrelevant. Given candidate solution v_0 v_1 v_2 ... v_n v_m v_n+1 ... the weight is c(v_s, v_0) + c(v_0, v_1) + ... + c(v_n-1, v_n) where c(v_a, v_b) is the cost of the edge ab.

I want to know memory manament and data display algorithm in LabViEW.Where can I get some materials about them?

I want to know memory manament and data display algorithm in LabViEW.Where can I get some materials about them? I need them to complete my essay .Thank you for your help.

Memory menagement:
http://zone.ni.com/devzone/conceptd.nsf/2d17d611efb58b22862567a9006ffe76/732cec772aa4fbe586256a37005541d3?OpenDocument
About data display: what type of data you wat display?

Genetic Algorithm

Hello everyone,
am a little stuck, heck alot stuck. Need to create a genetic algorithm in Java and need help fast! Where do i even start?

Perhaps you should start with trying to implement a popular genetic (evolutionary) genetic algorithm like SPEA2, NSGA-II, or PAES. SPEA2 takes the general form below:
1. Generate an initial population (P) and an empty archive set (A).
2. Calculate the fitness of individuals in P and A.
3. Add all non-dominated individuals in P and A to A + 1. If A + 1 is smaller than the N, then add more members; if larger than N, then reduce (based on density info.).
4. If the number of iterations you want to complete is met (or some other condition met) then stop, else continue.
5. Fill a mating pool using binary tournament selection with replacement using A + 1.
6. Apply recombination and mutation to the mating pool to generate P + 1.
You can find more information at:
http://www.tik.ee.ethz.ch/~zitzler/
http://www.lania.mx/~ccoello/EMOO/
http://www.cs.cf.ac.uk/user/L.Raisanen/links.html (look under algorithm links)
While this looks daunting, let me take you through what is happening:
Let's create the genetic material of an individual (of integers). For example, an individual may be represented by 5 number: 5, 3, 2, 1, 4.
We then "decode" this individual (depending on what we are measuring)and say he is 41 out of 60 on some trait we desire.
Let's say there are 19 other people in this same population, and that this one falls somewhere in between on his "score" out of 60.
We may choose to take the best 10 individuals (for arguments sake) from this population and recombine them to form a new population of 20. Each person, in this instance, has two children. Let's say our man is somewhere in this "chosen" population.
After the genetic material from our man, and some other person in the population is recombined, let's say his childrens genetic material now looks as follows: 3, 4, 2, 5, 4 & 4, 1, 3, 2, 2
Let's say they now scores 46 and 32 out of 60 on some trait. So, the one child improved on the parent, and the other was worse.
Well, we continue to take the best 10 and recombine them (and throw in a bit of mutation as well) until we meet some condition. We then stop with a final population.
As you can see, the code for this can become quite long... and time-consuming to build. You will need some of the following methods:
public void makePopulation( sizeOfPop, numberOfIntegers )
public void mutatePopulation
public void recombinePopulation
public void decodePopulation
..etc., etc.
These may vary quite substantially depending on what you are trying to accomplish. Obviously, this is a very large subject area, but hopefully this gives a flavour of what is happening. If you want to see some bit of code in particular, write back here and I will be happy to help.

