Method for optimizing traffic flow planning and scheduling of open-pit iron ore

Abstract

The invention provides an opencast iron ore traffic flow planning and scheduling optimization method, and relates to the technical field of opencast iron ore production operation. The method comprises the following steps of: firstly, pre-estimating the quantity of dispatched trucks according to a daily planned ore blending production proportion given by an ore dressing process department, and establishing a multi-target optimization model taking the minimum total transportation cost, the maximum ore transportation volume, the minimum rock transportation volume and the minimum total idle time as targets; providing an improved MOPSO algorithm with an adaptive weight strategy, a congestion degree sorting strategy, a local search mechanism and a constraint processing mechanism, and solving a traffic flow planning and scheduling model; and correcting daily plan actual yield and grade according to an optimization result, and providing a reference basis for scheduling operation decision. According to the method, multiple targets and complex constraints in the opencast iron ore traffic flow planning and scheduling problem can be effectively processed, and the solving efficiency and the scheduling quality are improved; the method is helpful for a dispatcher to evaluate the feasibility and effect of the dispatching scheme before actual execution, and improves the production efficiency and saves the production cost on the premise of meeting the production yield and grade requirements.

Description

Technical Field

[0001] The invention relates to the technical field of open-pit iron ore production operations, and in particular to an open-pit iron ore vehicle flow planning and scheduling optimization method. Background Art

[0002] Open-pit iron ore production is a production activity based on mining and linked by transportation, integrating mining, transportation and unloading. The transportation of ore and rock is the key link connecting this series of production activities. Open-pit iron ore mainly relies on mining trucks to complete the loading, unloading and transportation of ore and rock excavated by electric shovels. An efficient and reasonable mining truck dispatching plan is conducive to saving transportation costs and improving the economic benefits of mining enterprises.

[0003] In the prior art, open-pit mining enterprises use the dispatching system to dispatch mining trucks to complete production tasks, but the mining truck planning and dispatching optimization methods adopted by the existing systems do not take into account the unique operating conditions of open-pit iron mines, making the existing methods inapplicable. First, there are ore crushing stations and rock crushing stations in open-pit iron mines, and the production process requires that the grade and output of the two be focused on and constrained. Secondly, in actual production, it is necessary to determine the number of trucks required for each line according to the requirements of the ore distribution ratio, and then formulate a detailed dispatching plan to ensure the feasibility of dispatching. Finally, the actual output and grade obtained daily after the open-pit mine truck dispatching plan is implemented will deviate from the daily plan, resulting in data distortion. From the above, it can be seen that the existing methods do not take into account the particularity of open-pit iron mines, resulting in inaccurate models and weak practicality of the obtained solutions, which makes it difficult to provide strong support for the actual production of open-pit iron mines. In addition, multiple types of actual features lead to complex models, and the existing algorithms are difficult to solve efficiently and the planning quality is not high. Therefore, there is an urgent need for an open-pit iron ore traffic planning and scheduling optimization method that is more targeted and closer to actual needs, which fully considers practical factors such as the constraints of ore crushing stations and rock crushing stations, ore and vehicle matching processes and plan deviation corrections, so as to achieve an effective and high-quality solution to the open-pit iron ore truck planning and scheduling problem. Summary of the invention

[0004] The present invention provides an open-pit iron mine traffic flow planning and scheduling optimization method, which comprehensively considers the constraints of ore crushing stations and rock crushing stations inside the open-pit iron mine, the demand for ore and vehicle matching, and the current status of output grade deviation, and designs an improved MOPSO algorithm combining an adaptive weight strategy, a congestion sorting strategy, a local search mechanism, and a constraint processing mechanism, so as to better guide the actual production of the open-pit iron mine.

[0005] In order to solve the above technical problems, the technical solution adopted by the present invention is:

[0006] A method for optimizing open pit iron mine traffic flow planning and scheduling, comprising the following steps:

[0007] Step 1: Determine the truck usage for daily scheduling according to the reference truck configuration ratio for each ore loading area in the daily plan given by the ore dressing process department. Minimize the truck usage to save costs while meeting the production plan's requirements for output and grade;

[0008] Step 2: Establish the objective function of the multi-objective traffic flow planning and scheduling model for open-pit iron ore; the established model takes the minimum total transportation cost, the maximum ore output, and the minimum total idle time as the objective function;

[0009] Step 3: Set the constraints of the multi-objective traffic flow planning and scheduling model for open-pit iron ore, including the ore grade constraint of the ore crushing station, the output plan constraint of the ore crushing station, the production capacity constraint of the electric shovel loading point, the traffic continuity constraint, the remaining oil constraint and the truck transportation number constraint;

[0010] Step 4: Based on the truck usage calculated in step 1 and the given number of shovel loading areas on the day, combined with the objective functions and constraints of steps 2 and 3, a multi-objective intelligent optimization algorithm is applied to solve the model to obtain the Pareto optimal solution set;

[0011] Step 5: According to the Pareto optimal solution set obtained in step 4, select an optimal solution as the final scheduling plan. Based on the truck scheduling operation results provided by the selected plan, revise the actual daily planned output and grade for reference decision-making of scheduling operations.

[0012] Furthermore, the step 1 specifically includes:

[0013] Step 1.1: Calculate the maximum number of times n that a single truck can load and unload ore per day at the shovel ore loading area i∈(1,p) r1 , as follows:

[0014]

[0015] Among them, i∈(1,p) is the shovel ore loading area, i∈(p+1,I) is the shovel rock loading area; the daily plan includes two shifts, day shift and night shift, and the working time of each shift is T l ; T i1 is the average time taken by trucks for a loading and unloading cycle in the shovel ore loading area, d i1 is the optimal distance for trucks from ore loading point i to the ore crushing station, s f is the average speed of the truck when it is heavily loaded, s n is the average speed of the truck when it is unloaded, T z is the average time taken to load a truck, T q is the average time taken for truck unloading; γ is the control coefficient, which is adjusted according to the actual daily planned suspension or maintenance time of the truck; T c It is the time for the truck to be serviced;

