Multi-objective mining truck online scheduling method and system for real-time ore distribution

Abstract

The present invention relates to the field of mine operation optimization, and provides a multi-objective mine truck online scheduling method and system for real-time ore matching, including: obtaining actual production environment information; calculating and obtaining the optimal path from each shoveling point to each unloading point through the actual production environment information; constructing a multi-objective mine truck online scheduling model for real-time ore matching; solving the multi-objective mine truck online scheduling model through the optimal path, constraint processing and NSGA-III algorithm to obtain a scheduling plan. The present invention solves the problem of large fluctuations in the grade of selected copper due to the difficulty in controlling ore matching and the problem that it is difficult to uniformly schedule the mixed transportation of ore and waste rock by mine trucks of different models, and directly saves transportation costs to improve the economic benefits of mining enterprises, providing technical support for the construction of smart mines and unmanned mines.

Description

Technical Field

[0001] The present invention relates to the field of mine operation optimization, and in particular to a multi-objective mine truck online scheduling method and system for real-time ore distribution. Background Art

[0002] The open-pit mining process can be divided into five major links: geological exploration, drilling and blasting, shoveling and transportation, unloading and crushing, and ore dressing. The five links are closely linked, and the execution of the first four links will have a direct impact on the ore dressing results. If the grade of the original metal ore entering the crushing station fluctuates greatly, it is difficult to control the stability of the grade and output of the metal ore entering the ore dressing plant. Therefore, under the condition that the geological resources of the ore are determined after the step blasting, it becomes crucial to reduce the grade fluctuation of the original metal ore entering the crushing station, thereby maintaining the stability of the metal concentrate output and improving the recovery rate of the concentrate product.

[0003] As the main transportation equipment, large mining trucks account for about 40%-60% of the total investment of the mine, and the transportation cost accounts for about 30%-40% of the ore production cost. In addition, due to the different purchase times of equipment in the mine operation site, the models of mining trucks are not exactly the same, and the parameters such as transportation capacity and fuel consumption are also different. The size of the unloading port of the crushing station is also different for the number of different types of mining trucks, which also puts forward strict requirements on the scheduling method. At the loading point shoveling site, ore and waste rock are produced alternately, and mining trucks need to be dispatched to different unloading points according to the type of stone. However, the existing scheduling method only schedules a single transportation task for a single type of mining truck, and cannot schedule different types of mining trucks for different transportation tasks. In addition, existing mining enterprises have problems such as low ore distribution accuracy, insufficient ore dressing, low ore recovery rate, low transportation efficiency due to shovel queues, and offline scheduling solutions that cannot handle on-site emergencies in a timely manner.

[0004] In addition, most of the existing technologies for truck and forklift dispatching are based on offline dispatching, where a dispatching plan for a mining truck is generated before work and the truck driver is executed according to the dispatching plan after work. However, there are many uncertainties in actual mining. When encountering emergencies, the mining truck cannot get a new dispatching command and still executes according to the previous offline dispatching plan, which cannot solve the current emergencies. As a result, one mining truck fails to execute according to the dispatching plan, which makes the dispatching plan of all vehicles invalid.

[0005] The above contents are only used to assist in understanding the technical solution of the present invention and do not constitute an admission that the above contents are prior art. Summary of the invention

[0006] In order to solve the above technical problems, the present invention provides a multi-objective mining truck online scheduling method for real-time ore distribution, comprising:

[0007] S1: Obtain actual production environment information;

[0008] S2: Calculate and obtain the optimal path from each loading point to each unloading point through the actual production environment information;

[0009] S3: Build a multi-objective mining truck online scheduling model for real-time mining allocation;

[0010] S4: Solving the multi-objective mining truck online scheduling model through the optimal path, constraint processing and NSGA-III algorithm to obtain a scheduling solution.

[0011] Preferably, step S1 specifically comprises:

[0012] S11: Obtain the number of shoveling points, the location of shoveling points, unloading point information, the number of different types of mining trucks, the load capacity of different types of mining trucks, the unit time fuel consumption of different types of mining trucks, road-related information and production plans through the mine's backend statistical system;

[0013] The unloading point information includes: the number of ore crushing stations, the location of the ore crushing stations, the number of waste rock crushing stations, the location of the waste rock crushing stations, the number of spoil dumps and the location of the spoil dumps;

[0014] S12: Obtain underground three-dimensional grade distribution information of open-pit mines through exploration data;

[0015] S13: Calculate the travel time of different models of mining trucks on different road sections and the locations of the mining trucks through the GPS location information and speed information of each mining truck.

[0016] Preferably, step S3 specifically comprises:

[0017] S31: By analyzing the key factors that affect the excessive fluctuation of copper grade and the low efficiency of mining truck transportation, a multi-objective mining truck online scheduling model for real-time ore distribution is constructed;

[0018] S32: Determine the target to be optimized in the multi-target mining truck online scheduling model;

[0019] S33: Determine the constraints satisfied by the multi-objective mining truck online scheduling model.

[0020] Preferably, the objectives to be optimized include:

[0021] The first optimization goal is to wait for the mining truck to travel from the loading point to the unloading point within three round trips. 1 (S)Minimum:

[0022]

[0023] Among them, p represents the number of the shovel loading point, n represents the total number of shovel loading points; q represents the number of the unloading point, m represents the total number of unloading points; r represents the number of the mining truck, k represents the total number of mining trucks; It indicates the waiting time of the rth mining card at the pth shovel loading point. represents the waiting time of mine No. r at mine No. q;

[0024] The second optimization goal is the deviation F between the average ore grade of each mining truck before and after entering the crushing station and the target grade interval of the crushing station 2 (S)Minimum:

[0025]

[0026] Where u represents the number of the mining trucks sorted in the order of arrival at the crushing station, v represents the number of the first mining truck for which the average ore grade is to be calculated, and w represents the total number of mining trucks for which the average ore grade is to be calculated within the time window before and after arrival at the crushing station; C u represents the amount of ore loaded by the u-th mining truck arriving at the crushing station; α u represents the ore grade loaded by the uth mining truck arriving at the crushing station; q 1 Represents the number of the crushing station. Represents q 1 Target ore grade for crushing station No.

