Off-line approximate dynamic programming for the vehicle routing problem with a highly variable customer basis and stochastic demands
نویسندگان
چکیده
We study a stochastic variant of the vehicle routing problem (VRP) arising in context domestic donor collection services. The we consider combines following attributes. Customers requesting services are variable, sense that they stochastic, but not restricted to predefined set. Furthermore, demand volumes also and observed upon visiting customers. objective is maximize expected served demands while meeting capacity time restrictions. call this VRP with highly Variable Customer basis Stochastic Demands (VRP-VCSD). first propose classical Markov Decision Process (MDP) formulation for VRP-VCSD. resulting model is, however, unusable due explosion dimension state action spaces. To solve VRP-VCSD, number methodological contributions aimed at reducing reformulate MDP as an consecutive selection procedure. In formulation, enforce treatment single (as opposed multiple vehicles) each decision epoch. then introduce observation function selects subset available information, which deemed relevant considered develop Q-learning algorithm called QN-CO. particular, use continuous representation incorporate two-layer artificial neural network approximate Q values. aggregation strategy yielding fixed-size output. Finally, enhance our Replay Memory Double Network. conduct thorough computational analysis. Results show QN-CO considerably outperforms five benchmark policies. Moreover, can compete specialized methods developed particular case VRP-VCSD where customer locations known advance.
منابع مشابه
An approximate dynamic programming approach for the vehicle routing problem with stochastic demands
This paper examines approximate dynamic programming algorithms for the single-vehicle routing problem with stochastic demands from a dynamic or reoptimization perspective. The methods extend the rollout algorithm by implementing different base sequences (i.e. a priori solutions), look-ahead policies, and pruning schemes. The paper also considers computing the cost-to-go with Monte Carlo simulat...
متن کاملCompetitive Vehicle Routing Problem with Time Windows and Stochastic Demands
The competitive vehicle routing problem is one of the important issues in transportation area. In this paper a new method for competitive VRP with time windows and stochastic demand is introduced. In the presented method a three time bounds are given and the probability of arrival time between each time bound is assumed to be uniform. The demands of each customer are different in each time wind...
متن کاملComparing neuro-dynamic programming algorithms for the vehicle routing problem with stochastic demands
The paper considers a version of the vehicle routing problem where customers' demands are uncertain. The focus is on dynamically routing a single vehicle to serve the demands of a known set of geographically dispersed customers during real-time operations. The goal consists of minimizing the expected distance traveled in order to serve all customers' demands. Since actual demand is revealed upo...
متن کاملFuzzy Vehicle Routing Problem with Split Delivery: Trade-off between Air Pollution and Customer Satisfaction
In large-scale emergency, the vehicle routing problem focuses on finding the best routes for vehicles. The equitable distribution has a vital role in this problem to decrease the number of death and save people's lives. In addition to this, air pollution is a threat to people’s life and it can be considered to omit other kinds of disasters happens because of it. So, a new MINLP model presented ...
متن کاملذخیره در منابع من
با ذخیره ی این منبع در منابع من، دسترسی به آن را برای استفاده های بعدی آسان تر کنید
ژورنال
عنوان ژورنال: Computers & Operations Research
سال: 2023
ISSN: ['0305-0548', '1873-765X']
DOI: https://doi.org/10.1016/j.cor.2023.106338