The probabilistic nature of traffic breakdown and freeway capacity at an on-ramp bottleneckis discussed based on the author's three-phase traffic flow theory. It is stressed that theprobabilistic nature of traffic breakdown leads to a limitation for reliable applications of free flow control approach in which free flow should be maintained at the bottleneck. It is explainedthat earlier traffic flow models and related simulation tools cannot explain and cannotpredict traffic breakdown at the bottleneck as it is observed in empirical (measured) data. Forthis reason, the related simulations of freeway control and management strategies cannot predict many of the freeway traffic phenomena that would occur through the use of thesestrategies. Explanations of this critical conclusion are made. An introduction tomicroscopic freeway simulations of freeway control in the context of three-phase traffictheory is outlined, which shows and explains empirical features of traffic breakdown andresulting congested patterns. In the following parts II and III, the consequences ofprobabilistic nature of traffic breakdown for onramp metering strategies are discussed. Firstly(part II), a critical analysis of the well-known ALINEA method for feedback on-ramp meteringis performed. In part III, a possible solution of the on-ramp metering problem - a recentcongested pattern control approach (ANCONA) - is considered. Parts II and III will be publishedin the February and March issues of TEC.
Abstract