A Chance-Constrained Vehicle Routing Problem for Wet Waste Collection and Transportation Considering Carbon Emissions

Abstract

In order to solve the optimization problem of wet waste collection and transportation in Chinese cities, this paper constructs a chance-constrained low-carbon vehicle routing problem (CCLCVRP) model in waste management system and applies certain algorithms to solve the model. Considering the environmental protection point of view, the CCLCVRP model combines carbon emission costs with traditional waste management costs under the scenario of application of smart bins. Taking into the uncertainty of the waste generation rate, chance-constrained programming is applied to transform the uncertain model to a certain one. The initial optimal solution of this model is obtained by a proposed hybrid algorithm, that is, particle swarm optimization (PSO); and then the further optimized solution is obtained by simulated annealing (SA) algorithm, due to its global optimization capability. The effectiveness of PSOSA algorithm is verified by the classic database in a capacitated vehicle routing problem (CVRP). What’s more, a case of waste collection and transportation is applied in the model for acquiring reliable conclusions, and the application of the model is tested by setting different waste fill levels (WFLs) and credibility levels. The results show that total costs rise with the increase of credibility level reflecting dispatcher’s risk preference; the WFL value range between 0.65 and 0.75 can obtain the optimal solution under different credibility levels. Finally, according to these results, some constructive proposals are propounded for the government and the logistics organization dealing with waste collection and transportation.

Designing of Wet Transformers as Artificial Soil for Flower Tree Cultivation

“…Responding data as the low carbon society concept idea: it focuses on the procedure design as the sustainable development to reduce the green gas effect from the transforming procedure of wet garbage as the artificial soil causing the increasing expenditure or gaining the new procedure as the result to promote the low carbon society (Shams, S., et al, 2017), as well as reducing the releasing of Co2 in transformation procedure suitably (da Graça Carvalho, M., et al, 2011;Wu, H., et al, 2020).…”

Optimization of Vehicle Routing for Waste Collection and Transportation

“…This paper randomly selects 8 cases from the database including small-scale study (Pro1, Pro2, Pro3, Pro4), medium-scale study (Pro5, Pro6, Pro7) and large-scale study (Pro8), the detailed information is shown in Table 5. Parameters of vehicles are shown in Table 6 according to references [21,50,51] and parameters of the proposed algorithm are shown in Table 7 according to references [10,[52][53][54].…”

“…Parameters of vehicles are shown in Table 6 according to references [ 21 , 50 , 51 ] and parameters of the proposed algorithm are shown in Table 7 according to references [ 10 , 52 , 53 , 54 ].…”

Quarantine Vehicle Scheduling for Transferring High-Risk Individuals in Epidemic Areas

“…Traditional exact methods are only applicable to small problem instances [5]. In recent years, considerable efforts have been devoted to evolutionary algorithms [6][7][8], including genetic algorithms (GA) [9][10][11][12][13], particle swarm optimization [14][15][16][17][18], ant colony optimization (ACO) [19][20][21][22][23], artificial bee colony [24][25][26], biogeography-based optimization [27,28], etc., for VRPs. Compared to exact algorithms and construction heuristics, evolutionary algorithms are more capable of jumping out of local optima and obtaining optimal or near-optimal solutions within an acceptable time by evolving a population of candidate solutions to simultaneously explore multiple regions of the search space [29].…”

“…The hybrid WWO and neighborhood search algorithm for the quarantine vehicle routing problem. 1 Use the procedures described in Section 4.1 to initialize a population P of solutions; 2 while the stopping condition is not met do 3 foreach solution Z = (X, Y) in the population do 4 Calculate the wavelength of Z according to Equation (17); 5 for d = 1 to N X do 6 Letd = rand(1, n X ); 7 Reverse the subsequence X[d,d]; 8 Call Algorithm 5 to improve the direction vector Y according to the new sequence vector X; 9 Call Algorithm 4 to improve the new solution Z by balancing the areas among the subsequences; 10 if Z is better than the original solution then 11 Replace the original solution with Z in the population; 12 if Z is the newly found best solution then 13 for i = 1 to m do 14 for k = 1 to αn i do 15 Clone Z to a new solution Z ; 16 Swap two randomly selected areas in x i ; 17 if Z is better than the Z then 18 Let Z = Z ; 19 Update the population size; 20 return the best solution found so far.…”