23
Progress Report Presenter : Min-chia Chang Advisor : Prof. Jane Hsu Date : 2011 - 03 - 04

Progress Report

  • Upload
    huey

  • View
    56

  • Download
    0

Embed Size (px)

DESCRIPTION

Progress Report. Presenter : Min- chia Chang Advisor : Prof. Jane Hsu Date : 201 1 - 03 - 04. Outline. Prediction of AC State (revised) Definition of AC Waste Analysis Result of AC Waste Analysis Difference Control of Central AC Schedule and Goal. Outline. - PowerPoint PPT Presentation

Citation preview

Page 1: Progress  Report

Progress ReportPresenter : Min-chia ChangAdvisor : Prof. Jane HsuDate : 2011 - 03 - 04

Page 2: Progress  Report

OutlinePrediction of AC State (revised)Definition of AC Waste AnalysisResult of AC Waste AnalysisDifference Control of Central ACSchedule and Goal

2011/03/04 2NTU CSIE iAgent Lab

Page 3: Progress  Report

OutlinePrediction of AC State (revised)Definition of AC Waste AnalysisResult of AC Waste AnalysisDifference Control of Central ACSchedule and Goal

2011/03/04 3NTU CSIE iAgent Lab

Page 4: Progress  Report

Data- label: (OFF, ONno green, ONgreen) define⟶ y={0,1,2}

- OFF : close- ONno green : Tindoor < TuserSetting , valve = OFF- ONgreen : Tindoor > TuserSetting , valve = ON

- feature : define⟶ x , which is a vector- (Tindoor, Hindoor, Tvent, Hvent, Toutdoor, Houtdoor)- context information

2011/03/04 4NTU CSIE iAgent Lab

Page 5: Progress  Report

Dataset D={(x n,yn)}, where n=1 to N

- each minute of labeled period (original : intersection of vent and indoor) - labeled by camera (original : controlled on purpose by duck) - size = 77,439

Time- R336 : 2010-12-18 ~ 2011-01-06- R204 : 2011-01-06 ~ 2011-01-17- R324 : 2011-01-20 ~ 2011-01-30

Current condition : continue collecting the new data into dataset 2011/03/04 5NTU CSIE iAgent Lab

Page 6: Progress  Report

Execution environment Weka Function: SVM

- Kernel: RBF Cross Validation: 3-fold

- In each iteration :

2011/03/04 6NTU CSIE iAgent Lab

DatasetTrainingDataTestingData note : NEVER use testing data before

you predict.

Page 7: Progress  Report

Generate feature x=(Tindoor , Hindoor , Tvent , Hvent , Toutdoor , Houtdoor , context information)

2011/03/04 7NTU CSIE iAgent Lab

context data dimensions valuechilled water host 3 {0,1}chilled water temperature 1 integerrotation speed of pump 1 floatnew or old (building) 2 {0,1}floor 5 {0,1}room type 6 {0,1}zone 6 {0,1}area 1 Floatday of the week 7 {0,1}weekday or weekend 2 {0,1}semester or vacation 2 {0,1}hour of the day 24 {0,1}

total : 66 dimensions

Page 8: Progress  Report

Bagging (bootstrap aggregation)

2011/03/04 8NTU CSIE iAgent Lab

DatasetTrainingDataTestingData

K=?, S=?• K fixed - If S decreases, then time decreases.

• S fixed - If K increases, then the result of the vote is more convinced.

……

K training datasize = Ssize = S size = S

re-sampling

Page 9: Progress  Report

Bagging

2011/03/04 9NTU CSIE iAgent Lab

……

K training datasize = Ssize = S size = S

TrainingData re-samplingy=0 y=1 y=2

S/3 S/3 S/3

• K fixed - If S decreases, then time decreases.

size = 51626

Page 10: Progress  Report

2011/03/01 10NTU CSIE iAgent Lab

x=60dim K=1 K=2 K=3 K=4 K=5 K=10 K=30 K=100S=3 47.76%0m12s 52.80%0m21s 55.80%0m30s 58.13%0m46s 55.69%0m49s 67.49%1m36s 65.32%4m47s 68.65%16m14sS=300 62.03%0m18s 65.81%0m33s 71.87%0m48s 66.84%1m15s 74.32%1m22s77.50%2m37s 80.47%8m03s 81.00%26m41sS=1500 85.41%0m40s 84.60%1m13s 90.35%1m52s 85.11%2m40s 89.09%3m11s 90.15%7m41s 91.35%18m41s

91.61%65m13sS=3000 89.58%1m37s 87.76%2m26s 91.48%4m14s 91.47%8m30s 92.48%7m8s92.50%16m6s 93.45%44m04sS=7500 93.55%3m56s 92.45%7m30s 94.32%10m36s 93.90%16m51s

94.66%19m33s 94.99%37m33sS=15000 94.67%15m05s96.16%26m19s 95.25%41m41s 95.05%62m45s

95.42%74m23sS=22500 95.19%23m53s94.62%47m0s 95.60%74m22s

S=30000 95.30%37m07s94.77%72m12s

baseline:96.11%92m50sK=?, S=?

