Historical SpeedPatterns in tour optimization

This forum deals with any kind of trip optimization whether it is automatic planning or manual dispatching, refering to transport orders or service planning.
Post Reply
User avatar
Bernd Welter
Site Admin
Posts: 2564
Joined: Mon Apr 14, 2014 10:28 am
Contact:

Historical SpeedPatterns in tour optimization

Post by Bernd Welter »

Hello there,

today I've been asked (once more) whether and how it is possible to consider our well known PTV_SpeedPatterns in the context of a tour optimization via xTour. The expectation of the players is that we consider the daytime dependent driving time (and distance) during the optimization.

Well, the answer is: sorry, that doesn't work :cry: . Reason: Though the SpeedPatterns enable you to compute the driving time and distance of a relation in a simple 1:1 routing based on a given START TIME (or END TIME) this is a mechanism that is not available within the required distance matrices of tour optimization: A distance matrix is like a snapshot considering ONE timestamp for all the relations being stored but as soon as the algorithm retrieves distances and periods from the DIMA there is no longer a context of time.

I know that this is a meaningful wish but technologywise it is not covered by the current interfaces.

Best regards Bernd
Bernd Welter
Technical Partner Manager Developer Components
PTV Logistics - Germany

Bernd at... The Forum,LinkedIn, Youtube, StackOverflow
I like the smell of PTV Developer in the morning... :twisted:
User avatar
Bernd Welter
Site Admin
Posts: 2564
Joined: Mon Apr 14, 2014 10:28 am
Contact:

Re: Historical SpeedPatterns in tour optimization

Post by Bernd Welter »

And: almost the same answer if someone asks for PTV_TrafficIncidents (Live traffic info). This only works if the planning scope is right now and on a short horizon.

What is possible (but doesn't solve the whole story) is this: you can set the reference time of the distance matrix to a specific point in time. But in the planning itself usually the horizon is larger than just 1 hour.
Optimization in Paris - distance matrix based on 24:00
Optimization in Paris - distance matrix based on 24:00
Optimization in Paris - distance matrix based on 12:00
Optimization in Paris - distance matrix based on 12:00
As the distance matrices are different for these two timestamps the sequences and periods changed.

Furthermore we know that tour planning is based on two independent usecase scenarios:
  • construction of tours under optimal conditions (designing tours)
  • updating tours due to live info, e.g. "vehicle delay" or "vehicle breakdown" (repair mode). For this case we have to develop other approaches. The current xTour 1.24.x is focusing on the DESIGN.
Paris with SPEED PATTERNS at midnight
Paris with SPEED PATTERNS at midnight
Paris with SPEED PATTERNS at noon
Paris with SPEED PATTERNS at noon
Bernd Welter
Technical Partner Manager Developer Components
PTV Logistics - Germany

Bernd at... The Forum,LinkedIn, Youtube, StackOverflow
I like the smell of PTV Developer in the morning... :twisted:
Post Reply