[0016] Here, we consider that the model of each truck is the same, that is, the loading capacity is the same; the ore processing department roughly estimates the required ore volume based on the required ore grade and the ore type and grade of each shovel loading area, which is used as a reference for the daily operation plan of the production scheduling room. However, there is a large error between the estimated ore output and the actual production volume;

[0017] Step 1.2: Reference truck configuration ratio m according to the daily plan 1 :m 2 :...:m p , calculate the daily planned number of vehicles z for each shovel ore loading area i∈(1,p) i1 , round up, the formula is as follows:

[0018]

[0019] If the ratio of shovel to truck is 1:2 or 1:3, then z i1 ={2,3}, and m i is the number of times trucks load and unload ore for the ore produced in the shovel loading zone i;

[0020] Step 1.3: Calculate the maximum number of times n that a single truck can load and unload ore per day at the shovel rock loading area i∈(p+1,I) r2 , the formula is as follows:

[0021]

[0022] Among them, T i2 is the average time taken by a truck to load and unload a rock in the shovel loading area, d i2 is the optimal distance for trucks from rock loading point i to the rock crushing station;

[0023] Step 1.4: Calculate the daily planned vehicle number z for each shovel rock loading area i∈(p+1,I) i2 ; Since the mineral processing department only gives the planned ratio of ore demand, the rock demand is evenly distributed according to the annual plan, with slight adjustments made every day, which are given by the production scheduling room. Therefore, it is only necessary to meet the target given by the production scheduling room. The formula is as follows:

[0024]

[0025] If the ratio of electric shovel to truck is 1:2 or 1:3, then z i2 ={2,3}, and f 2 is the minimum production requirement of the rock unloading point within one shift, C r is the load of the truck;

[0026] Step 1.5: Calculate the number of trucks to be deployed per day:

[0027]

[0028] Furthermore, the objective function of step 2 specifically includes:

[0029] The objectives of the truck dispatch daily plan include minimizing truck transportation costs, maximizing ore production, and minimizing the total idle time of trucks; the model objective function is:

[0030] F(S)={F 1 (S), -F 2 (S), F 3 (S)};

[0031] Among them, F 1 (S), F 2 (S), F 3 (S) are the sub-objectives in the objective function, representing minimization of total truck transportation cost, maximization of ore output, and minimization of total truck idle time;

[0032] Truck transportation costs include three parts: the operating cost of a heavily loaded truck from the loading point to the unloading point, the operating cost of an empty truck from the unloading point to the loading point or to and from the refueling point, and the maintenance cost of the truck; the objective function expression for minimizing the total truck transportation cost is:

[0033]

[0034] The objective function of maximizing ore production is specifically expressed as maximizing ore transportation volume while minimizing rock transportation volume, and its expression is:

[0035]

[0036] and converts to:

[0037]

[0038] The idle time of a single truck is the difference between the daily planned working time of each shift and the necessary driving, loading, unloading, refueling and inspection time. The objective function expression for minimizing the total idle time of the truck is:

[0039]

[0040] Where i is the loading point index, indicating the i-th loading point, i.e., the position of the electric shovel, i=1,2,...,I; j is the unloading point index, indicating the j-th unloading point, i.e., the position of the crushing station, j=1,2,...,J; r is the truck index, indicating the r-th truck, r=1,2,...,R; xrij is the number of times the rth truck goes from loading point i to unloading point j, x rji is the number of times the rth truck goes from unloading point j to loading point i, d i,j The optimal distance from loading point i to unloading point j, d jO The optimal distance from unloading point j to refueling point O, d Oi The optimal distance from refueling point O to loading point i, C r1 is the unit distance cost of the rth truck when it is heavily loaded, C r2 is the unit distance cost when the rth truck is empty, C r3 is the wear and tear cost per unit time of the r-th truck, i.e., the maintenance cost, f 1 is the minimum production requirement of the ore unloading point within one shift, T O Average time to refuel a truck, K rjO is the number of times truck r runs from unloading point j to refueling point O, K rOi is the number of trips of truck r from refueling point O to loading point i.

[0041] Furthermore, the constraints in step 3 specifically include:

[0042] Ore grade constraint of ore crushing station: The deviation between the actual grade of the daily plan and the given grade shall not exceed the maximum grade deviation β, as shown in the following formula:

[0043]

[0044] Among them, α i is the ore grade at loading point i, β is the allowable deviation of ore grade, G 1 is the target grade of the ore crushing station;

[0045] Ore crushing station production plan constraint: The total amount of ore unloaded by the ore crushing station must meet the minimum demand of each shift in the ore production daily plan. 1 The total amount of rock unloaded by the rock crushing station must meet the minimum demand of each shift in the daily rock production plan. 2 , as follows:

[0046]

[0047]

[0048] The total amount of ore loaded at each shovel loading point must not exceed the maximum production capacity of the loading point. i , as follows:

[0049]

[0050] Traffic flow continuity constraint: The traffic flow in and out of the loading / unloading point must be equal, as shown in the following formula:

[0051]

[0052] Remaining fuel constraint: It is necessary to monitor the remaining fuel of each truck. When it drops to the minimum value A, it is necessary to refuel at a refueling point. The remaining fuel A = the maximum fuel consumption from the unloading point to the loading point + the maximum fuel consumption from the loading point to the unloading point + the maximum fuel consumption from the unloading point to the refueling point. The constraint formula is as follows:

[0053]

[0054] Among them, E is the fuel tank capacity of the truck, A is the minimum remaining fuel of the truck, and E r1 is the fuel consumption per unit distance of the r-th truck when it is heavily loaded, E r2 is the fuel consumption per unit distance of the rth truck when it is unloaded;

[0055] Truck transport times constraint: The truck transport times must be a positive integer, and when the ore and rock are fully loaded, they must go to the ore and rock unloading points for unloading, while the empty trucks at the unloading points can go to any loading points; for the traffic planning and scheduling problem of mixed ore and rock transportation, let the first p of I shovel loading points be ore loading points and the last Ip be rock loading points, let the first of J unloading points be an ore crushing station and the second be a rock crushing station, the constraint formula is as follows:

[0056] x rij ∈{0,1,2...},x rji ∈{0,1,2...};