[0027] The third optimization goal is to achieve the energy consumption F of the mining truck within three round trips from the loading point to the unloading point. 3 (S)Minimum:

[0028]

[0029] in, Represents the heavy-load fuel consumption of mining truck No. r from loading point No. p to unloading point No. q, Represents the no-load fuel consumption of mining truck r from unloading point q to loading point p. Represents the number of times the mining truck r has moved from the loading point p to the unloading point q. Represents the number of times mining truck r has moved from unloading point q to loading point p.

[0030] Preferably, the constraints include:

[0031] The first constraint is that mining trucks with different load capacities can only go to the loading points corresponding to their loading capacities:

[0032] xT r ∈G r

[0033] Among them, X r Represents the next destination of the mining card No. r, G rRepresents the set of loading points where the shovel trucks are located corresponding to the loading capacity of the mining truck;

[0034] The second constraint is that the current ore output at the loading point is greater than the total load capacity of the ore truck to be loaded at that point:

[0035]

[0036] in, represents the amount of ore loaded by the r mining truck at the p shovel loading point, D p Represents the current ore reserves of shovel loading point p;

[0037] The third constraint is that the amount of ore entering the crushing station per unit time is less than or equal to the rated crushing amount of ore in the crushing station:

[0038]

[0039] in, Representatives arrive at q 1 The amount of ore loaded by the No. r mining truck of the No. crushing station, D q1 Represents q 1 The current silo volume of crushing station No.

[0040] The fourth constraint is that the distance between mining trucks on the main road is greater than the safety distance:

[0041]

[0042] Among them, g represents the time, Represents the position of the rth mining card at time g, represents the position of the mining card adjacent to the mining card r at time g, L safe Represents the minimum safe distance between two adjacent mining trucks to ensure safe driving;

[0043] The fifth constraint is the ratio constraint of powder ore and lump ore:

[0044]

[0045] Among them, C u represents the amount of ore loaded by the u-th mining truck arriving at the crushing station, γ u Represents the powder content in the ore transported by the mining truck arriving at the crushing station at the uth time, Represents q 1 The requirements for the ratio of powder ore and lump ore in crushing station No.

[0046] The sixth constraint condition is that the kaolin content in the crushing station silo is less than 20%:

[0047]

[0048] Among them, βu Represents the kaolin content in the ore transported by the u-th mining truck arriving at the crushing station, Represents q 1 Requirements for kaolin content in crushing station No.

[0049] The seventh constraint is that the forklift does not stop working:

[0050]

[0051] in, It represents the time when the rth mine card leaves the pth shovel loading point. The time when the next mining truck after the mining truck with the number R leaves the shovel loading point with the number P;

[0052] The eighth constraint condition is that there should be no interruption events that affect the operation of the mining truck due to lack of oil, driver fatigue and tire break:

[0053]

[0054] Among them, I r Represents the status of mining card No. r.

[0055] Preferably, step S4 is specifically:

[0056] S41: by constraint processing, solving the multi-objective mining truck online scheduling model is constructed as a dynamic constrained multi-objective optimization problem for solving four optimization objectives;

[0057] S42: According to the actual operation problem of the mining truck scheduling, the encoding of the solution of the dynamic constraint multi-objective optimization problem is designed through the optimal path to obtain the scheduling solution chromosome;

[0058] S43: Iteratively solving the dynamic constrained multi-objective optimization problem based on the NSGA-III algorithm, representing the solution of the dynamic constrained multi-objective optimization problem through the scheduling scheme chromosome, and outputting the scheduling scheme.

[0059] Preferably, step S41 is specifically as follows:

[0060] S411: The normalized average degree of constraint violation of the parameter x to be solved on all constraints is taken as the default value target, and the formula is as follows:

[0061]

[0062] Among them, h represents the number of constraints, P 0 represents the initial population, G i (x) represents the degree to which x violates the constraint under the i-th constraint condition;

[0063] S412: The expression of the dynamic constraint multi-objective optimization problem is:

[0064]

[0065]

[0066]

[0067] F 4 (x) = mincv(x)

[0068]

[0069]

[0070] Among them, ε (t) represents the dynamic constraint boundary, t represents the number of environmental changes, and T represents the maximum number of environmental changes, satisfying θ(x)≤ε (t) The solution is called an ε-feasible solution; otherwise, it is called an ε-infeasible solution.

[0071] Preferably, the expression of the scheduling scheme chromosome X is:

[0072]

[0073] Among them, [ABC] represents the set of shovel loading points, A, B and C represent shovel loading points; [abc] represents the set of crushing stations, a, b and c represent crushing stations; [AaBbCc] means the route of the mining truck is A to a to B to b to C to c.