Page 11: Progress  Report

Process the missing value missing value : Tindoor , Hindoor , Tvent , Hvent processing method :

method 1 : encoding e.g. : (?, ?, 15.2, ?) => (0, ?, 0, ?, 1, 15.2, 0, ?)

method 2 : interpolation   (linear) e.g. : 2011-01-31 23:50 : (20, 45, 10, 70) 2011-01-31 23:51 : (?, 45.2, 9.9, 70.2)…… 2011-02-01 00:00 : (20.1,45.5,10.2,69.2)=> ? = 20.01

method 3 : encoding + interpolation

2011/03/04 11NTU CSIE iAgent Lab

Page 12: Progress  Report

Result

2011/03/04 12NTU CSIE iAgent Lab

baseline bagging(K=?, S=?)(Tvent) 72.66%12m32s6 dim 93.21%45m12s6 dim +generate features (total : 60 dim)96.11%92m50s60 dim + missingValue : Encode 96.07%73m43s60 dim + missingValue : Interpolation 99.79%54m41s60 dim + missingValue : Encode, Interpolation 99.60%93m40s

60 dim + missingValue : Interpolation + normalize 97.55%

Page 13: Progress  Report

OutlinePrediction of AC State (revised)Definition of AC Waste AnalysisResult of AC Waste AnalysisDifference Control of Central ACSchedule and Goal

2011/03/04 13NTU CSIE iAgent Lab

Page 14: Progress  Report

Problem definition : energy(AC) waste analysis Component 2 – AC state predictor:

- input : AC information- output : AC state ( yn={0, 1, 2} )

Component 3 – thermal comfort calculator :- input : thermal comfort questionnaire, Toutdoor- output : thermal comfort range

Component 4 – AC waste analysis : - input : mn , yn , thermal comfort range, Tindoor- output : proportion of AC waste

2011/03/04 14NTU CSIE iAgent Lab

Page 15: Progress  Report

System overview

2011/03/04 15NTU CSIE iAgent Lab

AC statepredictorthermal comfortcalculator

AC wasteanalysisACinformationthermal comfort questionnaire

Toutdoor

motion sensorstateACstate

thermal comfortrange Tindoor

proportion of AC waste

Page 16: Progress  Report

Condition of AC waste state of motion sensormn state of ACyn Tindoor ? TcomfortableRange waste or notN 0 higher NN 0 among NN 0 lower NN 1 higher Y N 1 among YN 1 lower YN 2 higher YN 2 among YN 2 lower YY 0 higher NY 0 among NY 0 lower NY 1 higher NY 1 among NY 1 lower NY 2 higher abnormalY 2 among NY 2 lower Y

2011/03/04 16NTU CSIE iAgent Lab

Page 17: Progress  Report

OutlinePrediction of AC State (revised)Definition of AC Waste AnalysisResult of AC Waste AnalysisDifference Control of Central ACSchedule and Goal

2011/03/04 17NTU CSIE iAgent Lab

Page 18: Progress  Report

Condition of AC   waste situation 1. mn = no and (yn = 1 or yn = 2) 2. mn = yes and yn = 2 and Tindoor < TcomfortableRange abnormal situation 1. mn = yes and yn = 2 and Tindoor > TcomfortableRange

2011/03/04 18NTU CSIE iAgent Lab

Page 19: Progress  Report

Proportion of AC waste waste situation 1. mn = no and (yn = 1 or yn = 2) 2. mn = yes and yn = 2 and Tindoor < TcomfortableRange

2011/02/21 19NTU CSIE iAgent Lab

place mn=no mn=yes yn=0 yn=1 yn=2 waste 1 waste 2 abnormal336_2 58% 42% 26% 61% 13% 36.7% 9.5% 0%204_1 47% 53% 23% 70% 7% 33.8% 4.0% 0%204_2 43% 57% 19% 39% 42% 35.8% 17.8% 0%204_3 48% 52% 53% 44% 3% 23.9% 16.7% 0%204_4 57% 43% 53% 41% 6% 28.2% 4.0% 0%204_5 65% 35% 15% 13% 72% 57.2% 20.8% 0%204_6 65% 35% 55% 41% 5% 31.9% 2.3% 0%204 33% 67% 7% - - 31.0% - -

2011.01

Page 20: Progress  Report

OutlinePrediction of AC State (revised)Definition of AC Waste AnalysisResult of AC Waste AnalysisDifference Control of Central ACSchedule and Goal

2011/03/04 20NTU CSIE iAgent Lab

Page 21: Progress  Report

OutlinePrediction of AC State (revised)Definition of AC Waste AnalysisResult of AC Waste AnalysisDifference Control of Central ACSchedule and Goal

2011/03/04 21NTU CSIE iAgent Lab

Page 22: Progress  Report

Schedule and Goal Schedule (March)

- next step after result of AC waste analysis- definition of thesis part 2- data aggregation of thesis part 2 - thesis writing : AC waste analysis (CH1, CH3) - (?) implementation of thesis part 2

Goal (this semester) - 100 年 6 月順利口試

2011/03/04 22NTU CSIE iAgent Lab

Page 23: Progress  Report

Thank you for listening !

2011/03/04 23NTU CSIE iAgent Lab