[0057] in,

[0058] Furthermore, the multi-objective intelligent optimization algorithm in step 4 adopts an improved multi-objective particle swarm optimization algorithm to solve the multi-objective optimization problem of open-pit iron ore traffic planning and scheduling; the algorithm introduces a number of improvements on the basis of the standard multi-objective particle swarm optimization algorithm to improve its performance in dealing with complex constraints and multi-objective optimization problems; the specific steps are as follows:

[0059] Step 4.1: Particle encoding and initialization;

[0060] Each particle represents a feasible scheduling solution, encoded as a multidimensional vector X = (x 1 ,x 2 ,...,x n ), where n is the number of decision variables; during initialization, the particle positions are randomly generated while satisfying the constraints;

[0061] Step 4.2: Based on the objective function established in step 2, define the multi-objective fitness function: F(S) = {F 1 (S), -F 2 (S), F 3 (S)}, where F 1 (S) represents the objective of minimizing the total truck transportation cost, F 2 (S) represents the goal of maximizing ore production, F 3 (S) represents the objective of minimizing the total idle time of trucks;

[0062] Step 4.3: Introduce the adaptive weight strategy, the particle speed and position update formula is as follows:

[0063]

[0064] Among them, v id (t+1) is the velocity of the ith particle in the dth dimension at time t+1; v id t is the velocity of the ith particle in the dth dimension at time t; c 1 is the individual learning factor, which determines the intensity of the tendency of the particle to move to the individual historical optimal position; c 2 is the social learning factor, which determines the intensity of the tendency of particles to move toward the historical optimal position of the group; r 1 、r 2 is a random number between [0,1], which is regenerated in each iteration to increase the randomness of the search; p id t is the individual historical optimal position of the ith particle in the dth dimension, i.e., the individual optimal solution; p gd t is the historical optimal position of the group in the dth dimension, i.e., the global optimal solution, x id t is the current position of the ith particle in the dth dimension at time t, x id (t+1) is the current position of the ith particle in the dth dimension at time t+1; w(t) is the adaptive inertia weight, which is dynamically adjusted by the following formula:

[0065] w(t)=w max -(w max -w min )*(t / T max ) 2 ;

[0066] Among them, w max and w min are the maximum and minimum values ​​of the inertia weight, t is the current iteration number, T max is the maximum number of iterations;

[0067] Step 4.4: Maintenance of non-dominated solution set;

[0068] The fast non-dominated sorting algorithm is used to sort the particles, and the crowding distance is used to maintain the diversity of the non-dominated solution set; the crowding distance calculation formula is as follows:

[0069]

[0070] Among them, CD i is the crowding distance of the ith solution, M is the number of objective functions, and are the function values ​​of the adjacent solutions of the mth objective function after sorting, f m max and f m min are the maximum and minimum values ​​of the mth objective function respectively;

[0071] Step 4.5: Local search mechanism;

[0072] After each iteration, the simulated annealing strategy is used to perform local search on some solutions in the non-dominated solution set. In order to facilitate the algorithm to jump out of the local optimal solution during local search, it is allowed to accept a worse solution according to a certain probability, that is, the acceptance probability. The calculation formula for the acceptance probability is:

[0073] P(ΔE,T)=exp(-ΔE / T);

[0074] Among them, ΔE is the difference between the objective function of the new solution and the current solution, and T is the current temperature, which gradually decreases with the number of iterations;

[0075] Step 4.6: Constraint handling mechanism;

[0076] Use the improved constraint governance rule to handle constraints and define constraint violation:

[0077]

[0078] Among them, g q (X)≤0 is an inequality constraint, q=1,2,...,Q; h k (X) = 0 is an equality constraint, k = 1, 2, ..., K; when comparing two solutions, their constraint violation degrees are compared first, and a solution with a smaller constraint violation degree is better than a solution with a larger constraint violation degree.

[0079] Step 4.7: Algorithm termination condition;

[0080] The algorithm iterates until any of the following conditions is met: the preset maximum number of iterations is reached, and the non-dominated solution set does not change after N consecutive iterations.

[0081] Furthermore, the details of step 5 are as follows:

[0082] According to the Pareto optimal solution set obtained in step 4, the TOPSIS (Technique for Order Preference by Similarity to an Ideal Solution) method is used to select an optimal solution from the final non-dominated solution set as the final scheduling solution; the TOPSIS method is based on the following steps:

[0083] Construct a normalized decision matrix: Calculate a weighted normalized decision matrix; determine the ideal solution (the hypothetical solution where each indicator achieves the best value) and the negative ideal solution (the hypothetical solution where each indicator achieves the worst value); calculate the distance of each solution to the ideal solution and the negative ideal solution; calculate the relative proximity of each solution; sort the solutions according to the relative proximity;

[0084] Compute the distance of each solution to the ideal solution and the negative ideal solution:

[0085] Let the weighted normalized value of the i-th solution on the j-th indicator be v ij , the ideal solution value at the jth index is v j+ , the value of the negative ideal solution at the jth index is v j- , then the distance S from the i-th solution to the ideal solution i+ and to the negative ideal solution S i- The distance can be calculated using the Euclidean distance:

[0086]

[0087] Among them, n is the number of evaluation indicators.

[0088] Compute the relative closeness of each solution:

[0089] The relative closeness C of the i-th solution i It can be calculated using the following formula:

[0090]

[0091] Relative proximity C i The value range of is [0,1], and the closer it is to 1, the better the solution.

[0092] Sort the solutions by relative proximity:

[0093] All the solutions are sorted according to their relative closeness C i Sort from largest to smallest, C i The solution with the largest value is the optimal solution.

[0094] According to the truck dispatching operation results of the selected decision-making scheme, the actual daily planned output and grade are revised for reference decision-making of dispatching operations. The revised daily planned ore output, rock output and grade formulas are as follows:

[0095]

[0096] in, is the revised daily planned ore production, is the corrected rock yield, is the corrected grade, α i is the ore grade at loading point i.