[0074] Preferably, step S43 is specifically as follows:

[0075] S431: For the dynamic constrained multi-objective optimization problem, parameters are initialized and reference points on the hyperplane are determined. The calculation formula for the number of reference points Q is as follows:

[0076]

[0077] Where M represents the dimension of the target vector, and H represents the number of target divisions;

[0078] S432: Construct the multi-objective mining truck online scheduling model to obtain the initial population P 0 ;

[0079] S433: Through the NSGA-III algorithm, the parent population P is formed t After that, the tournament selection mechanism was introduced from P t Select parent individuals from the generation to construct the offspring population Q t ;

[0080] S434: For the P t , Qt The population R after merging t Perform non-dominated rank sorting;

[0081] S435: R t Perform adaptive normalization, individual association reference points, and microhabitat preservation operations, and select the dominant individuals to enter the next generation population P after environmental selection. t+1 ;

[0082] S436: repeat steps S433 to S435 until the maximum number of iterations is reached and then proceed to step S437;

[0083] S437: Calculate and obtain the optimal compromise solution, specifically:

[0084] Calculate a set of Pareto solutions for the dynamic constrained multi-objective optimization problem, and the jth objective value f of the i-th Pareto solution ij The membership function h ij The calculation formula is:

[0085]

[0086] Among them, f jmax represents the maximum value of the jth objective function, f jmin Represents the minimum value of the jth objective function;

[0087] For the i-th Pareto solution, its standardized membership function h i The calculation formula is:

[0088]

[0089] Where I is the total number of Pareto solutions;

[0090] Select the membership function h i The solution with the largest value is the compromise optimal solution;

[0091] S438: Sending the scheduling plan corresponding to the compromise optimal solution to the mining card through the instruction system;

[0092] S439: Repeat steps S432 to S438 until the scheduling plans received by all mining cards are completed.

[0093] A multi-objective mining truck online scheduling system for real-time ore distribution, comprising:

[0094] An actual production environment information acquisition module is used to acquire actual production environment information;

[0095] An optimal path calculation module, used to calculate and obtain the optimal path from each loading point to each unloading point through the actual production environment information;

[0096] A multi-objective mining truck online scheduling model construction module is used to build a multi-objective mining truck online scheduling model for real-time ore distribution;

[0097] The scheduling scheme generation module is used to solve the multi-objective mining truck online scheduling model through the optimal path, constraint processing and NSGA-III algorithm to obtain a scheduling scheme.

[0098] The present invention has the following beneficial effects:

[0099] 1. Solve the problem of the difficulty of unified online scheduling of different types of mining trucks transporting ore and waste rock together by solving the established multi-objective mining truck online scheduling model for real-time ore distribution with the goal of minimizing transportation energy consumption, minimizing grade fluctuation of selected metal ore, and minimizing waiting time of mining trucks.

[0100] 2. During the transportation stage, the vehicle flow is controlled to carry out real-time ore distribution to solve the problem of large fluctuations in the grade and output of finished ore caused by large fluctuations in the grade of the selected metal ore. This is the most direct way to save transportation costs and improve the economic benefits of mining companies, providing technical support for the construction of smart mines and unmanned mines. BRIEF DESCRIPTION OF THE DRAWINGS

[0101] Figure 1 This is a flow chart of a method according to an embodiment of the present invention;

[0102] Figure 2 This is a schematic diagram of the road network in the open-pit metal mine area;

[0103] Figure 2 The numbers on the middle line represent weights, and the numbers in the circles represent the node numbers of the road network;

[0104] The realization of the purpose, functional features and advantages of the present invention will be further explained in conjunction with embodiments and with reference to the accompanying drawings. DETAILED DESCRIPTION

[0105] It should be understood that the specific embodiments described herein are only used to explain the present invention, and are not used to limit the present invention.

[0106] Reference Figure 1The present invention provides a multi-objective online scheduling method for real-time ore matching, which is improved based on the NSGA-III algorithm framework, and the encoding and local search operator of the solution are designed according to the characteristics of the problem. The multi-objective online scheduling model for real-time ore matching with the minimum transportation energy consumption, the minimum fluctuation of the grade of the selected metal ore, and the minimum waiting time of the ore truck is solved to solve the problem that it is difficult to uniformly schedule the mixed transportation of ore and waste rock by ore trucks of different models. In the transportation stage, the real-time ore matching is carried out by controlling the traffic flow to solve the problem of large fluctuations in the grade and output of the finished ore caused by the large fluctuations in the grade of the selected metal ore, which directly saves the transportation cost to improve the economic benefits of the mining enterprise, and provides technical support for the construction of smart mines and unmanned mines.

[0107] Specifically include:

[0108] S1: Obtain actual production environment information;

[0109] S2: Calculate and obtain the optimal path from each loading point to each unloading point through the actual production environment information;

[0110] Specifically, the optimal path is determined by setting the path with the smallest average historical time taken by the driver to drive the mining truck to complete the path according to the road conditions as the optimal path; according to the mine road network information, the driver's driving time on each road section is used as the weight value, and the Floyd algorithm is used to calculate the optimal path from each shovel loading point to each crushing station;

[0111] S3: Build a multi-objective mining truck online scheduling model for real-time mining allocation;

[0112] S4: Solving the multi-objective mining truck online scheduling model through the optimal path, constraint processing and NSGA-III algorithm to obtain a scheduling solution.

[0113] In this embodiment, step S1 is specifically as follows:

[0114] S11: Through the mine's backend statistical system, obtain the number of shoveling points, the location of shoveling points, unloading point information, the number of different types of mining trucks, the load capacity of different types of mining trucks, the unit time fuel consumption of different types of mining trucks, road related information (length, width, slope, friction coefficient, number of turns) and production plan; also includes vehicle conditions (fuel consumption per unit time, health parameters);

[0115] The unloading point information includes: the number of ore crushing stations, the location of the ore crushing stations, the number of waste rock crushing stations, the location of the waste rock crushing stations, the number of spoil dumps and the location of the spoil dumps;

[0116] S12: Obtain underground three-dimensional grade distribution information of open-pit mines through exploration data;

[0117] S13: Calculate the travel time of different models of mining trucks on different road sections and the locations of the mining trucks through the GPS location information and speed information of each mining truck.

