Paper Titles

A Study on the Effect of Process Parameters on Surface Topography of Al Thin Films on Various Substrates Using AFM
p.903

Preparation of Geopolymer Using the Slag from Direct Reduction-Magnetic Separation of Refractory Iron Ore (SDRMS)
p.911

Research of Person-Parachute System Motion in Free-Fall Phase
p.916

Dexterous Robot’s Finger Actuated by Pneumatic Artificial Muscle
p.920

Study on PMV Index Forecasting Method Based on Fuzzy C-Means Clustering
p.925

RVE Based Numerical Evaluation on Effective Mechanical Properties of Composite with Randomly Distributed Multi-Phase Inclusions
p.931

Preparation of Modified Cu-Antibacterial White Carbon Black by Microwave Drying
p.935

Study on Morphological Structure and Application of Ag-La-Inorganic Antibacterial Material Dying by Microwave
p.941

A Study on Ammonia Nitrogen, Phosphorus and Organic Pollutants Removal of Zeolites by Salt Modification
p.947

HomeAdvanced Materials ResearchAdvanced Materials Research Vols. 383-390Study on PMV Index Forecasting Method Based on...

Study on PMV Index Forecasting Method Based on Fuzzy C-Means Clustering

Article Preview

Abstract:

In order to improve the forecasting accuracy of indoor thermal comfort, the basic principle of fuzzy c-means clustering algorithm (FCM) and support vector machines (SVM) is analyzed. A kind of SVM forecasting method based on FCM data preprocess is proposed in this paper. The large data sets can be divided into multiple mixed groups and each group is represented by a single regression model using the proposed method. The support vector machines based on fuzzy c-means clustering algorithm (FCM+SVM) and the BP neural network based on fuzzy c-means clustering algorithm (FCM+BPNN) are respectively applied to forecast PMV index. The experimental results demonstrate that the FCM+SVM method has better forecasting accuracy compared with FCM+BPNN method.

You might also be interested in these eBooks

Manufacturing Science and Technology, ICMST2011

Info:

Periodical:

Advanced Materials Research (Volumes 383-390)

Pages:

925-930

DOI:

https://doi.org/10.4028/www.scientific.net/AMR.383-390.925

Citation:

Cite this paper

Online since:

November 2011

Authors:

Chun Cheng Zhang, Xiang Guang Chen, Yuan Qing Xu

Keywords:

BP, FCM Algorithm, PMV Index, Support Vector Machine (SVM)

Export:

RIS, BibTeX

Price:

Permissions CCC:

Request Permissions

Permissions PLS:

Request Permissions

Сopyright:

© 2012 Trans Tech Publications Ltd. All Rights Reserved

Share:

Citation:

[1] O. Gassmann, H. H. Meixner. Sensors in Intelligent Buildings[M]. Weinheim(Germany): WILET-VCH Verlag GmbH, ISBN: 3-527-29557-7, (2001).

[2] Schenck, Paula. Mold in the indoor environment and health[C]. International Conference on Indoor Air Qulity/Healthy Indoor Environment, Vancouver, BC, Canada, 2003, 73-88.

[3] Fanger P O. Thermal Comfort [M]. New York: McGraw-Hill Inc, (1970).

[4] P. O. Fanger. Thermal Comfort, Analysis and Applications in Environmental Engineering[M]. New York: McGraw-Hill Inc., (1972).

[5] J. C. Bezdek. Fuzzy mathematics in pattern classification[D]. New York: Cornell University, (1973).

[6] J. C. Dunn. A fuzzy relative of the ISODATA process and its use in detecting compact, well-separated clusters[J]. J. Cybernet, 1974, (3): 32-57.

DOI: 10.1080/01969727308546046

[7] R. N. Dave, K. Bhaswan. Adaptive fuzzy c-shells clustering and detection of ellipses[J]. IEEE Trans. Neural Networks, 1992, 3(5): 643-662.

