Abstract

Rainfall is regarded as a major uncertainty factor that has adverse impacts on productivity and duration of highway construction activities. In practice, given the location, type, start date, and original duration of the activities, a common approach for construction schedulers to assess the effect of rain is by adding a certain percentage of time to tasks. However, this method depends mainly on the experience and subjective judgment of the schedulers, who may be unfamiliar with the rainfall pattern and its impact on productivity of the operations, and thus, oftentimes produces inaccurate results. This paper presents a model that utilizes historical daily rainfall data and experts' knowledge, and employs fuzzy set concept for assessing the impact of rain on project completion. Based on the model, a fuzzy reasoning knowledge-based scheduling system (FRESS) is proposed. A case study involving a highway construction project implemented in two geographic areas with different rainfall environments is presented to illustrate the salient features of the system that allows users to simulate experts' judgment and to demonstrate the capability and effectiveness of the system that can assist contractors to better estimate activity durations for projects in geographical locations having rainfall data.

Keywords: Rain impact; Activity duration; Knowledge base; Fuzzy logic

Nomenclature

C_ij

fuzzy set of adverse consequence regarding rain level i and state factor state j;

D

direct productivity loss caused by a particular amount of precipitation;

d_p

direct productivity loss with respect to the jth column element x_j in Eq. (17);

F_ij

fuzzy set of frequency of occurrence regarding rain level i and state factor j;

i

indicator of rain level, 1, 2, 3, 4, and 5 representing drizzle, sprinkle, medium rain, drench, and cloudburst, respectively;

j

indicator of state factor, 1 and 2 representing exposure level to rain and soil drainage condition, respectively;

L_D

direct productivity loss;

L_j

indirect productivity loss of a single rainfall for the following jth day;

L_i

productivity loss on the ith day due to multiple rainfalls;

L_iD

direct productivity loss corresponding to a particular precipitation on the ith day;

P

matrix of direct productivity loss regarding each rain level i;

the average direct productivity loss with respect to a particular rain level i;

P_ij

fuzzy set of productivity loss associated with rain level i and state factor j;

R

fuzzy set of rain level;

R_i

fuzzy subset representing rain level i;

S_i

total effect of C_ij and P_ij;

S_ij

fuzzy relation between C_ij and P_ij;

T

the period in which activity productivity is affected under a particular amount of rainfall;

T_D

the period of rainfall causing a direct productivity loss;

T_I

period between the end of the rainfall and the no longer loss in productivity;

T_i

total effect of F_ij and C_ij;

T_ij

fuzzy relation between F_ij and C_ij;

T_O

original activity duration;

T_R

impacted activity duration attributed to a single and multiple rainfalls;

Y

referenced value of chance-of-work;

Z

matrix of the period that activity productivity is affected by each rain level i;

1−L_i

residual productivity;

α_ij

matching degree between input variable x_j and rule condition R_i about x_j;

μ_Ci(y)

membership value in the rule's consequents;

μ_R1(x_i)

membership function for element x_i with respect to the fuzzy subset R₁;

ω

minimum threshold value of chance-of-work;

∑L_ij

sum of following productivity losses due to the previous rainfall on the ith day;

∩

fuzzy intersection;

fuzzy union;

fuzzy composition.

Article Outline

Nomenclature

1. Introduction

2. Methodology

2.1. Rain classification and rain-related variables

2.2. Fuzzy logic model for rain impact assessment

2.3. Fuzzy relation and fuzzy composition

2.4. Direct productivity loss

2.5. Indirect productivity loss of a single rainfall

2.6. Indirect productivity loss of multiple rainfalls

2.7. Assessment of extended activity duration

2.8. Decision of work or no-work

3. Fuzzy reasoning knowledge-based scheduling system

4. A case study

5. Conclusion

Acknowledgements

References