[0118] In this embodiment, step S3 is specifically as follows:

[0119] S31: By analyzing the key factors that affect the excessive fluctuation of copper grade and the low efficiency of mining truck transportation, a multi-objective mining truck online scheduling model for real-time ore distribution is constructed;

[0120] Specifically, the goal is to minimize the waiting time of the mining truck from the shoveling point to the crushing station in one round trip, minimize the deviation between the average value of the ore grade loaded by several mining trucks before and after entering the crushing station and the target grade range of the crushing station, and minimize the energy consumption of the mining truck from the shoveling point to the crushing station in one round trip. The constraints are the shoveling capacity of the mining point, the crushing capacity of the crushing station, the number of mining trucks to be loaded at the mining point, the amount of ore to be unloaded at the crushing station, the type of mining trucks entering the shoveling point, the ore powder-lump ratio and the kaolin content in the silo of the crushing station, and a multi-objective mining truck online scheduling model for real-time ore distribution is established.

[0121] S32: Determine the target to be optimized in the multi-target mining truck online scheduling model;

[0122] S33: Determine the constraints satisfied by the multi-objective mining truck online scheduling model.

[0123] In this embodiment, the objectives to be optimized include:

[0124] The first optimization goal is to wait for the mining truck to travel from the loading point to the unloading point within three round trips. 1 (S)Minimum:

[0125]

[0126] Among them, p represents the number of the shovel loading point, n represents the total number of shovel loading points; q represents the number of the unloading point, m represents the total number of unloading points; r represents the number of the mining truck, k represents the total number of mining trucks; It indicates the waiting time of the rth mining card at the pth shovel loading point. Represents the waiting time of the rth ore card at the qth crushing station;

[0127] The second optimization goal is the deviation F between the average ore grade of each mining truck before and after entering the crushing station and the target grade interval of the crushing station 2 (S)Minimum:

[0128]

[0129] Where u represents the number of the mining trucks sorted in the order of arrival at the crushing station, v represents the number of the first mining truck for which the average ore grade is to be calculated, and w represents the total number of mining trucks for which the average ore grade is to be calculated within the time window before and after arrival at the crushing station; C u represents the amount of ore loaded by the u-th mining truck arriving at the crushing station; α u represents the ore grade loaded by the uth mining truck arriving at the crushing station; q 1 Represents the number of the crushing station. Represents q 1 Target ore grade for crushing station No.

[0130] The third optimization goal is to achieve the energy consumption F of the mining truck within three round trips from the loading point to the unloading point. 3 (S)Minimum:

[0131]

[0132] in, Represents the heavy-load fuel consumption of mining truck No. r from loading point No. p to unloading point No. q, Represents the no-load fuel consumption of mining truck r from unloading point q to loading point p. Represents the number of times the mining truck r has moved from the loading point p to the unloading point q. Represents the number of times mining truck r has moved from unloading point q to loading point p.

[0133] In this embodiment, the constraints include:

[0134] The first constraint is that mining trucks with different load capacities can only go to the loading points corresponding to their loading capacities:

[0135] xT r ∈G r (4)

[0136] Among them, X r Represents the next destination of the mining card No. r, G r Represents the set of loading points where the shovel trucks are located corresponding to the loading capacity of the mining truck;

[0137] The second constraint is that the current ore output at the loading point is greater than the total load of the ore truck planned to be loaded at that point:

[0138]

[0139] in, represents the amount of ore loaded by the r mining truck at the p shovel loading point, D p Represents the current ore reserves of shovel loading point p;

[0140] The third constraint is that the amount of ore entering the crushing station per unit time is less than or equal to the rated crushing amount of ore in the crushing station:

[0141]

[0142] in, Representatives arrive at q 1 The amount of ore loaded by the No. R mining truck of the No. Represents q 1 The current silo volume of crushing station No.

[0143] The fourth constraint is that the distance between mining trucks on the main road is greater than the safety distance:

[0144]

[0145] Among them, g represents the time, Represents the position of the rth mining card at time g, represents the position of the mining card adjacent to the mining card r at time g, L safe Represents the minimum safe distance between two adjacent mining trucks to ensure safe driving;

[0146] The fifth constraint is the ratio constraint of powder ore and lump ore:

[0147]

[0148] Among them, C u represents the amount of ore loaded by the u-th mining truck arriving at the crushing station, γ u Represents the powder content in the ore transported by the mining truck arriving at the crushing station at the uth time, Represents q 1 The requirements for the ratio of powder ore and lump ore in crushing station No.

[0149] The sixth constraint condition is that the kaolin content in the crushing station silo is less than 20%:

[0150]

[0151] Among them, β u Represents the kaolin content in the ore transported by the u-th mining truck arriving at the crushing station, Represents q 1 Requirements for kaolin content in crushing station No.

[0152] The seventh constraint is that the forklift does not stop working:

[0153]

[0154] in, It represents the time when the rth mine card leaves the pth shovel loading point. The time when the next mining truck after the mining truck with the number R leaves the shovel loading point with the number P;

[0155] The eighth constraint condition is that there should be no interruption events that affect the operation of the mining truck due to lack of oil, driver fatigue and tire break:

[0156]

[0157] Among them, I r Represents the status of mining card No. r.

