Hongtao Lei

The capacitated vehicle routing problem with stochastic demands and time windows

The capacitated vehicle routing problem with stochastic demands and time windows is an extension of the capacitated vehicle routing problem with stochastic demands, in which demands are stochastic and a time window is imposed on each vertex. A vertex failure occurring when the realized demand exceeds the vehicle capacity may trigger a chain reaction of failures on the remaining vertices in the same route, as a result of time windows. This paper models this problem as a stochastic program with recourse, and proposes an adaptive large neighborhood search heuristic for its solution. Modified Solomon benchmark instances are used in the experiments. Computational results clearly show the superiority of the proposed heuristic over an alternative solution approach.

Districting for routing with stochastic customers

We introduce the vehicle routing and districting problem with stochastic customers (VRDPSC). This problem is modelled and solved as a two-stage stochastic program during which the districting decisions are made in the first stage and the Beardwood–Halton–Hammersley formula is used to approximate the expected routing cost of each district in the second stage. District compactness is also considered as part of the objective function. We have developed a large neighbourhood search heuristic for VRDPSC. The heuristic was tested on modified Solomon instances and on modified Gehring and Homberger instances. Extensive computational results confirm the effectiveness of the proposed heuristic.

Solving a multi-objective dynamic stochastic districting and routing problem with a co-evolutionary algorithm

This study considers a multi-objective dynamic stochastic districting and routing problem in which the customers of a territory stochastically evolve over several periods of a planning horizon, and where the number of service vehicles, the compactness of the districts, the dissimilarity measure of the districts and an equity measure of vehicles profit are considered as objectives. The problem is modeled and solved as a two-stage stochastic program, where in each period, districting decisions are made in the first stage, and the Beardwood-Halton-Hammersley formula is used to approximate the expected routing cost of each district in the second stage. An enhanced multi-objective evolutionary algorithm (MOEA), i.e., the preference-inspired co-evolutionary algorithm using mating restriction, is developed for the problem. The algorithm is tested on randomly generated instances and is compared with two state-of-the-art MOEAs. Computational results confirm the superiority and effectiveness of the proposed algorithm. Moreover, a procedure for selecting a preferred design for the proposed problem is described. HighlightsWe consider a multi-objective dynamic stochastic districting and routing problem.An enhanced co-evolutionary algorithm called PICEA-g-mr is proposed.The PICEA-g-mr outperforms two MOEAs on randomly generated instances.A procedure for selecting a preferred design for the problem is illustrated.

A multi-objective co-evolutionary algorithm for energy-efficient scheduling on a green data center

Nowadays, the environment protection and the energy crisis prompt more computing centers and data centers to use the green renewable energy in their power supply. To improve the efficiency of the renewable energy utilization and the task implementation, the computational tasks of data center should match the renewable energy supply. This paper considers a multi-objective energy-efficient task scheduling problem on a green data center partially powered by the renewable energy, where the computing nodes of the data center are DVFS-enabled. An enhanced multi-objective co-evolutionary algorithm, called OL-PICEA-g, is proposed for solving the problem, where the PICEA-g algorithm with the generalized opposition based learning is applied to search the suitable computing node, supply voltage and clock frequency for the task computation, and the smart time scheduling strategy is employed to determine the start and finish time of the task on the chosen node. In the experiments, the proposed OL-PICEA-g algorithm is compared with the PICEA-g algorithm, the smart time scheduling strategy is compared with two other scheduling strategies, i.e., Green-Oriented Scheduling Strategy and Time-Oriented Scheduling Strategy, different parameters are also tested on the randomly generated instances. Experimental results confirm the superiority and effectiveness of the proposed algorithm. HighlightsWe consider multi-objective energy efficient scheduling on a green data center.The task model, energy model and scheduling model are defined for the problem.An enhanced PICEA-g algorithm is proposed for solving the problem.The proposed algorithm is compared and tested on the generated instances.