[0097] The beneficial effects of adopting the above technical scheme are as follows: the open-pit iron ore traffic planning and scheduling optimization method provided by the present invention gives priority to the problem of ore matching and vehicle matching, determines the truck usage according to the ore matching requirements of the ore dressing process, and then optimizes the traffic scheduling, which conforms to the logical sequence of actual production decisions of the open-pit iron ore and is operational; constructs a multi-objective optimization model, comprehensively considers multiple optimization goals such as transportation cost, ore output, rock output, idle time, and sets practical constraints, so that the optimization result is more comprehensive and reliable, and balances multiple factors such as cost and output; efficiently solves the multi-objective optimization model through an intelligent optimization algorithm, overcomes the defect that traditional scheduling methods are difficult to deal with complex problems with multiple objectives and multiple constraints, and greatly improves the efficiency and effect of traffic scheduling optimization; uses the optimized scheduling plan to correct the actual output and grade, provides an important basis for formulating a reliable daily plan scheduling strategy, and improves the reference and guiding significance of the traffic planning and scheduling plan. BRIEF DESCRIPTION OF THE DRAWINGS

[0098] Figure 1 A schematic diagram of a flow chart of a method for optimizing open pit iron mine traffic flow planning and scheduling provided by an embodiment of the present invention;

[0099] Figure 2 A flow chart of an improved multi-objective particle swarm optimization algorithm provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0100] The specific implementation of the present invention is further described in detail below in conjunction with the accompanying drawings and examples. The following examples are used to illustrate the present invention, but are not intended to limit the scope of the present invention.

[0101] like Figure 1 As shown, the method of this embodiment is described as follows.

[0102] A method for optimizing open pit iron mine traffic flow planning and scheduling, comprising the following steps:

[0103] Step 1: Determine the truck usage for daily scheduling based on the reference truck configuration ratio for each ore loading area in the daily plan given by the ore dressing process department. Under the premise of meeting the production plan's requirements for output and grade, minimize the truck usage to save costs. Specifically include:

[0104] Step 1.1: Calculate the maximum number of times n that a single truck can load and unload ore per day at the shovel ore loading area i∈(1,p) r1 , as follows:

[0105]

[0106] Among them, i∈(1,p) is the shovel ore loading area, i∈(p+1,I) is the shovel rock loading area; the daily plan includes two shifts, day shift and night shift, and the working time of each shift is T l ; T i1 is the average time taken by trucks for a loading and unloading cycle in the shovel ore loading area, d i1 is the optimal distance for trucks from ore loading point i to the ore crushing station, s f is the average speed of the truck when it is heavily loaded, s n is the average speed of the truck when it is unloaded, T z is the average time taken to load a truck, T q is the average time taken for truck unloading; γ is the control coefficient, which is adjusted according to the actual daily planned suspension or maintenance time of the truck; T c Time for a truck overhaul.

[0107] Here we consider that each truck has the same model and loading capacity. The mineral processing department roughly estimates the required ore volume based on the required ore grade and the ore type and grade of each shovel loading area, which serves as a reference for the daily operation plan of the production scheduling room. However, there is a large error between the estimated ore output and the actual production volume.

[0108] Step 1.2: Reference truck configuration ratio m according to the daily plan 1 :m 2 :...:m p , calculate the daily planned number of vehicles z for each shovel ore loading area i∈(1,p) i1 , round up, the formula is as follows:

[0109]

[0110] If the ratio of shovel to truck is 1:2 or 1:3, then z i1 ={2,3}, and m i The number of times trucks load and unload ore is the reference number of ore produced in the shovel loading area i.

[0111] Step 1.3: Calculate the maximum number of times n that a single truck can load and unload ore per day at the shovel rock loading area i∈(p+1,I) r2 , the formula is as follows:

[0112]

[0113] Among them, T i2 is the average time taken by a truck to load and unload a rock in the shovel loading area, d i2 is the optimal distance for the truck from the rock loading point i to the rock crushing station.

[0114] Step 1.4: Calculate the daily planned vehicle number z for each shovel rock loading area i∈(p+1,I) i2 ; Since the mineral processing department only gives the planned ratio of ore demand, the rock demand is evenly distributed according to the annual plan, with slight adjustments made every day, which are given by the production scheduling room. Therefore, it is only necessary to meet the target given by the production scheduling room. The formula is as follows:

[0115]

[0116] If the ratio of electric shovel to truck is 1:2 or 1:3, then z i2 ={2,3}, and f 2 is the minimum production requirement of the rock unloading point within one shift, C r is the load capacity of the truck.

[0117] Step 1.5: Calculate the number of trucks to be deployed per day:

[0118]

[0119] Step 2: Establish the objective function of the multi-objective traffic flow planning and scheduling model for open-pit iron ore. The model established by the present invention takes the minimum total transportation cost, the maximum ore output, and the minimum total idle time as the objective function, which is as follows:

[0120] The objectives of the truck dispatch daily plan include minimizing truck transportation costs, maximizing ore production, and minimizing the total idle time of trucks. The model objective function is:

[0121] F(S)={F 1 (S), -F 2 (S), F 3 (S)};

[0122] Among them, F 1 (S), F 2 (S), F 3(S) are the sub-objectives in the objective function, which respectively represent minimization of the total truck transportation cost, maximization of ore output, and minimization of the total truck idle time.

[0123] Truck transportation costs include three parts: the operating cost of a heavily loaded truck from the loading point to the unloading point, the operating cost of an empty truck from the unloading point to the loading point or to and from the refueling point, and the maintenance cost of the truck. The objective function expression for minimizing the total truck transportation cost is:

[0124]

[0125] The objective function of maximizing ore production is specifically expressed as maximizing ore transportation volume while minimizing rock transportation volume, and its expression is:

[0126]

[0127] and converts to:

[0128]

[0129] The idle time of a single truck is the difference between the daily planned working time of each shift and the necessary driving, loading, unloading, refueling and inspection time. The objective function expression for minimizing the total idle time of the truck is:

[0130]

[0131] Where i is the loading point index, indicating the i-th loading point, i.e., the position of the electric shovel, i=1,2,...,I; j is the unloading point index, indicating the j-th unloading point, i.e., the position of the crushing station, j=1,2,...,J; r is the truck index, indicating the r-th truck, r=1,2,...,R; x rij is the number of times the rth truck goes from loading point i to unloading point j, x rji is the number of times the rth truck goes from unloading point j to loading point i, d i,j The optimal distance from loading point i to unloading point j, d jO The optimal distance from unloading point j to refueling point O, d Oi The optimal distance from refueling point O to loading point i, C r1 is the unit distance cost of the rth truck when it is heavily loaded, C r2 is the unit distance cost when the rth truck is empty, C r3 is the wear and tear cost per unit time of the r-th truck, i.e., the maintenance cost, f 1 is the minimum production requirement of the ore unloading point within one shift, T O Average time to refuel a truck, K rjO is the number of times truck r runs from unloading point j to refueling point O, K rOiis the number of trips of truck r from refueling point O to loading point i.