[0158] In this embodiment, step S4 is specifically as follows:

[0159] S41: by constraint processing, solving the multi-objective mining truck online scheduling model is constructed as a dynamic constrained multi-objective optimization problem for solving four optimization objectives;

[0160] S42: According to the actual operation problem of the mining truck scheduling, the encoding of the solution of the dynamic constraint multi-objective optimization problem is designed through the optimal path to obtain the scheduling solution chromosome;

[0161] S43: Iteratively solving the dynamic constrained multi-objective optimization problem based on the NSGA-III algorithm, representing the solution of the dynamic constrained multi-objective optimization problem through the scheduling scheme chromosome, and outputting the scheduling scheme.

[0162] In this embodiment, step S41 is specifically as follows:

[0163] S411: This framework takes the default value as an additional objective to transform an m-objective constrained optimization problem:

[0164] (f 1 (x),…,f m (x)) is transformed into an m+1 objective constrained optimization problem (f 1 (x),…,f m (x),cv(x));

[0165] The normalized average degree of constraint violation of the parameter x to be solved on all constraints is taken as the default value target, and the formula is as follows:

[0166]

[0167] Among them, h represents the number of constraints, P 0 represents the initial population, G i (x) represents the degree to which x violates the constraint under the i-th constraint condition;

[0168] S412: The expression of the dynamic constraint multi-objective optimization problem is:

[0169]

[0170]

[0171]

[0172] F 4 (x) = mincv(x)

[0173]

[0174]

[0175] Among them, ε (t) represents the dynamic constraint boundary, t represents the number of environmental changes, and T represents the maximum number of environmental changes, satisfying θ(x)≤ε (t) The solution is called an ε-feasible solution; otherwise, it is called an ε-infeasible solution.

[0176] refer to Figure 2 In this embodiment, based on the actual operation problem of mining truck scheduling, character encoding is used to simply and accurately express the operation route of mining trucks; loading points and crushing stations are represented by uppercase and lowercase letters respectively, so the operation route of mining trucks is represented by a string consisting of alternating uppercase and lowercase letters; such as [ABCDE] represents the set of loading points, [abc] represents the set of crushing stations, and [AaBbCcDaEb] represents the operation route of a mining truck (the mining truck runs from loading point A to crushing station a and then to loading point B, and so on). At the same time, the operation routes of all mining trucks in a fixed time period are combined into one solution, so a scheduling plan for all mining trucks can be represented by one solution, such as matrix X represents one solution and one scheduling plan, the number of rows represents 1# mining truck to n# mining truck, and the number of columns represents the target point set of the mining truck;

[0177] The expression of the scheduling scheme chromosome X is:

[0178]

[0179] Among them, [ABC] represents the set of shovel loading points, A, B and C represent shovel loading points; [abc] represents the set of crushing stations, a, b and c represent crushing stations; [AaBbCc] means that the route of 1# mining truck is A to a to B to b to C to c.

[0180] In this embodiment, step S43 is specifically as follows:

[0181] S431: For the dynamic constrained multi-objective optimization problem, parameter initialization is performed and reference points on the hyperplane are determined. Parameter initialization includes setting the population size, the maximum number of iterations, the crossover probability, and the mutation probability. The number of reference points depends on the dimension M of the target vector and the number H of parts into which each target is divided.

[0182] The calculation formula of the number of reference points Q is as follows:

[0183]

[0184] Where M represents the dimension of the target vector, and H represents the number of target divisions;

[0185] S432: Construct the multi-objective mining truck online scheduling model to obtain the initial population P 0 ;

[0186] During the first run, the initial population P that meets the conditions is generated according to the established multi-objective mine truck online scheduling model for real-time ore distribution. 0 When it is not the first time to run, the initial population of this round is generated according to the execution of the scheduling plan by all vehicles;

[0187] S433: Through the NSGA-III algorithm, the parent population P is formed t After that, the tournament selection mechanism was introduced from P t Select parent individuals from the generation to construct the offspring population Q t ;

[0188] In the specific implementation, according to the characteristics of this problem, the following seven search operators are designed for tournament selection. LS1-LS4 are random search operators, and LS5-LS7 are heuristic search operators:

[0189] LS1: Randomly select a mining truck transportation route, then randomly select two sets of sequences from the shovel loading point to the crushing station from the route and exchange the two sets of transportation sequences;

[0190] LS2: Randomly select a mining truck transportation route, then randomly select two sequences from the shovel loading point to the crushing station from the route and exchange the two transportation sequences;

[0191] LS3: Randomly select two mining truck transport routes, then randomly select two sets of sequences from the shovel loading point to the crushing station from the route and exchange the two sets of transport sequences;

[0192] LS4: Randomly select two mining truck transport routes, then randomly select two sequences from the shovel loading point to the crushing station from the route and exchange the two transport sequences;

[0193] LS5: Find the transportation route with the longest waiting time from the solution, select the loading and unloading point combination with the longest waiting time in the mining truck operation route, and then replace the selected loading and unloading points with the loading and unloading points that make the mining truck wait the shortest time;

[0194] LS6 finds the transport route with the highest energy consumption from the solution, selects the combination of loading and unloading points with high energy consumption in the mining truck operation route, and then replaces the selected loading and unloading points with the combination of loading and unloading points that makes the mining truck consume the lowest energy;

[0195] LS7 finds the relevant transportation routes with the largest deviation from the target grade from the solution, selects a group of routes with higher grade deviation among the mining truck operation routes (the average grade of several vehicles before and after), and then replaces the selected group of routes with higher grade deviation with the group of routes with the lowest grade deviation;

[0196] S434: For the P t , Q t The population R after merging t Perform non-dominated rank sorting;