Help With Genetic Algorithms

My source code compile well with to .Classs file but it shows some error when i wanted to execute the .Class file the error :
EXCEPTION in he thread main java.lang.NoSuchMethod error main
I'm i leaving out some method please take a look at the code
import java.util.Random;
public class GeneticTools
private static Random myRandom = new Random(10);
public static void determineFitness(Candidate myCandidate)
       double fitnessResult;
       fitnessResult = 0.0; //initial value to overwrite
      myCandidate.setFitness(fitnessResult);
//selection
public static void selection(Candidate populationList[])
      Candidate newPopulationList[];
       newPopulationList = new Candidate[populationList.length];
       //sum up all the fitnesses
       int percentages[];
       percentages = new int[populationList.length];
       double sum;
       sum = 0.0;
       for(int i=0;i<populationList.length;i++)
          sum += populationList.getFitness();
     //determined percentages for next recombination
     for(int i=0;i<populationList.length;i++)
//how many of each to put back into the population
          percentages[i] =(int) Math.ceil(((populationList[i].getFitness() / sum)
               * populationList.length));
     int countSum;
     countSum=0;
     for(int i=0;i<percentages.length;i++)
countSum += percentages[i];
     if ( (populationList.length - countSum) > 0) //not creating enough
     //just add it to the strongest candidate to date
percentages[0] += (populationList.length - countSum);
     if ((populationList.length - countSum) < 0) //toomany
for(int i=0;i<countSum-populationList.length;i++)
if (percentages[i]>0)
               percentages[i]--;
     countSum=0;
     for(int i=0;i<percentages.length;i++)
countSum += percentages[i];
     //create new PopulationList for recombination
     //for each percentage>0 insert that candidate that many time into the new list
int newListIndex = 0;
for(int i=0; i<percentages.length;i++)
for(int j=0;j<percentages[i];j++)
newPopulationList[newListIndex] = populationList[i];
               newListIndex++;
//populationList = newPopulationList; //wont work, pass by value
//insert new population back into the old population list
     for(int i=0;i<populationList.length;i++)
populationList[i] = newPopulationList[i];
     } //end of selection
public static void recombination(Candidate populationList[], int mutationRate)
     Candidate newOffspringList[] = new Candidate[populationList.length+1];
     Candidate offspring;
int parentA;
     int parentB;
     for(int i=0;i<populationList.length;i++)
//randomly pick two unique from top halfto "mate"
          } //end of for loop to create new kiddies
     //see if the "best parent is more fit than the kids
insertCandidate(newOffspringList,populationList[0],populationList.length);
     //create the same number of offspring as parents putting them back into the populationList
     for(int i=0;i<populationList.length;i++)
populationList[i] = newOffspringList[i];
     } //end of recombination
//recombination
public static void mutation(Candidate myCandidate, int mutationThreshold)
     int currentRandomMutation;
     int randomPosition;
     currentRandomMutation = myRandom.nextInt(100);
     } //end of mutation
public static Candidate crossover(Candidate candidateA, Candidate candidateB)
Candidate newCandidate;
     newCandidate = new Candidate();
     int crossoverPoint;
return(newCandidate);
     } //end of crossover
// insert the candidates into the list sorted by fitness
public static void insertCandidate(Candidate populationList[],
     Candidate currentCandidate, int size)
     } //end of insertCandidate
public static String displayCurrentPopulation(Candidate populationList[])
String retval;
     retval = "\n";
     return(retval);
} //end of GeneticTools