[0132] Step 3: Set the constraints of the multi-objective traffic flow planning and scheduling model for open-pit iron ore, including the ore grade constraint of the ore crushing station, the output plan constraint of the ore crushing station, the production capacity constraint of the electric shovel loading point, the traffic continuity constraint, the remaining oil constraint and the truck transportation frequency constraint. Specifically include:

[0133] Ore grade constraint of ore crushing station: The deviation between the actual grade of the daily plan and the given grade shall not exceed the maximum grade deviation β, as shown in the following formula:

[0134]

[0135] Among them, α i is the ore grade at loading point i, β is the allowable deviation of ore grade, G 1 It is the target grade of the ore crushing station.

[0136] Ore crushing station production plan constraint: The total amount of ore unloaded by the ore crushing station must meet the minimum demand of each shift in the ore production daily plan. 1 The total amount of rock unloaded by the rock crushing station must meet the minimum demand of each shift in the daily rock production plan. 2 , as follows:

[0137]

[0138] The total amount of ore loaded at each shovel loading point must not exceed the maximum production capacity of the loading point. i , as follows:

[0139]

[0140] Traffic flow continuity constraint: The traffic flow in and out of the loading / unloading point must be equal, as shown in the following formula:

[0141]

[0142] Remaining fuel constraint: It is necessary to monitor the remaining fuel of each truck. When it drops to the minimum value A, it is necessary to refuel at a refueling point. The remaining fuel A = the maximum fuel consumption from the unloading point to the loading point + the maximum fuel consumption from the loading point to the unloading point + the maximum fuel consumption from the unloading point to the refueling point. The constraint formula is as follows:

[0143]

[0144] Among them, E is the fuel tank capacity of the truck, A is the minimum remaining fuel of the truck, and E r1 is the fuel consumption per unit distance of the r-th truck when it is heavily loaded, E r2is the fuel consumption per unit distance of the rth truck when it is unloaded.

[0145] Truck transport times constraint: The truck transport times must be a positive integer, and when the ore and rock are fully loaded, they must go to the ore and rock unloading points for unloading, while the empty trucks at the unloading points can go to any loading points; for the traffic planning and scheduling problem of mixed ore and rock transportation, let the first p of I shovel loading points be ore loading points and the last Ip be rock loading points, let the first of J unloading points be an ore crushing station and the second be a rock crushing station, the constraint formula is as follows:

[0146] x rij ∈{0,1,2...},x rji ∈{0,1,2...};

[0147] in,

[0148] Step 4: Based on the truck usage calculated in step 1 and the given number of shovel loading areas on the day, combined with the objective functions and constraints of steps 2 and 3, the multi-objective intelligent optimization algorithm is used to solve the model and obtain the Pareto optimal solution set. The multi-objective intelligent optimization algorithm uses an improved multi-objective particle swarm optimization algorithm (MOPSO) to solve the multi-objective optimization problem of open-pit iron ore traffic planning and scheduling; the algorithm introduces a number of improvements on the basis of the standard multi-objective particle swarm optimization algorithm MOPSO to improve its performance in dealing with complex constraints and multi-objective optimization problems. Figure 2 As shown, the specific steps are as follows:

[0149] Step 4.1: Particle encoding and initialization;

[0150] Each particle represents a feasible scheduling solution, encoded as a multidimensional vector X = (x 1 ,x 2 ,...,x n ), where n is the number of decision variables; during initialization, the particle positions are randomly generated while satisfying the constraints.

[0151] Step 4.2: Based on the objective function established in step 2, define the multi-objective fitness function: F(S) = {F 1 (S), -F 2 (S), F 3 (S)}, where F 1 (S) represents the objective of minimizing the total truck transportation cost, F 2 (S) represents the goal of maximizing ore production, F 3 (S) represents the objective of minimizing the total idle time of trucks.

[0152] Step 4.3: Introduce the adaptive weight strategy, the particle speed and position update formula is as follows:

[0153]

[0154] Among them, v id (t+1) is the velocity of the ith particle in the dth dimension at time t+1; v id t is the velocity of the ith particle in the dth dimension at time t; c 1 is the individual learning factor, which determines the intensity of the tendency of the particle to move to the individual historical optimal position; c 2 is the social learning factor, which determines the intensity of the tendency of particles to move toward the historical optimal position of the group; r 1 、r 2 is a random number between [0,1], which is regenerated in each iteration to increase the randomness of the search; p id t is the individual historical optimal position of the ith particle in the dth dimension, i.e., the individual optimal solution; p gd t is the historical optimal position of the group in the dth dimension, i.e., the global optimal solution, x id t is the current position of the ith particle in the dth dimension at time t, x id (t+1) is the current position of the ith particle in the dth dimension at time t+1; w(t) is the adaptive inertia weight, which is dynamically adjusted by the following formula:

[0155] w(t)=w max -(w max -w min )*(t / T max ) 2 ;

[0156] Among them, w max and w min are the maximum and minimum values ​​of the inertia weight, t is the current iteration number, T max is the maximum number of iterations.

[0157] Step 4.4: Maintenance of non-dominated solution set.

[0158] The fast non-dominated sorting algorithm is used to sort the particles, and the crowding distance is used to maintain the diversity of the non-dominated solution set; the crowding distance calculation formula is as follows:

[0159]

[0160] Among them, CD iis the crowding distance of the ith solution, M is the number of objective functions, and are the function values ​​of the adjacent solutions of the mth objective function after sorting, f m max and f m min are the maximum and minimum values ​​of the mth objective function respectively.

[0161] Step 4.5: Local search mechanism.