[0197] Specifically, in order to t The optimal N solutions of the current generation are selected to enter the next iteration process. First, the non-dominated sorting method is used to sort R t Divide into several groups of non-dominated layers of different levels; then, add the solutions starting from non-dominated level 1 to S in sequence t In, until S t The number of solutions is equal to N or greater than N for the first time;

[0198] S435: R t Perform adaptive normalization, individual association reference points, and microhabitat preservation operations, and select the dominant individuals to enter the next generation population P after environmental selection. t+1 ;

[0199] Specifically, the process of adaptive normalization is:

[0200] Calculate the minimum values ​​of the four established objective functions, assuming that the corresponding minimum value obtained on the target axis i is and The set is the ideal point set mentioned in the NSGA-III algorithm, and then equation (16) is used to transform the objective function value;

[0201]

[0202] In order to find the extreme point, it is necessary to use the scalarizing function (ASF) as shown in formula (17);

[0203]

[0204] Among them, e i is the target axis f i The axis direction satisfies if i≠j, then, e i,j=0; otherwise, e i,j =1, for e i,j =0, then use a very small value of 10 -6 Instead, traverse each objective function and find the individual with the lowest ASF value to form the extreme point. The extreme point and the origin (ideal point) form three lines, which can form a hyperplane. The intersection between the hyperplane and the coordinate axis is the intercept required, and then normalized by equation (18). Among them, is the normalized objective function value;

[0205]

[0206] The process of associating individual reference points is:

[0207] After normalization, individuals need to be associated with reference points. Use the line formed by the reference point and the origin as the baseline, then traverse all reference lines to find the reference line closest to each individual, record the corresponding reference point and the shortest distance, and then calculate the number of individuals associated with each reference point;

[0208] The process of microhabitat preservation operation is as follows:

[0209] Definition S t+1 is the set of all individuals in the population from non-dominated level 1 to non-dominated level L, and then traverses each reference point to see if the reference point is S t+1 The number of citations by individuals outside the non-dominated hierarchy L is ρ j ; First, determine whether there is an individual associated with this reference point. If not, change the reference point. If there is a reference point associated with it, then determine ρ j ; if ρ j = 0, then select the solution with the minimum distance from the non-dominated level L to the reference point j and add it to P t+1 ; if ρ j >1, then randomly select a solution associated with a non-dominated level L reference point and add it to P t+1 until P t+1 The number of individuals in is equal to the size of the original population;

[0210] S436: repeat steps S433 to S435 until the maximum number of iterations is reached and then proceed to step S437;

[0211] S437: Use fuzzy decision-making method to calculate the optimal compromise solution from the Pareto solution set, specifically:

[0212] Calculate a set of Pareto solutions for the dynamic constrained multi-objective optimization problem, and the jth objective value f of the i-th Pareto solution ij The membership function hij The calculation formula is:

[0213]

[0214] Among them, f jmax represents the maximum value of the jth objective function, f jmin Represents the minimum value of the jth objective function;

[0215] For the i-th Pareto solution, its standardized membership function h i The calculation formula is:

[0216]

[0217] Where I is the total number of Pareto solutions;

[0218] Select the membership function h i The solution with the largest value is the compromise optimal solution;

[0219] S438: Sending the scheduling plan corresponding to the compromise optimal solution to the mining card through the instruction system;

[0220] S439: Repeat steps S432 to S438 until the cumulative transportation volume of all mining trucks reaches the set transportation target.

[0221] A multi-objective mining truck online scheduling system for real-time ore distribution, comprising:

[0222] An actual production environment information acquisition module is used to acquire actual production environment information;

[0223] An optimal path calculation module, used to calculate and obtain the optimal path from each loading point to each unloading point through the actual production environment information;

[0224] A multi-objective mining truck online scheduling model construction module is used to build a multi-objective mining truck online scheduling model for real-time ore distribution;

[0225] The scheduling scheme generation module is used to solve the multi-objective mining truck online scheduling model through the optimal path, constraint processing and NSGA-III algorithm to obtain a scheduling scheme.

[0226] It should be noted that, in this article, the terms "include", "comprises" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, article or system including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or system. In the absence of further restrictions, an element defined by the sentence "comprises a ..." does not exclude the existence of other identical elements in the process, method, article or system including the element.

[0227] The serial numbers of the embodiments of the present invention are for descriptive purposes only and do not represent the superiority or inferiority of the embodiments. In a unit claim that lists several means, several of these means may be embodied by the same hardware item. The use of the words first, second, and third, etc. does not indicate any order and these words may be interpreted as identifiers.

[0228] The above are only preferred embodiments of the present invention, and are not intended to limit the patent scope of the present invention. Any equivalent structure or equivalent process transformation made using the contents of the present invention specification and drawings, or directly or indirectly applied in other related technical fields, are also included in the patent protection scope of the present invention.