DOI: 10.1109/72.159055

[8] R. Krishnapuram, O. Nasraoui, J. Keller. The fuzzy c spherical shells algorithm: a new approach[J]. IEEE Trans. Neural Networks, 1992, 3(5): 663-671.

DOI: 10.1109/72.159056

[9] Y. Man, I. Gath. Detection and separation of ring-shaped clusters using fuzzy clustering[J]. IEEE Trans. Pattern Anal. Mach. Intell, 1994, 16(8): 855-861.

DOI: 10.1109/34.308484

[10] Deng naiyang, Tian yingjie. A New Method of Data Mining-Support Vector Machines[M]. Beijing：Science Press, 2004. TableI Clustering centers and corresponding PMV Num. Temperature (℃) Humidity (%) Wind speed (m/s) Average radiation temperature (℃) Vapor partial pressur(e/Pa) Clothing temperature (℃) Clothing thermal resistance (clo) PMV P1.

[23] 4.

[55] 5.

42.

[24] 4 3256. 8.

[29] 7.

83.

14 P2.

[26] 2.

[58] 5.

45.

[27] 2 3672. 4.

[32] 4.

98.

[1] 28 P3.

[15] 4.

[59] 4.

42.

[16] 7 2046. 3.

[21] 7.

[1] 02 -2. 20 P4.

[28] 8.

[61] 4.

40.

[29] 7 4064. 0.

[35] 0.

85.

[1] 67 P5 P6 P7.

[20] 4.

[11] 9.

[18] 1.

[60] 7.

[62] 4.

[60] 2.

51.

37.

31.

[21] 4.

[12] 9.

[19] 0 2794. 7 1514. 8 2447. 3.

[26] 6.

[18] 1.

[24] 3.

93.

[1] 20.

85 -0. 70 -2. 77 -1. 45 Table II FCM + SVM based PMV Forecasting value and error Num. Computational FCM+SVM Error 1.

[1] 11.

83 -0. 28 2.

[1] 52.

[1] 43 -0. 09 3.

[1] 46.

[1] 38 -0. 08 4.

[1] 41.

[1] 37 -0. 05 5.

97.

[1] 03.

06 56 -0. 52 -0. 46.

06 57.

52.

53.

01 58.

95.

[1] 15.

20 59 -0. 18 -0. 12.

06 60.

04.

07.

03 Table V Euclidean distance obtained with 2 prediction methods prediction method FCM+SVM FCM+BPNN Euclidean distance.

[1] 641.

[1] 650 Table III FCM + BPNN based PMV Forecasting value and error Num. Computational FCM+BPNN Error 1.

[1] 11.

82 -0. 29 2.

[1] 52.

[1] 45 -0. 07 3.

[1] 46.

[1] 40 -0. 06 4.

[1] 41.

[1] 36 -0. 05 5.

97.

[1] 04.

08 56 -0. 52 -0. 43.

09 57.

52.

54.

02 58.

95.

[1] 18.

23 59 -0. 18 -0. 14.

04 60.

04.

07.

03 Table IV Comparison on FCM+SVM and FCM+BPNN Num. Computa- tional PMV FCM+ SVM Error (FCM+ SVM) FCM+ BPNN Error (FCM+ BPNN) 1.

DOI: 10.1109/icnc.2008.274

[1] 11.

83 -0. 28.

82 -0. 29 2.

[1] 52.

[1] 43 -0. 09.

[1] 45 -0. 07 3.

[1] 46.

[1] 38 -0. 08.

[1] 40 -0. 06 4.

[1] 41.

[1] 37 -0. 05.

[1] 36 -0. 05 5.

97.

[1] 03.

06.

[1] 04.

08 56 -0. 52 -0. 46.

06 -0. 43.

09 57.

52.

53.

01.

54.

02 58.

95.

[1] 15.

20.

[1] 18.

23 59 -0. 18 -0. 12.

06 -0. 14.

04 60.

04.

07.

03.

07.

03.