[0162] After each iteration, the simulated annealing strategy is used to perform local search on some solutions in the non-dominated solution set. In order to facilitate the algorithm to jump out of the local optimal solution during local search, it is allowed to accept a worse solution according to a certain probability, that is, the acceptance probability. The calculation formula for the acceptance probability is:

[0163] P(ΔE,T)=exp(-ΔE / T);

[0164] Among them, ΔE is the difference between the objective function of the new solution and the current solution, and T is the current temperature, which gradually decreases with the number of iterations.

[0165] Step 4.6: Constraint handling mechanism.

[0166] Use the improved constraint governance rule to handle constraints and define constraint violation:

[0167]

[0168] Among them, g q (X)≤0 is an inequality constraint, q=1,2,...,Q; h k (X) = 0 is an equality constraint, k = 1, 2, ..., K; when comparing two solutions, first compare their constraint violation degrees, and the solution with a smaller constraint violation degree is better than the solution with a larger constraint violation degree. When the constraint violation degrees are equal, then compare the objective function values.

[0169] Step 4.7: Algorithm termination condition.

[0170] The algorithm iterates until any of the following conditions is met: the preset maximum number of iterations is reached, and the non-dominated solution set does not change after N consecutive iterations.

[0171] Step 5: According to the Pareto optimal solution set obtained in step 4, select an optimal solution as the final scheduling plan. Based on the truck scheduling operation results provided by the selected plan, revise the actual daily planned output and grade for reference decision-making of scheduling operations.

[0172] Since the optimization model is a multi-objective model, a set of complementary Pareto optimal solutions will be obtained in the process of solving the model. These solutions all meet the constraints in the model but correspond to different objective function values. The three objectives in the model are conflicting to a certain extent.

[0173] According to the Pareto optimal solution set obtained in step 4, the TOPSIS (Technique for Order Preference by Similarity to an Ideal Solution) method is used to select an optimal solution from the final non-dominated solution set as the final scheduling solution; the TOPSIS method is based on the following steps:

[0174] Construct a normalized decision matrix: Calculate the weighted normalized decision matrix; determine the ideal solution (the hypothetical solution in which each indicator achieves the best value) and the negative ideal solution (the hypothetical solution in which each indicator achieves the worst value); calculate the distance of each plan to the ideal solution and the negative ideal solution; calculate the relative proximity of each plan; and sort the plans according to their relative proximity.

[0175] The method for calculating the distance of each solution to the ideal solution and the negative ideal solution is:

[0176] Let the weighted normalized value of the i-th solution on the j-th indicator be v ij , the ideal solution value at the jth index is v j+ , the value of the negative ideal solution at the jth index is v j- , then the distance S from the i-th solution to the ideal solution i+ and to the negative ideal solution S i- The distance can be calculated using the Euclidean distance:

[0177]

[0178] Among them, n is the number of evaluation indicators.

[0179] The relative closeness of each solution is calculated as:

[0180] The relative closeness C of the i-th solution i It can be calculated using the following formula:

[0181]

[0182] Relative proximity C i The value range of is [0,1], and the closer it is to 1, the better the solution.

[0183] The method to sort the solutions according to their relative closeness is:

[0184] All the solutions are sorted according to their relative closeness Ci Sort from largest to smallest, C i The solution with the largest value is the optimal solution.

[0185] According to the truck dispatching operation results of the selected decision-making scheme, the actual daily planned output and grade are revised for reference decision-making of dispatching operations. The revised daily planned ore output, rock output and grade formulas are as follows:

[0186]

[0187] in, is the revised daily planned ore production, is the corrected rock yield, is the corrected grade, α i is the ore grade at loading point i.

[0188] This embodiment proposes a method for open-pit iron ore traffic planning and scheduling that first considers ore matching and then performs multi-objective optimization. This method first estimates the truck usage based on the daily plan reference ore matching ratio given by the ore dressing process department, and then establishes a multi-objective optimization model with the goals of minimizing total transportation cost, maximizing ore output, and minimizing total idle time. This method is more in line with actual production needs and can achieve cost savings under the premise of ensuring production plans. This embodiment uses an improved multi-objective particle swarm optimization algorithm (MOPSO) to solve the traffic planning and scheduling model. The algorithm introduces an adaptive weight strategy, a congestion sorting strategy, and a local search mechanism, and designs a special constraint processing mechanism. This improved algorithm can effectively handle multiple objectives and complex constraints in the open-pit iron ore traffic planning and scheduling problem, and improves the efficiency and quality of the solution. The method of this embodiment can correct the actual output and grade of the daily plan according to the optimization results, and provide a reference basis for scheduling operation decisions. This correction function enables the dispatcher to evaluate the feasibility and effect of the scheduling plan before actual execution, which helps to improve the accuracy of decision-making and production efficiency.

[0189] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the above embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the above embodiments, or replace some or all of the technical features therein with equivalents. However, these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the scope defined by the claims of the present invention.

Claims

1. A method for optimizing open pit iron mine traffic flow planning and scheduling, characterized in that: The following steps are involved: Step 1: Determine the truck usage for daily scheduling according to the reference truck configuration ratio for each ore loading area in the daily plan given by the ore dressing process department; minimize the truck usage to save costs while meeting the production plan's requirements for output and grade; Step 2: Establish the objective function of the multi-objective traffic flow planning and scheduling model for open-pit iron ore; the established model takes the minimum total transportation cost, the maximum ore output, and the minimum total idle time as the objective function; Step 3: Set the constraints of the multi-objective traffic flow planning and scheduling model for open-pit iron ore, including the ore grade constraint of the ore crushing station, the output plan constraint of the ore crushing station, the production capacity constraint of the electric shovel loading point, the traffic continuity constraint, the remaining oil constraint and the truck transportation number constraint; Step 4: Based on the truck usage calculated in step 1 and the given number of shovel loading areas on the day, combined with the objective functions and constraints of steps 2 and 3, a multi-objective intelligent optimization algorithm is applied to solve the model to obtain the Pareto optimal solution set; Step 5: According to the Pareto optimal solution set obtained in step 4, select an optimal solution as the final scheduling plan. Based on the truck scheduling operation results provided by the selected plan, revise the actual daily planned output and grade for reference decision-making of scheduling operations.