Claims

1. A multi-objective mining truck online scheduling method for real-time mining allocation, It is characterized in that include: S1: Obtain actual production environment information; S2: Calculate and obtain the optimal path from each loading point to each unloading point through the actual production environment information; S3: Build a multi-objective mining truck online scheduling model for real-time mining allocation; S4: solving the multi-objective mining truck online scheduling model through the optimal path, constraint processing and NSGA-III algorithm to obtain a scheduling solution; Step S3 is specifically as follows: S31: By analyzing the key factors that affect the excessive fluctuation of copper grade and the low efficiency of mining truck transportation, a multi-objective mining truck online scheduling model for real-time ore distribution is constructed; S32: Determine the target to be optimized in the multi-target mining truck online scheduling model; S33: Determine the constraint conditions satisfied by the multi-objective mining truck online scheduling model; The objectives to be optimized include: The first optimization goal is to wait for the mining truck to travel from the loading point to the unloading point within three round trips. 1 (S) minimum; The second optimization objective is the deviation F of the average grade of the ore loaded by each mining truck before and after entering the crushing station from the target grade range of the crushing station. 2 (S) is minimized; The third optimization goal is to achieve the energy consumption F of the mining truck within three round trips from the loading point to the unloading point. 3 (S) minimum; The constraints include: The first constraint is that mining trucks with different load capacities can only go to the loading points corresponding to their loading capacities. The second constraint condition is that the current ore output at the loading point is greater than the total load capacity of the ore truck to be loaded at that point; The third constraint condition is that the amount of ore entering the crushing station per unit time is less than or equal to the rated crushing amount of ore in the crushing station; The fourth constraint condition is that the distance between mining trucks on the main road is greater than the safety distance; The fifth constraint is the ratio constraint of powder ore and lump ore; The sixth constraint is that the kaolin content in the crushing station silo is less than 20%; The seventh constraint is that the forklift does not stop working; The eighth constraint condition is that there should be no interruption events that affect the operation of the mining truck due to lack of oil, driver fatigue rest, or flat tire conditions.

2. The multi-objective mining truck online scheduling method for real-time ore distribution according to claim 1, It is characterized in that Step S1 is specifically as follows: S11: Obtain the number of shoveling points, the location of shoveling points, unloading point information, the number of different types of mining trucks, the load capacity of different types of mining trucks, the unit time fuel consumption of different types of mining trucks, road-related information and production plans through the mine's backend statistical system; The unloading point information includes: the number of ore crushing stations, the location of the ore crushing stations, the number of waste rock crushing stations, the location of the waste rock crushing stations, the number of spoil dumps and the location of the spoil dumps; S12: Obtain underground three-dimensional grade distribution information of open-pit mines through exploration data; S13: Calculate the travel time of different models of mining trucks on different road sections and the locations of the mining trucks through the GPS location information and speed information of each mining truck.

3. The multi-objective mining truck online scheduling method for real-time ore distribution according to claim 1, Features: The expression of the first optimization objective is: Among them, p represents the number of the shovel loading point, n represents the total number of shovel loading points; q represents the number of the unloading point, m represents the total number of unloading points; r represents the number of the mining truck, k represents the total number of mining trucks; It indicates the waiting time of the rth mining card at the pth shovel loading point. represents the waiting time of mine No. r at mine No. q; The expression of the second optimization objective is: Where u represents the number of the mining trucks sorted in the order of arrival at the crushing station, v represents the number of the first mining truck for which the average ore grade is to be calculated, and w represents the total number of mining trucks for which the average ore grade is to be calculated within the time window before and after arrival at the crushing station; C u represents the amount of ore loaded by the u-th mining truck arriving at the crushing station; α u represents the ore grade loaded by the uth mining truck arriving at the crushing station; q 1 Represents the number of the crushing station. Represents q 1 Target ore grade for crushing station No. The expression of the third optimization objective is: in, Represents the heavy-load fuel consumption of mining truck No. r from loading point No. p to unloading point No. q, Represents the no-load fuel consumption of mining truck r from unloading point q to loading point p. Represents the number of times the mining truck r has moved from the loading point p to the unloading point q. Represents the number of times mining truck r has moved from unloading point q to loading point p.

4. The multi-objective mining truck online scheduling method for real-time ore distribution according to claim 1, Features : The expression of the first constraint is: s.t.X r ∈G r Among them, X r Represents the next destination of the mining card No. r, G r Represents the set of loading points where the shovel trucks are located corresponding to the loading capacity of the mining truck; The expression of the second constraint is: in, represents the amount of ore loaded by the r mining truck at the p shovel loading point, D p represents the current ore reserves of the shovel loading point p; k represents the total number of ore cards; The expression of the third constraint is: in, Representatives arrive at q 1 The amount of ore loaded by the No. R mining truck of the No. Represents q 1 The current silo volume of crushing station No. The expression of the fourth constraint is: Among them, g represents the time, Represents the position of the rth mining card at time g, represents the position of the mining card adjacent to the mining card r at time g, L safe Represents the minimum safe distance between two adjacent mining trucks to ensure safe driving; The expression of the fifth constraint is: Among them, C u represents the amount of ore loaded by the u-th mining truck arriving at the crushing station, γ u Represents the powder content in the ore transported by the mining truck arriving at the crushing station at the uth time, Represents q 1 The requirements for the ratio of powder ore and lump ore in crushing station No. The expression of the sixth constraint is: Among them, β u Represents the kaolin content in the ore transported by the u-th mining truck arriving at the crushing station, Represents q 1 Requirements for kaolin content in crushing station No. The expression of the seventh constraint is: Among them, represents the time when the r-th mining truck arrives at the p-th loading point, represents the time when the mining truck following the r-th mining truck arrives at the p-th loading point; The expression of the eighth constraint is: Among them, I r Represents the status of mining card No. r.

5. The multi-objective mining truck online scheduling method for real-time ore distribution according to claim 1, It is characterized in that Step S4 is specifically as follows: S41: by constraint processing, solving the multi-objective mining truck online scheduling model is constructed as a dynamic constrained multi-objective optimization problem for solving four optimization objectives; S42: According to the actual operation problem of the mining truck scheduling, the encoding of the solution of the dynamic constraint multi-objective optimization problem is designed through the optimal path to obtain the scheduling solution chromosome; S43: Iteratively solving the dynamic constrained multi-objective optimization problem based on the NSGA-III algorithm, representing the solution of the dynamic constrained multi-objective optimization problem through the scheduling scheme chromosome, and outputting the scheduling scheme.

