Reliability of travel time estimation

A method for increasing accuracy in estimating an average time taken to travel through a chosen road segment, the method comprising:

• a road traffic sensing system collecting data corresponding to one or more vehicles travelling through road segments, thereby enabling determination of a time taken by the one or more vehicles to travel through the road segments;
• at least one processor identifying one or more correlated road segments for which the time taken to travel through the correlated road segments is correlated with the time taken to travel through the chosen road segment, the processor identifying one or more preferred road segments among the correlated road segments that increases the accuracy in determining the average time taken to travel through the chosen road segment;
• and
• the processor estimating the average time taken to travel through the chosen road segment using data corresponding to time taken to travel through the preferred road segments and the time taken to travel through the chosen road segment,
• characterized in that identifying the preferred road segments comprises:
• identifying the correlated road segments which comprise exclusive vehicles for which the time taken to travel is known for the correlated road segments and for which the time taken to travel is not known for the chosen road segment;
• estimating the time taken to travel by each of the exclusive vehicles on the chosen road segment using the correlation between the correlated road segments and the chosen road segment;
• determining, for each of the correlated road segments, an improvement that is achieved in estimating the average travel time by using data corresponding to an estimated time taken to travel by each of the exclusive vehicles on the chosen road segment;
• sorting the correlated road segments based on the improvement achieved in estimating the average travel time;
• and
• determining a reduction in error in estimating the average time taken to travel through the chosen road segment by using the time taken by exclusive vehicles, by considering the correlated road segments in decreasing order of reduction in error, till one of the correlated road segments does not reduce the error in the estimate compared to the reduction in error by the previous correlated road segment which provided the reduction in error, wherein the correlated road segments which provided the reduction in error in estimate are considered as preferred road segments
• wherein, determining the statistics related to the time taken to travel the chosen road segment comprises, estimating average time taken to travel through the chosen road segment using the equation:
• μ
• 1
• =
• (
• ∑
• i
• =
• 1
• M
• ⁢
• Y
• ⁡
• (
• i
• )
• +
• ∑
• j
• =
• 1
• N
• ⁢
• Y
• ⁡
• (
• j
• )
• 1
• )
• /
• (
• M
• +
• N
• )
• wherein
• μ1 is an estimate of the average travel time to travel through the chosen road segments,
• M is the number of vehicles for which the time taken to travel through the chosen road segments is available,
• N is the exclusive number of vehicles corresponding to the preferred road segment,
• Y(i) is the time taken to travel through the chosen road segment by each of the M vehicles, and
• Y(j)1 is the time taken to travel through the chosen road segment by each of the N exclusive vehicles, estimated using the correlation between the preferred road segment and the chosen road segment.

A system for increasing accuracy in estimating an average time taken to travel through a chosen road segment, the system comprising:

• a road traffic sensing system configured to collect data corresponding to one or more vehicles travelling through road segments, thereby enabling determination of a time taken by the one or more vehicles to travel through the road segments;
• at least one data repository configured to:
• store a list of one or more correlated road segments for which the time taken to travel through the correlated road segments is correlated with the time taken to travel through the chosen road segment;
• and
• store historical data corresponding to the time taken to travel through the road segments;
• and
• at least one processor configured to:
• determine the time taken by the one or more vehicles to travel through the road segments using the data collected by the road traffic sensing system, corresponding to one or more vehicles travelling through road segments;
• identify the one or more correlated road segments for which the time taken to travel through the correlated road segments is correlated with the time taken to travel through the chosen road segment;
• identify one or more preferred road segments among the correlated road segments that increases the accuracy in determining average time taken to travel through the chosen road segment;
• and
• determine the average time taken to travel through the chosen road segment using data corresponding to the time taken to travel through the preferred road segments and using data corresponding to the time taken to travel through the chosen road segment,
• characterized in that the processor is configured to:
• identify the correlated road segments which comprise exclusive vehicles for which time taken to travel is known for the correlated road segments and not known for the chosen road segment;
• estimate time taken to travel by each of the exclusive vehicles on the chosen road segment using the correlation between the correlated road segments and the chosen road segment;
• determine, for each of the correlated road segments, improvement that is achieved in error in estimating average travel time by using data corresponding to the estimated time taken to travel by each of the exclusive vehicles on the chosen road segment;
• sort the correlated road segments based on improvement achieved in error in estimating average travel time;
• and
• determine reduction in error in estimating the average time taken to travel through the chosen road segment by using the time taken by exclusive vehicles, by considering correlated road segments in decreasing order of reduction in error, till one of the correlated road segments does not reduce error in estimate compared to the reduction in error by previous correlated road segment which provided reduction in error, wherein the correlated road segments which provided reduction in error in estimate are considered as preferred road segments
• wherein the processor is configured to estimate the time taken to travel by each of the exclusive vehicles on the chosen road segment using the correlation between the correlated road segments and the chosen road segment, using the equation:
• Y(j)1=f(X(j)),1≦j≦N wherein
• N is the number of exclusive vehicles corresponding to one of the correlated road segments;
• X(j) is the time taken to travel by the Jth exclusive vehicle through the one of the correlated road segments,
• Y(j)1 is the estimated time taken to travel through the chosen road segment by each of the Jth exclusive vehicles, and
• f is a function.