2. The open-pit iron ore traffic planning and scheduling optimization method according to claim 1 is characterized in that: The step 1 specifically includes: Step 1.1: Calculate the maximum number of times n that a single truck can load and unload ore per day at the shovel ore loading area i∈(1,p) r1 , as follows: Among them, i∈(1,p) is the shovel ore loading area, i∈(p+1,I) is the shovel rock loading area; the daily plan includes two shifts, day shift and night shift, and the working time of each shift is T l ; T i1 is the average time taken by trucks for a loading and unloading cycle in the shovel ore loading area, d i1 is the optimal distance for trucks from ore loading point i to the ore crushing station, s f is the average speed of the truck when it is heavily loaded, s n is the average speed of the truck when it is unloaded, T z is the average time taken to load a truck, T q is the average time taken for truck unloading; γ is the control coefficient, which is adjusted according to the actual daily planned suspension or maintenance time of the truck; T c It is the time for the truck to be overhauled; Step 1.2: Refer to the truck configuration ratio m1:m2:...:m according to the daily plan p , calculate the daily planned number of vehicles z for each shovel ore loading area i∈(1,p) i1 , round up, the formula is as follows: If the ratio of shovel to truck is 1:2 or 1:3, then z i1 ={2,3}, and m i is the number of times trucks load and unload ore for the ore produced in the shovel loading zone i; Step 1.3: Calculate the maximum number of times n that a single truck can load and unload ore per day at the shovel rock loading area i∈(p+1,I) r2 , the formula is as follows: Among them, T i2 is the average time taken by a truck to load and unload a rock in the shovel loading area, d i2 is the optimal distance for trucks from rock loading point i to the rock crushing station; Step 1.4: Calculate the daily planned vehicle number z for each shovel rock loading area i∈(p+1,I) i2 ; Since the mineral processing department only gives the planned ratio of ore demand, the rock demand is evenly distributed according to the annual plan, with slight adjustments made every day, which are given by the production scheduling room. Therefore, it is only necessary to meet the target given by the production scheduling room. The formula is as follows: If the ratio of electric shovel to truck is 1:2 or 1:3, then z i2 ={2,3}, and f2 is the minimum production demand of the rock unloading point within one shift, C r is the load of the truck; Step 1.5: Calculate the number of trucks to be deployed per day:

3. The open-pit iron ore traffic planning and scheduling optimization method according to claim 2 is characterized in that: The objective function of step 2 specifically includes: The objectives of the truck dispatch daily plan include minimizing truck transportation costs, maximizing ore production, and minimizing the total idle time of trucks; the model objective function is: F(S)={F1(S),-F2(S),F3(S)}; Among them, F1(S), F2(S), and F3(S) are the sub-objectives in the objective function, which represent minimizing the total truck transportation cost, maximizing the ore output, and minimizing the total truck idle time, respectively; Truck transportation costs include three parts: the operating cost of a heavily loaded truck from the loading point to the unloading point, the operating cost of an empty truck from the unloading point to the loading point or to and from the refueling point, and the maintenance cost of the truck; the objective function expression for minimizing the total truck transportation cost is: The objective function of maximizing ore production is specifically expressed as maximizing ore transportation volume while minimizing rock transportation volume, and its expression is: and converts to: The idle time of a single truck is the difference between the daily planned working time of each shift and the necessary driving, loading, unloading, refueling and inspection time. The objective function expression for minimizing the total idle time of the truck is: Where i is the loading point index, indicating the i-th loading point, i.e., the position of the electric shovel, i=1,2,...,I; j is the unloading point index, indicating the j-th unloading point, i.e., the position of the crushing station, j=1,2,...,J; r is the truck index, indicating the r-th truck, r=1,2,...,R; x rij is the number of times the rth truck goes from loading point i to unloading point j, x rji is the number of times the rth truck goes from unloading point j to loading point i, d i,j The optimal distance from loading point i to unloading point j, d jO The optimal distance from unloading point j to refueling point O, d Oi The optimal distance from refueling point O to loading point i, C r1 is the unit distance cost of the rth truck when it is heavily loaded, C r2 is the unit distance cost when the rth truck is empty, C r3 is the wear and tear cost of the rth truck per unit time, i.e., the maintenance cost, f1 is the minimum production demand of the ore unloading point within one shift, T O Average time to refuel a truck, K rjO is the number of times truck r runs from unloading point j to refueling point O, K rOi is the number of trips of truck r from refueling point O to loading point i.

