Method for predicting cycle time

1. A computer implemented predicting cycle time, comprising the following steps:

collecting a plurality of known sets of data, the known sets of data having at least two reference cycle times and an objective cycle time;

using a clustering method to classify the known sets of data into a plurality of clusters;

using a decision tree method to build a classification rule of the clusters;

building a prediction model of each cluster, the prediction model indicating the relationship of the variation between the reference cycle times and the objective cycle time;

preparing a predicted set of data, the predicted set of data having of at least two or more known reference cycle times and one unknown objective cycle time;

using the classification rule to determine that to which clusters the predicted set of data belongs; and

using the prediction model of the cluster to estimate the objective cycle time of the predicted set of data, acquiring a prediction value for the objective cycle time.

2. The method for predicting cycle time according to claim 1 , further comprising the following step:

using a fuzzy system to adjust the predicted value of the objective cycle time for the predicted set of data.

3. The method for predicting cycle time according to claim 2 , wherein the method of building the fuzzy system comprises the following steps:

respectively defining a standard value for the reference cycle times; and

defining a member function, which generates an adjusting value of an objective cycle time based on the difference value between the actual value and the standard of the reference cycle times.

4. The method for predicting cycle time according to claim 3 , wherein the use of the fuzzy system comprises the following steps:

calculating the difference values between the actual values and standard values of the reference cycle times of the predicted set of data;

inputting the difference values into the fuzzy system to obtain an adjusting value for the objective cycle time; and

adding the predicted value of the objective cycle time to the adjusting value.

5. The method for predicting cycle time according to claim 3 , wherein the member function is a triangular member function.

6. The method for predicting cycle time according to claim 3 , wherein the standard values for the reference cycle times are defined by experts.

7. The method for predicting cycle time according to claim 1 , further comprising the following steps:

obtaining an actual value of the objective cycle time for the predicted set of data;

comparing the difference between the predicted value and the actual value of the objective cycle time for the predicted set of data; and

in case that the difference between the predicted value and the actual value exceeds an acceptable error range, then recollecting the known sets of data.

8. The method for predicting cycle time according to claim 7 , further comprising the following step:

in case that the difference between the predicted value and the actual value is within the acceptable error range, then not recollecting the known sets of data.

9. The method for predicting cycle time according to claim 1 , further comprising the following steps:

defining an update interval;

determining whether the update interval has been reached; and

in case that the update interval has been reached, then recollecting the known sets of data.

10. The method for predicting cycle time according to claim 9 , further comprising the following step:

in case that the update interval has not been reached, then not recollecting the known sets of data.

11. The method for predicting cycle time according to claim 1 , wherein the clustering method is a K-means clustering method.

12. The method for predicting cycle time according to claim 1 , wherein the prediction model is a polynomial model.

13. The method for predicting cycle time according to claim 1 , wherein the prediction model is a neural network.

14. The method for predicting cycle time according to claim 13 , wherein the neural network is a Back Propagation Neural Network (BPNN).

15. The method for predicting cycle time according to claim 1 , wherein the total number of the clusters is smaller than the total number of the known sets of data.

16. A computer implemented predicting cycle time, comprising the following steps:

collecting a plurality of known sets of data, the known sets of data having at least two or more reference cycle times and one objective cycle time;

classifying the know sets of data into a plurality of clusters;

building a classification rule of the clusters;

building a prediction model of each cluster, the prediction model indicating the relationship of the variation between the reference cycle times and the objective cycle time;

preparing a predicted set of data, the predicted set of data having of at least two or more known reference cycle times and one unknown objective cycle time;

using the classification rule to determine that to which clusters the predicted set of data belongs; and

using the prediction model of the cluster to estimate the objective cycle time of the predicted set of data, acquiring a prediction value for the objective cycle time.

17. The method for predicting cycle time according to claim 16 , further comprising the following step:

adjusting the predicted value of the objective cycle time for the predicted set of data.

18. The method for predicting cycle time according to claim 17 , further comprising the following step:

obtaining an actual value of the objective cycle time for the predicted set of data;

comparing the difference between the predicted value and the actual value of the objective cycle time for the predicted set of data; and

in case that the difference between the predicted value and the actual value is exceedingly large, then recollecting the known sets of data.

19. The method for predicting cycle time according to claim 17 , further comprising the following steps:

defining an update interval;

determining whether the update interval has been reached; and

in case that the update interval has been reached, then recollecting the known sets of data.