6. The multi-objective mining truck online scheduling method for real-time ore distribution according to claim 5, It is characterized in that Step S41 is specifically as follows: S411: The normalized average degree of constraint violation of the parameter x to be solved on all constraints is taken as the default value target, and the formula is as follows: Among them, h represents the number of constraints, P 0 represents the initial population, G i (x) represents the degree to which x violates the constraint under the i-th constraint condition; S412: The expression of the dynamic constraint multi-objective optimization problem is: F 4 (x)=mincv(x) Among them, p represents the number of the shovel loading point, n represents the total number of shovel loading points; q represents the number of the unloading point, m represents the total number of unloading points; r represents the number of the mining truck, k represents the total number of mining trucks; It indicates the waiting time of the rth mining card at the pth shovel loading point. represents the waiting time of mine No. r at mine No. q; u represents the number of the haul truck sorted in the order of arrival at the crushing station, v represents the number of the first haul truck for calculating the average ore grade, and w represents the total number of haul trucks for calculating the average ore grade within the time window before and after arriving at the crushing station; C u represents the amount of ore loaded by the u-th haul truck arriving at the crushing station; α u represents the ore grade of the ore loaded by the u-th haul truck arriving at the crushing station; q 1 represents the number of the crushing station, represents q 1 the target ore grade of the q-th crushing station; Represents the heavy-load fuel consumption of mining truck No. r from loading point No. p to unloading point No. q, Represents the no-load fuel consumption of mining truck r from unloading point q to loading point p. Represents the number of times the mining truck r has moved from the loading point p to the unloading point q. Represents the number of times the mining truck r has traveled from the unloading point q to the loading point p; represents the amount of ore loaded by the r mining truck at the p shovel loading point, D p Represents the current ore reserves of shovel loading point p; Representatives arrive at q 1 The amount of ore loaded by the No. R mining truck of the No. Represents q 1 The current silo volume of crushing station No. g represents the time, Represents the position of the rth mining card at time g, represents the position of the mining card adjacent to the mining card r at time g, L safe Represents the minimum safe distance between two adjacent mining trucks to ensure safe driving; γ u Represents the powder content in the ore transported by the mining truck arriving at the crushing station at the uth time, Represents q 1 The requirements for the ratio of powder ore and lump ore in crushing station No. β u Represents the kaolin content in the ore transported by the u-th mining truck arriving at the crushing station, Represents q 1 Requirements for kaolin content in crushing station No. It represents the time when the rth mine card leaves the pth shovel loading point. The time when the next mining truck after the mining truck with the number R leaves the shovel loading point with the number P; ε (t) represents the dynamic constraint boundary, t represents the number of environmental changes, and T represents the maximum number of environmental changes, satisfying θ(x)≤ε (t) The solution is called an ε-feasible solution; otherwise, it is called an ε-infeasible solution.

7. The multi-objective mining truck online scheduling method for real-time ore distribution according to claim 5, It is characterized in that The expression of the scheduling scheme chromosome X is: Among them, [ABC] represents the set of shovel loading points, A, B and C represent shovel loading points; [abc] represents the set of crushing stations, a, b and c represent crushing stations; [AaBbCc] means the route of the mining truck is A to a to B to b to C to c.

8. The multi-objective mining truck online scheduling method for real-time ore distribution according to claim 5, It is characterized in that Step S43 is specifically as follows: S431: For the dynamic constrained multi-objective optimization problem, parameters are initialized and reference points on the hyperplane are determined. The calculation formula for the number of reference points Q is as follows: Where M represents the dimension of the target vector, and H represents the number of target divisions; S432: Construct the multi-objective mining truck online scheduling model to obtain the initial population P 0 ; S433: Through the NSGA-III algorithm, the parent population P is formed t After that, the tournament selection mechanism was introduced from P t Select parent individuals from the generation to construct the offspring population Q t ; S434: For the P t , Q t The population R after merging t Perform non-dominated rank sorting; S435: R t Perform adaptive normalization, individual association reference points, and microhabitat preservation operations, and select the dominant individuals to enter the next generation population P after environmental selection. t+1 ; S436: repeat steps S433 to S435 until the maximum number of iterations is reached and then proceed to step S437; S437: Calculate and obtain the optimal compromise solution, specifically: Calculate a set of Pareto solutions for the dynamic constrained multi-objective optimization problem, and the jth objective value f of the i-th Pareto solution ij The membership function h ij The calculation formula is: Among them, f jmax represents the maximum value of the jth objective function, f jmin Represents the minimum value of the jth objective function; For the i-th Pareto solution, its standardized membership function h i The calculation formula is: Where I is the total number of Pareto solutions; Select the membership function h i The solution with the largest value is the compromise optimal solution; S438: Sending the scheduling plan corresponding to the compromise optimal solution to the mining card through the instruction system; S439: Repeat steps S432 to S438 until the cumulative transportation volume of the mining truck reaches the set transportation target.

9. A multi-objective mining truck online scheduling system for real-time ore matching, used to implement the multi-objective mining truck online scheduling method for real-time ore matching as claimed in any one of claims 1 to 8, It is characterized in that include: An actual production environment information acquisition module is used to acquire actual production environment information; An optimal path calculation module, used to calculate and obtain the optimal path from each loading point to each unloading point through the actual production environment information; A multi-objective mining truck online scheduling model construction module is used to build a multi-objective mining truck online scheduling model for real-time ore distribution; The scheduling scheme generation module is used to solve the multi-objective mining truck online scheduling model through the optimal path, constraint processing and NSGA-III algorithm to obtain a scheduling scheme.

Images