4. The open-pit iron ore traffic planning and scheduling optimization method according to claim 3 is characterized in that: The constraints in step 3 specifically include: Ore grade constraint of ore crushing station: The deviation between the actual grade of the daily plan and the given grade shall not exceed the maximum grade deviation β, as shown in the following formula: Among them, α i is the ore grade at loading point i, β is the allowable deviation of ore grade, and G1 is the target grade of the ore crushing station; Ore crushing station production plan constraints: The total amount of ore unloaded by the ore crushing station must meet the minimum demand f1 of each shift in the ore production daily plan, and the total amount of rock unloaded by the rock crushing station must meet the minimum demand f2 of each shift in the rock production daily plan, as shown in the following two formulas: The total amount of ore loaded at each shovel loading point must not exceed the maximum production capacity of the loading point. i , as follows: Traffic flow continuity constraint: The traffic flow in and out of the loading / unloading point must be equal, as shown in the following formula: Remaining fuel constraint: It is necessary to monitor the remaining fuel of each truck. When it drops to the minimum value A, it is necessary to refuel at a refueling point. The remaining fuel A = the maximum fuel consumption from the unloading point to the loading point + the maximum fuel consumption from the loading point to the unloading point + the maximum fuel consumption from the unloading point to the refueling point. The constraint formula is as follows: Among them, E is the fuel tank capacity of the truck, A is the minimum remaining fuel of the truck, and E r1 is the fuel consumption per unit distance of the r-th truck when it is heavily loaded, E r2 is the fuel consumption per unit distance of the rth truck when it is unloaded; Truck transport times constraint: The truck transport times must be a positive integer, and when the ore and rock are fully loaded, they must go to the ore and rock unloading points for unloading, while the empty trucks at the unloading points can go to any loading points; for the traffic planning and scheduling problem of mixed ore and rock transportation, let the first p of I shovel loading points be ore loading points and the last Ip be rock loading points, let the first of J unloading points be an ore crushing station and the second be a rock crushing station, the constraint formula is as follows: x rij ∈{0,1,2...},x rji ∈{0,1,2...}; in, 5. The open-pit iron ore traffic planning and scheduling optimization method according to claim 4 is characterized in that: The multi-objective intelligent optimization algorithm in step 4 adopts an improved multi-objective particle swarm optimization algorithm to solve the multi-objective optimization problem of open-pit iron ore traffic planning and scheduling; the algorithm introduces a number of improvements on the basis of the standard multi-objective particle swarm optimization algorithm to improve its performance in dealing with complex constraints and multi-objective optimization problems; the specific steps are as follows: Step 4.1: Particle encoding and initialization; Each particle represents a feasible scheduling solution, encoded as a multidimensional vector X = (x1, x2, ..., x n ), where n is the number of decision variables; during initialization, the particle positions are randomly generated while satisfying the constraints; Step 4.2: Based on the objective function established in step 2, define the multi-objective fitness function: F(S) = {F1(S), -F2(S), F3(S)}, where F1(S) represents the objective of minimizing the total truck transportation cost, F2(S) represents the objective of maximizing the ore output, and F3(S) represents the objective of minimizing the total truck idle time; Step 4.3: Introduce the adaptive weight strategy, the particle speed and position update formula is as follows: Among them, v id (t+1) is the velocity of the ith particle in the dth dimension at time t+1; v id t is the velocity of the ith particle in the dth dimension at time t; c1 is the individual learning factor, which determines the intensity of the tendency of the particle to move to the individual historical optimal position; c2 is the social learning factor, which determines the intensity of the tendency of the particle to move to the group historical optimal position; r1 and r2 are random numbers between [0,1], which are regenerated in each iteration to increase the randomness of the search; p id t is the individual historical optimal position of the ith particle in the dth dimension, i.e., the individual optimal solution; p gd t is the historical optimal position of the group in the dth dimension, i.e., the global optimal solution, x id t is the current position of the ith particle in the dth dimension at time t, x id (t+1) is the current position of the ith particle in the dth dimension at time t+1; w(t) is the adaptive inertia weight, which is dynamically adjusted by the following formula: w(t)=w max -(w max -w min )*(t / T max ) 2 ; Among them, w max and w min are the maximum and minimum values ​​of the inertia weight, t is the current iteration number, T max is the maximum number of iterations; Step 4.4: Maintenance of non-dominated solution set; The fast non-dominated sorting algorithm is used to sort the particles, and the crowding distance is used to maintain the diversity of the non-dominated solution set; the crowding distance calculation formula is as follows: Among them, CD i is the crowding distance of the ith solution, M is the number of objective functions, and are the function values ​​of the adjacent solutions of the mth objective function after sorting, f m max and f m min are the maximum and minimum values ​​of the mth objective function respectively; Step 4.5: Local search mechanism; After each iteration, the simulated annealing strategy is used to perform local search on some solutions in the non-dominated solution set. During the local search, in order to facilitate the algorithm to jump out of the local optimal solution, it is allowed to accept the worse solution according to a certain probability, that is, the acceptance probability. The calculation formula of the acceptance probability is: P(ΔE,T)=exp(-ΔE / T); Among them, ΔE is the difference between the objective function of the new solution and the current solution, and T is the current temperature, which gradually decreases with the number of iterations; Step 4.6: Constraint handling mechanism; Use the improved constraint governance rule to handle constraints and define constraint violation: Among them, g q (X)≤0 is an inequality constraint, q=1,2,...,Q; h k (X) = 0 is an equality constraint, k = 1, 2, ..., K; when comparing two solutions, first compare their constraint violation degrees, and the solution with a smaller constraint violation degree is better than the solution with a larger constraint violation degree; Step 4.7: Algorithm termination condition; The algorithm iterates until any of the following conditions is met: the preset maximum number of iterations is reached, and the non-dominated solution set does not change after N consecutive iterations.

6. The method for optimizing open pit iron mine traffic flow planning and scheduling according to claim 5, characterized in that: The details of step 5 are as follows: According to the Pareto optimal solution set obtained in step 4, use the TOPSIS method to select an optimal solution from the final non-dominated solution set as the final scheduling solution; According to the truck dispatching operation results of the selected decision-making scheme, the actual daily planned output and grade are revised for reference decision-making of dispatching operations. The revised daily planned ore output, rock output and grade formulas are as follows: in, is the revised daily planned ore production, is the corrected rock yield, is the corrected grade, α i is the ore grade at loading point i.

7. The method for optimizing open-pit iron ore traffic flow planning and scheduling according to claim 6 is characterized in that: The TOPSIS method is based on the following steps: Construct a normalized decision matrix: Calculate a weighted normalized decision matrix; determine the ideal solution and the negative ideal solution, where the ideal solution refers to a hypothetical solution in which each indicator obtains the best value, and the negative ideal solution refers to a hypothetical solution in which each indicator obtains the worst value; calculate the distance of each solution to the ideal solution and the negative ideal solution; calculate the relative proximity of each solution; and sort the solutions according to the relative proximity; The method for calculating the distance of each solution to the ideal solution and the negative ideal solution is: Let the weighted normalized value of the i-th solution on the j-th indicator be v ij , the ideal solution value at the jth index is v j+ , the value of the negative ideal solution at the jth index is v j- , then the distance S from the i-th solution to the ideal solution i+ and to the negative ideal solution S i- The distance is calculated using the Euclidean distance as follows: Among them, n is the number of evaluation indicators; The method for calculating the relative proximity of each solution is: The relative closeness C of the i-th solution i Calculated using the following formula: Relative proximity C i The value range of is [0,1], the closer to 1, the better the solution; The method for sorting the solutions according to relative proximity is: All the solutions are sorted according to their relative closeness C i Sort from largest to smallest, C i The solution with the largest value is the optimal solution.

Images