A system for increasing accuracy in estimating an average time taken to travel through a chosen road segment, the system comprising:

• a road traffic sensing system configured to collect data corresponding to one or more vehicles travelling through road segments, thereby enabling determination of a time taken by the one or more vehicles to travel through the road segments;
• at least one data repository configured to:
• store a list of one or more correlated road segments for which the time taken to travel through the correlated road segments is correlated with the time taken to travel through the chosen road segment;
• and
• store historical data corresponding to the time taken to travel through the road segments;
• and
• at least one processor configured to:
• determine the time taken by the one or more vehicles to travel through the road segments using the data collected by the road traffic sensing system, corresponding to one or more vehicles travelling through road segments;
• identify the one or more correlated road segments for which the time taken to travel through the correlated road segments is correlated with the time taken to travel through the chosen road segment;
• identify one or more preferred road segments among the correlated road segments that increases the accuracy in determining average time taken to travel through the chosen road segment;
• and
• determine the average time taken to travel through the chosen road segment using data corresponding to the time taken to travel through the preferred road segments and using data corresponding to the time taken to travel through the chosen road segment,
• characterized in that the processor is configured to:
• identify the correlated road segments which comprise exclusive vehicles for which time taken to travel is known for the correlated road segments and not known for the chosen road segment;
• estimate time taken to travel by each of the exclusive vehicles on the chosen road segment using the correlation between the correlated road segments and the chosen road segment;
• determine, for each of the correlated road segments, improvement that is achieved in error in estimating average travel time by using data corresponding to the estimated time taken to travel by each of the exclusive vehicles on the chosen road segment;
• sort the correlated road segments based on improvement achieved in error in estimating average travel time;
• and
• determine reduction in error in estimating the average time taken to travel through the chosen road segment by using the time taken by exclusive vehicles, by considering correlated road segments in decreasing order of reduction in error, till one of the correlated road segments does not reduce error in estimate compared to the reduction in error by previous correlated road segment which provided reduction in error, wherein the correlated road segments which provided reduction in error in estimate are considered as preferred road segments, wherein the processor is configured to determine the statistics related to the time taken to travel the chosen road segment by estimating the average time taken to travel through the chosen road segment using the equation:
• μ
• 1
• =
• (
• ∑
• i
• =
• 1
• M
• ⁢
• Y
• ⁡
• (
• i
• )
• +
• ∑
• j
• =
• 1
• N
• ⁢
• Y
• ⁡
• (
• j
• )
• 1
• )
• /
• (
• M
• +
• N
• )
• wherein
• μ1 is an estimate of the average travel time to travel through the chosen road segment,
• M is the number of vehicles for which the time taken to travel through the chosen road segment is available,
• N is the exclusive number of vehicles corresponding to the preferred road segment,
• Y(i) is the time taken to travel through the chosen road segment by each of the M vehicles, and
• Y(j)1 is the time taken to travel through the chosen road segment by each of the N exclusive vehicles, estimated using the correlation between the preferred road segment and the chosen road segment.