THE ESTIMATION OF SATURATION FLOW, EFFECTIVE GREEN TIME AND PASSENGER CAR EQUIVALENTS AT TRAFFIC SIGNALS BY MULTIPLE LINEAR REGRESSION.

THE SATURATION FLOW AND EFFECTIVE GREEN TIME ON AN APPROACH TO A TRAFFIC SIGNAL ARE VERY IMPORTANT INPUTS TO METHODS OF ESTIMATING DELAY-MINIMIZING OR CAPACITY MAXIMIZING SIGNAL SETTINGS. THIS PAPER OUTLINES PRACTICAL METHODS WHICH USE MULTIPLE LINEAR REGRESSION TO ESTIMATE THESE PARAMETERS AND ALSO THE PCU VALUES OF DIFFERENT VEHICLE TYPES. IT IS SHOWN THAT THE ESTIMATED TIMES OF THE BEGINNING AND END OF THE EFFECTIVE GREEN PERIOD FOR AN APPROACH ARE DEPENDENT ON THE METHOD USED TO ESTIMATE STARTING DELAYS AND THE AMOUNT OF AMBER TIME THAT IS EFFECTIVELY GREEN. "SYNCHRONOUS" METHODS WHICH TERMINATE OBSERVATION AT THE INSTANT OF A VEHICLE DEPARTURE WILL PRODUCE ESTIMATES OF THESE TWO QUANTITIES WHICH ARE 1/(2S') (WHERE S' IS THE ESTIMATED SATURATION FLOW IN VEHICLES/UNIT TIME) SHORTER THAN ESTIMATES PRODUCED BY "ASYNCHRONOUS" METHODS WHICH TERMINATE OBSERVATIONS AT AN ARBITRARY POINT IN TIME. THIS MEANS THAT ASYNCHRONOUS METHODS ESTIMATE THE EFFECTIVE GREEN PERIOD TO OCCUR 1/(2S') LATER THAN DO SYNCHRONOUS METHODS, WITHOUT THE MAGNITUDE OF THE EFFECTIVE GREEN TIME BEING AFFECTED. THE SYNCHRONOUS AND ASYNCHRONOUS METHODS ARE APPLIED TO DATA FROM A SIMPLE SITE, AND TWO SOURCES OF BIAS ARE IDENTIFIED IN THE DATA FOR THE SYNCHRONOUS METHOD. WHEN THE DATA FOR THE SYNCHRONOUS METHOD ARE MODIFIED TO ELIMINATE THESE BIASES THE PARAMETER ESTIMATES FROM THE TWO METHODS ARE FOUND TO BE CONSISTENT WITH ONE ANOTHER AND WITH VALUES REPORTED ELSEWHERE.(Author/publisher).