doing project, for the first time to write blog
Step 1: particle coding
To solve the tasks scheduling problem, the first to build effective particle coding scheme, using indirect encoding, a coding length is the number of subtasks, and a child task corresponds to a resource nodes, such as 10 subtasks, 5 compute nodes, is the location of the individual particles I coding can be expressed as a 10 d vector xi={5,4,3,2,4,2,3,1,3,4}, the first bit code is 5 said task t, on the compute nodes (p5, second encoding said t2 on computing nodes p4 task execution,


Step 2: cloud computing task scheduling under represented
The third step, the algorithm flow



Step1: initialization parameters, set the maximum minimum parameters such as inertia weight value, set the number of iterations for the maximum number of iterations,

Step2: according to the last set of particle encoding, random initialization of the particle's position and velocity; To initialize the particle's velocity vi=xi,

Step 3: calculate the fitness of each particle,
Step4: the number of iterations t=1, the implementation of scheduling,

Step5: x will particles; Its fitness value and according to the experience, and experience the best location of pbest of fitness values, if more, then x; Record the best location for the particles experienced pbest,
Step6: x will particles; Its fitness value and according to the experience, and experience the best location of the fitness value gbest comparison, if better, then x records for the gbest particle has experienced the best position,

Step7: update the particle's velocity and position, the range of particle coding sequence of update may not comply with this code, so the absolute value more than integer for mapping method is adopted to update again, the location of the updated

Absolute value integer mapping principle is: when the distribution of the compute nodes p for [-m, m] decimal, for its absolute value after, take up the whole; When the range of (m, up) or (- up - m), is the absolute value after, to _ take on the whole, in addition to more than m o; Range of 0, then define value is 1, according to this way to update the position and velocity,

Step8: update particle inertia weight,

Step9: if the current number of iterations is less than the largest number of iterations, skip to step 3 to continue. Did not meet. To print gbest algorithm run over,

Reference: Jiang Yan paper

Did you do this algorithm, discuss

CodePudding user response:

Bloggers, did you get rid of this algorithm

CodePudding user response:

The

refer to the original poster qq_38001584 response:

doing project, for the first time to write blog
Step 1: particle coding
To solve the tasks scheduling problem, the first to build effective particle coding scheme, using indirect encoding, a coding length is the number of subtasks, and a child task corresponds to a resource nodes, such as 10 subtasks, 5 compute nodes, is the location of the individual particles I coding can be expressed as a 10 d vector xi={5,4,3,2,4,2,3,1,3,4}, the first bit code is 5 said task t, on the compute nodes (p5, second encoding said t2 on computing nodes p4 task execution,


Step 2: cloud computing task scheduling under represented
The third step, the algorithm flow



Step1: initialization parameters, set the maximum minimum parameters such as inertia weight value, set the number of iterations for the maximum number of iterations,

Step2: according to the last set of particle encoding, random initialization of the particle's position and velocity; To initialize the particle's velocity vi=xi,

Step 3: calculate the fitness of each particle,
Step4: the number of iterations t=1, the implementation of scheduling,

Step5: x will particles; Its fitness value and according to the experience, and experience the best location of pbest of fitness values, if more, then x; Record the best location for the particles experienced pbest,
Step6: x will particles; Its fitness value and according to the experience, and experience the best location of the fitness value gbest comparison, if better, then x records for the gbest particle has experienced the best position,

Step7: update the particle's velocity and position, the range of particle coding sequence of update may not comply with this code, so the absolute value more than integer for mapping method is adopted to update again, the location of the updated

Absolute value integer mapping principle is: when the distribution of the compute nodes p for [-m, m] decimal, for its absolute value after, take up the whole; When the range of (m, up) or (- up - m), is the absolute value after, to _ take on the whole, in addition to more than m o; Range of 0, then define value is 1, according to this way to update the position and velocity,

Step8: update particle inertia weight,

Step9: if the current number of iterations is less than the largest number of iterations, skip to step 3 to continue. Did not meet. To print gbest algorithm run over,

Reference: Jiang Yan paper

The original poster can be added to your contacts?

CodePudding user response:

Is there a code can be to look at

CodePudding user response:

references a 4th floor BUG# response:

do you have any code to see if we can expect

Please find the code

CodePudding user response:

reference 5 floor h_17 reply:

Quote: a reference 4 floor BUG# response:

Is there any code to see if we can expect

Please find the code

hey, do you have

CodePudding user response:

refer to 6th floor a BUG# reply:

Quote: refer to the fifth floor h_17 reply:

Quote: a reference 4 floor BUG# response:

Is there any code to see if we can expect

Please find the code

hey, do you have any

I don't have a