Dynamic design of sales territories

We introduce the Multiple Traveling Salesmen and Districting Problem with Multi-periods and Multi-depots. In this problem, the compactness of the subdistricts, the dissimilarity measure of districts and an equity measure of salesmen profit are considered as part of the objective function, and the salesman travel cost on each subdistrict is approximated by the Beardwood-Halton-Hammersley formula. An adaptive large neighbourhood search metaheuristic is developed for the problem. It was tested on modified Solomon and Gehring & Homberger instances. Computational results confirm the effectiveness of the proposed metaheuristic.

The Vehicle Routing Problem with Stochastic Demands and Split Deliveries

Abstract This paper proposes a paired vehicle recourse policy for the Vehicle Routing Problem with Stochastic Demands and Split Deliveries (VRPSDSD). Most publications on the stochastic vehicle routing problem focus on recourse policies which assume that the vehicles operate independently of each other. An adaptive large neighborhood search heuristic is developed for this problem. Extensive computational experiments demonstrate that allowing split deliveries is sometimes beneficial, namely when expected demands lie between 51% and 70% of the vehicle capacity.

SGEESS: Smart green energy-efficient scheduling strategy with dynamic electricity price for data center

Nowadays, it becomes a major trend to use the green renewable energy in the data center when considering the environment protection and the energy crisis. To improve the energy efficiency and save the system cost, the computational tasks of data center should match to the renewable energy supply. This paper aims to develop a smart green energy-efficient scheduling strategy to increase utilization of renewable energy, reduce system running cost and improve the task satisfaction rate in a data center partially powered by the renewable energy. We first define three mathematical models, i.e., task model, energy model and scheduling model for the proposed problem. Then, a smart green energy-efficient scheduling strategy is proposed for the task scheduling in the data center, based on the renewable energy prediction and the dynamic grid electricity price. In the experiments, three other scheduling strategies, i.e., Green-Scheduling Strategy, Price-Scheduling Strategy and Greedy-Energy-Efficient Strategy, are provided for comparisons, a real-world trace of Google cloud trace is also tested. The experimental results confirm the superiority and effectiveness of the proposed scheduling strategy.

Solving resource availability cost problem in project scheduling by pseudo particle swarm optimization

This paper considers a project scheduling problem with the objective of minimizing resource availability costs appealed to finish all activities before the deadline. There are finish-start type precedence relations among the activities which require some kinds of renewable resources. We predigest the process of solving the resource availability cost problem (RACP) by using start time of each activity to code the schedule. Then, a novel heuristic algorithm is proposed to make the process of looking for the best solution efficiently. And then pseudo particle swarm optimization (PPSO) combined with PSO and path relinking procedure is presented to solve the RACP. Finally, comparative computational experiments are designed and the computational results show that the proposed method is very effective to solve RACP.

Optimal microgrid operation based on model predictive control framework

This paper investigates the appliance of model predictive control framework for efficiently operating of a microgrid, which comprises distributed energy resources, energy storage systems, local smart loads such as HVAC and PHEVs, and some other uncontrollable loads. It can be operated in grid-connected mode and isolated mode autogenously. Model predictive control framework allows to take into consideration demand response, uncertainty power production of renewable energy resources, load demand and the varying of real-time electricity price, as well as to satisfy special constraints. The microgrid is considered as mixed logical dynamic system, the operation optimization of the microrgid is formulated with a mixed integer quadratic programming problem. Case studies are implemented and discussed to assess the performance of the proposed approach and the impacts of energy storage units, simulation results validate the efficiency of the proposed method.

Optimizing charging and discharging on a micro-grid with ESS and dynamic price

This paper considers the scheduling issue of charging and discharging on a micro-grid with ESS and dynamic price, where the micro-grid consists of an energy management system, a photovoltaic system, an energy storage system, normal loads, electric vehicles and their charging piles. The mathematical formulation of the problem is defined based on a day-ahead design mode of scheduling. An efficient algorithm is developed for the model. In the simulation, two compared algorithms have been also designed, and simulation results validate the effectiveness and superiority of our proposed algorithm.