Research On Vehicle Scheduling Problem In Mines Based On Time Cost Optimization

Mine vehicle transportation is an important logistics link in the mine production process.The efficiency of vehicle transportation is inversely proportional to the completion time of operation.The higher the efficiency of vehicle transportation,the shorter the completion time of operation and the lower the time cost of transportation.On the contrary,if the transportation scheduling plan is unreasonable,the time of operation completion is prolonged,and the time cost of transportation is relatively increased.The mine vehicle transportation operation includes three parts:the traffic distribution plan between the mine shovel and the unloading point,the truck transportation task assignment,and the truck transportation route planning.Unreasonable traffic distribution will cause large deviation of the ore demand and grade.And the unreasonable transportation task assignment will reduce the truck's utilization ratio,or increase transport time.Unreasonable trucking routes can cause excessive collisions between trucks when completing specific transportation tasks,causing long waiting times for transportation,thus affecting overall time costs.Therefore,is a big challenge for us to make the transportation plan of mine vehicles to ensure safe production of mines and improve transportation efficiency and save logistics costs.This thesis establishes a mixed integer programming model for the optimization of the overall transportation time cost of trucks during transportation.A heuristic-based two-stage model combined with differential evolution algorithm is designed to solve the large-scale problem which the solver cannot solve.Aiming at the problem of dynamic traffic allocation,due to its dynamism and state complexity,this thesis designs a corresponding approximate dynamic programming algorithm to solve it,and compares the result with the general myopia strategy.The main research contents of this thesis are as follows:(1)Establish mathematical programming model based on time cost optimization.Based on the classical mine traffic planning model,this thesis get a traffic planning integer model aimed to minimum total load time cost.After adding a assumption that the truck will have loading and unloading conflict at the loading or at the unloading point,the time cost is defined as the weighted sum of the total cumulative full-load transportation time,the total empty-time transportation time,and the total queued time,this thesis attempts to minimize the total time cost.A mixed integer programming model was established.The experimental results verify the correctness of the model.(2)Due to the mixed integer programming model has large number of variables and complexity constraints,that only can solve very small scale problem.Based on the traditional staged modeling method of mine vehicle scheduling,and considering the complexity impact of queue waiting time cost in this problem,the two stage mathematical model is proposed.The first stage establishes an integer programming model to assign a fixed number of transportation tasks to each truck,and the second stage uses a heuristic-based differential evolution algorithm combined with a simulation program to determine the truck transportation task sequence with the lowest waiting time cost.The experimental results show that the method can quickly find a viable scheduling solution.(3)An approximate dynamic programming algorithm for dynamic traffic allocation problem is proposed.Based on the dynamics of the multi-stage mines traffic distribution problem and the dimensionality disaster problem,an approximation function based on the post-decision state is proposed.This thesis use approximate dynamic programming algorithm and myopia strategy to solve the problem separately.The results show that the multi-stage allocation strategy based on approximate dynamic programming algorithm is more reasonable than the myopia allocation strategy.(4)An mine truck dispatch decision support system module has been designed and developed,embedding the solution algorithm involved in this thesis.The system module can store and update the production status information of the mine,and generate reasonable traffic distribution plan and truck scheduling plan according to the current production status information,providing reference for decision makers.