VerologSolverChallenge2017 | Project for Verolog Solver Challenge

VerologSolverChallenge2017 is a Python library. VerologSolverChallenge2017 has no bugs, it has no vulnerabilities and it has low support. However VerologSolverChallenge2017 build file is not available. You can download it from GitHub.

Project for Verolog Solver Challenge 2017. , more info about it: In this file is described the flow of the algorithm. Flow of algorithm consists of two main steps: 1) Decide which request will be executed on which day 2) Decide which vehicle will execute which request. For 1) we have two possible solutions: 1.1) Random 1.2) Greedy algorithm 1.3) Genetic algorithm 1.4) DFS with memoization. After we decide days, we create "negative requests", i.e. request when we need to pick up tools from customer. For 2) we propose Genetic algorithm as solution with several greedy optimizations. 1.1) Random - for each request, set some random (in bounds) day when it is executed - that didn't go so well :). 1.2) Greedy algorithm - for each request, set day when it is executed in the following way: 1. look for all days when it can be executed. 2. pick a day which is "most empty", i.e. day where the particular tool is least used. 1.3) Genetic Algorithm description 1.3.1) Chromosome description Chromosome consists of request list. daysMap where key of the map is day number and value is other map where key is tool id and value its usage. 1.3.2) Crossover description There are two types of crossovers. In crossover type 1 there is one cross point x which is random number between 0 and number of requests. Childs requests from 0 to x are setted at same days as first parents and other part is taken from other parent. Crossover type 2 focuses on days which exceeds maximal number of available tool. It chooses requests corresponding to day and tool slot which exceed number tools available. Day and tool type are selected using wheel selection. Crossover of type 2 is selected with probability of 0.6 only if there are slots which are exceeded. 1.3.3) Mutation description There are two types of mutation. In first type each request has probabiltiy to be assigned to random day. Second type selects again with wheel selection day and tool slot which exceed number of tools available, and each request corresponding to that slot has probability to be assigned to random day. Mutation of type 2 is selected with probability of 0.6 only if there are slots which are exceeded. 1.3.4) Selection description For selection 3-tournament is used. 1.4) DFS with memoization description - idea: Attach for each request on which day it is executed, but take care about count of tools - > - map for memoization - sort requests in array by tool they use, lastDayDelivery - firstDayDelivery, reverse(numberOfTools) - each node is requst with some day for delivery in interval [firstDay, lastDay] - children are nodes generated in the following way: - all children have request.id of next request in an array - picked day for delivery is one of the values between firstDay and lastDay - all children differ only in picked day - check if some branch is available by using a map above.