| [Previous: Introduction] | [Table Of Contents] | [Next: Crash Prediction Procedural Elements] |
In addition to the IHSDM system-level input (e.g., Project, Evaluation), the input data to the CPM consists of the following categories of data:
Data Requirements discusses the data used by all modules in more detail.
The evaluation data consist of the evaluation bounds and the evaluation period. The evaluation bounds are the beginning and ending stations for the portion of the project to be evaluated. The evaluation bounds must be stations within the project selected and are set from the "Set crash prediction attributes" screen in the Evaluation Wizard.
The evaluation period is the time period for which the safety performance is to be predicted. The evaluation period is usually (but not necessarily) a period in the future. The minimum time period for which safety performance can be predicted is one year; the CPM does not provide predictions for partial years. The evaluation period is set from the "Set crash prediction attributes" screen in the Evaluation Wizard.
The geometric and traffic control data required to run the CPM can be divided into two categories: the highway segment geometric and traffic control data, and the intersection geometric and traffic control data.
The following highway geometric design and traffic control data are required to run the CPM:
Values of some geometric characteristics may differ by direction of travel (i.e., lane width, shoulder width, shoulder type, and passing lanes). These data are entered separately for each direction of travel: "right" refers to the right side of the centerline in the direction of increasing stations and "left" refers to the left side of the centerline in the direction of increasing stations. Values for driveway density (DD) and roadside hazard rating (RHR) are for both sides of the road combined. Climbing lanes are treated like passing lanes by the crash prediction algorithm.
When historical crash data are not used in the evaluation, data for all the variables listed must be entered for the entire project within the evaluation bounds of the highway used in the after period or evaluation period (proposed highway). The CPM uses the highway geometric design and traffic control data to create a Highway Segment Data Table (see Section 4.1.1, Highway Segment Data (Current and Proposed)) for the proposed highway.
When historical crash data are used in the evaluation, data for all the variables listed must be entered for the entire project within the evaluation bounds of the highway used in the before period (current highway) and after period (proposed highway). The CPM uses these data to create a Highway Segment Data Table for both the current and proposed highway.
The following intersection geometric design and traffic control data are required to run the CPM:
The CPM uses these data to create an intersection data table Section 4.1.3, Intersection Data (Current and Proposed) for the proposed intersection if historical data are not used in the evaluation.
If historical data are used in the evaluation, the CPM uses these data to create an Intersection Data Table for both the current and proposed highways.
The traffic volume data required to run the CPM can be divided into two categories: the highway segment traffic volume data, and the intersection traffic volume data.
ADT data are needed for the entire project within the evaluation bounds for the evaluation period specified when historical crash data are not used in the evaluation.
ADT are needed for all years for both the period when crash history data are available (before period) and the evaluation (after) period when historical crash data are used in the evaluation.
If ADT data are missing for any years or segments that are included in the evaluation, then the CPM will interpolate missing ADT data based on rules described in the Crash Prediction Algorithm, Step 4.
ADT data are needed for each intersection for the evaluation period specified when historical crash data are not used in the evaluation.
ADT data are needed for all years for both the period when crash history data are available (before period) and the evaluation (after) period when historical crash data are used in the evaluation.
The major leg ADTs for each intersection can be determined directly from the highway segment ADTs. The minor leg ADTs need to be entered and cannot be interpolated by the CPM.
Crash history data can be used to improve the overall crash prediction from the CPM.
Section 3.5, Crash Prediction When Site-Specific Crash History Data are Availablediscusses when crash history data should and should not be used in the CPM. Values for the following variables are needed when historical crash data are used:
Crashes for which the relationship to intersection is stored as "intersection-related" or "non-intersection-related" are analyzed as such by the CPM. Crashes for which the relationship to intersection is stored as "unknown" and which fall within 76 m (250 ft) of the center of an intersection are attributed to that intersection. If the centers of two intersections are closer than 500 feet, the historical crashes between the two intersections are assigned to the closest intersection. Crashes stored as "unknown" and located more than 76 m (250 ft) from the center of an intersection are attributed to the highway segment within which the crash falls.
Crash history data are applied to the CPM in Step 14 of the Crash Prediction Algorithm.
Crashes assigned to highway segments are stored by specific location.
Crashes are assigned to a homogeneous analysis section (homogeneous highway segment or intersection).
Crashes that are stored as "intersection-related" and have a value defined for the "intersection location" are assigned to the intersection at the specified station. Crashes that are stored as "intersection-related" and do not have a value defined for the "intersection location" are assigned to the intersection closest to the crash location.
| [Previous: Introduction] | [Top] | [Next: Crash Prediction Procedural Elements] |