YES sorry i did not mention the candidate class i've create inte first play
this is what the candidate class
import java.util.Random;
public class Candidate
private int x;
private int y;
private int chromosome[];
private double fitness;
public final static int MAXGENES = 64;
private static Random myRandom = new Random(10);
public Candidate()
      x=myRandom.nextInt(100);
       y=myRandom.nextInt(100);
       chromosome = new int[MAXGENES];
       GeneticTools.determineFitness(this);
       makeChromosome();
     } //end of default constructor
public Candidate(int myX, int myY)
      x=myX;
       y=myY;
       chromosome = new int[MAXGENES];
       GeneticTools.determineFitness(this);
       makeChromosome();
     } //end of constructor
public void makeChromosome()
      String tempA;
       String tempB;
       String combined;
       tempA = Integer.toBinaryString(x);
       //now pad the leading zeros
       for(int i = tempA.length(); i < MAXGENES/2;i++)
          tempA = "0"+tempA;
       tempB = Integer.toBinaryString(y);
       for(int i = tempB.length(); i < MAXGENES/2;i++)
          tempB = "0"+tempB;
       String combinedResult = tempA +"-"+ tempB;
       combined = tempA +tempB;
      for(int i=0;i<combined.length();i++)
          chromosome[i] = combined.charAt(i)-'0';
     } //end of makeChromosome
public void displayCoordinates()
      String resultStringA,resultStringB,resultString;
       resultStringA = "";
       resultStringB = "";
       resultString = "";
       int resultA;
       int resultB;
       for(int i=0;i<chromosome.length/2;i++)
          resultStringA += chromosome;
     resultA = Integer.parseInt(resultStringA,2);
     for(int i=chromosome.length/2;i<chromosome.length;i++)
resultStringB += chromosome[i];
     resultB = Integer.parseInt(resultStringB,2);
     resultString += "("+resultA+","+resultB+")";
     System.out.println("coordinates = "+resultString);
     x = resultA;
     y = resultB;
     } //end of displayCoordinates
public void flipBit(int randomPosition)
if(randomPosition != 0 && randomPosition != MAXGENES/2) //dont make coordinate neg
          if (chromosome[randomPosition] == 1)
               chromosome[randomPosition] = 0;
          else
               chromosome[randomPosition] = 1;
     updateCoordinates(); //use the chromosome array to update x,y
     } //end of flipBit
//this method will analyze the new chromosome and reset the x and y
public void updateCoordinates()
String resultStringA,resultStringB,resultString;
     resultStringA = "";
     resultStringB = "";
     resultString = "";
     int resultA;
     int resultB;
     for(int i=0;i<chromosome.length/2;i++)
resultStringA += chromosome[i];
     resultA = Integer.parseInt(resultStringA,2);
     for(int i=chromosome.length/2;i<chromosome.length;i++)
resultStringB += chromosome[i];
     resultB = Integer.parseInt(resultStringB,2);
     resultString += "("+resultA+","+resultB+")";
     // System.out.println("coordinates = "+resultString);
     x = resultA;
     y = resultB;
     } //end of updateCoordinates
public void setX(int myX)
x = myX;
public void setY(int myY)
y = myY;
public int getX()
return(x);
public int getY()
return(y);
public int[] getChromosome()
return(chromosome);
public void setFitness(double myFitness)
fitness = myFitness;
public double getFitness()
return(fitness);
public String displayCandidateInfo()
String retval;
retval="";
     retval += "("+x+","+y+") "+ "fitness = "+fitness;
     return(retval);
} //end of class definition

Looking for VCDIFF (RFC 3284) algorithm implemented in LabVIEW

Hello forum,
I am looking to see if anyone knows of a publicly available implementation of the VCDIFF binary differencing algorithm in native LabVIEW (G) ?
Here is a link to a paper on the RFC 3284 standard for those interested: http://tools.ietf.org/html/rfc3284
There are a number of open source implementations in C/C++, but I would like to locate one in LabVIEW if possible.
Regards!

Implementing binary algorithmes in LabVIEW and especially compression algorithmes, while possibe is not a very smart idea. Aside from possible perfomance problems when you start to shuffle around on bit level, there is also the issue about how to proof the correctness of your implementation. Those open source C/C++ solutions have been used, peer reviewed and what else many times and all that has uncovered various flaws and bugs during the process.There is only a very small adience that will go and use your LabVIEW algorithme and even a much smaller one who is able to do any significant form of peer review.
So all in all you are likely to end up with a somewhat to almost working implementation that nobody is going to use, for the gain of bragging that you did it in LabVIEW.
Rolf Kalbermatter
CIT Engineering Netherlands
a division of Test & Measurement Solutions

FFT algorithm routines in Labview

Hello,
Real FFT.vi uses the fast radix-2 FFT routines or a mixed radix cooley-tukey algorithm in Labview.
Do these kinds of routines have the 2N-point real FFT method?
2N-point real FFT method is to utilize the single complex FFT by processing two seperated N points (even index data- real N-points input, odd index data-imaginary N-points input in single complex FFT routine)
And what are differences between radix-2 FFT and fast radix-2 FFT routines?
And what are differences between cooley-tukey algorithm and mixed radix cooley-tukey algorithm ?

fefelove wrote:
2N-point real FFT method is to utilize the single complex FFT by processing two seperated N points (even index data- real N-points input, odd index data-imaginary N-points input in single complex FFT routine)
If I understand you right, you could just typecast your interlaced array into a complex array of half the lenght before feeding it to the complex FFT.
Can you explain your concerns about the algorithms. Maybe wikipedia will help?
LabVIEW Champion . Do more with less code and in less time .

Genetic Algorithms in Labview

Similar Messages

Maybe you are looking for