A method and system for planning the movement of a loader in an open-pit mine for simulated ore blending.

Abstract

This invention provides a method and system for planning the movement of a loader in an open-pit mine for simulated ore blending. The method includes: acquiring actual mine operating environment information; calculating the optimal blast pile and ore output task information to minimize ore grade fluctuations based on the actual operating environment information; obtaining a loading operation set that covers each blast pile using a minimum set coverage and greedy algorithm; constructing a loader movement planning model for simulated ore blending; and solving the loader movement planning model using the blast pile and ore output task information, constraint processing, and the DCNSGA-III algorithm to obtain the loader's movement route. The beneficial effects of this invention are: solving the problems of low-grade ore waste and large fluctuations in ore grade caused by difficult ore blending, as well as the low efficiency and high energy consumption caused by unplanned movement. The most direct movement planning route is of great significance for improving the utilization rate of low-grade ore and the economic benefits of enterprises.

Description

Technical Field

[0001] This invention relates to the field of mine operation optimization, and in particular to a method and system for planning the movement of open-pit shovels in simulated ore blending. Background Technology

[0002] To improve the comprehensive utilization rate of ore resources and ensure that low-grade ore can meet certain performance requirements, it is necessary to use ore blending to control grade fluctuations and maintain stable ore production. However, the processes involved in ore blending are complex, and few studies have addressed this issue. Most research focuses only on the impact of truck scheduling on the blending process, neglecting the fact that the loader's travel path is a direct factor affecting the ore grade sequence. Therefore, research on loader travel planning is of great significance for improving the utilization rate of lean ore and enhancing the economic benefits of enterprises.

[0003] Regarding the ore blending problem, while some scholars have proposed planning the movement paths of the ore trucks to determine their arrival order at the crushing station, this approach overlooks the potential issue of all ore trucks arriving at the crushing station with grades higher or lower than the station's required crushing grade. This can lead to waste of high-quality ore and increased tailings. To minimize grade fluctuations upon arrival at the crushing station, refined ore blending is necessary. This requires not only planning the ore trucks themselves but also coordinating the planning of other equipment, such as loaders, to ensure that the grade of the ore loaded into the trucks meets the blending requirements. This prevents situations where all loaded ore fails to meet the blending criteria upon arrival at the crushing station.

[0004] In path planning problems, finding the optimal path typically involves using approximate search algorithms or metaheuristic algorithms to find a route that passes through all city points in a set and passes through each city point only once. However, this assumes that all points in the set are known. In actual production processes, excavation points (city points) are not pre-defined. Therefore, in mines, the challenge lies in dividing the blasted debris and finding loading points.

[0005] Most researchers treat the ore blending problem as a single-objective problem, but real-world production problems often involve multiple objectives. In the ore blending process, multiple loaders operate simultaneously, and to meet the grade requirements of the blended ore, coordinated scheduling between different loader operations is necessary. Therefore, a multi-objective optimization model for multi-loader scheduling is urgently needed to solve this problem. However, there are relatively few algorithms for this purpose, and no concrete and feasible algorithmic framework exists.

[0006] Most existing ore blending systems on the market are monolithic architectures. While these are simple in structure and have low deployment costs, the high coupling between functional modules makes later maintenance and upgrades difficult. The loader-driven shovel system to be built not only needs to be designed with more user-friendly functions to meet real-world requirements, but also needs to resolve the coupling between modules to facilitate future service upgrades and expansions. Summary of the Invention

[0007] To address the aforementioned problems, this invention provides a method for planning the movement of a loader in an open-pit mine based on simulated ore blending, mainly comprising:

[0008] S1: Obtain actual mine operating environment information, which includes: blast pile information, grade and location information of different blocks within the blast pile, number of different types of loaders, maximum production capacity and production plan of different types of loaders, and the production plan includes the ore output and ore grade information for each shift.

[0009] S2: Calculate and obtain the blasting and ore output task information that minimizes ore grade fluctuations based on the actual working environment information;

[0010] S3: Combine minimum set coverage and greedy algorithm to obtain the shovel operation set for each blast pile obtained in step S2 that can be covered;

[0011] S4: Construct a plan for the movement of a loader in an open-pit mine for simulated ore blending;

[0012] S5: Based on the shovel loading operation set, the shovel driving planning model is solved by the blasting pile and ore output task information, constraint processing and the third-generation dynamic constraint non-dominated sorting genetic algorithm (DCNSGA-III) to obtain the shovel driving route.

[0013] Furthermore, the specific implementation process of step S3 is as follows:

[0014] S31: Based on the height of the blast pile and the load capacity of the mining truck obtained in step S2, divide the blast pile into grids along the two adjacent sides and number them. The center of each grid can be used as the loading point of the loader.

[0015] S32: Based on the loader's loading operation point and the loader's maximum digging radius, the grid that can be covered is taken as the digging area of ​​each operation point;

[0016] S33: Based on the coordinate information of the loader's loading operation point and the blasting area, calculate the effective operating area of ​​each operation location point within the excavable area of ​​each operation point, and form a set of effective operating areas;

[0017] S34: Using a greedy algorithm, select the largest working area from the set of effective working areas, add its corresponding working location point to the loader loading work set, and delete the area from the set of effective working areas. Repeat step S34 until the set of effective working areas is empty. At this point, the blast pile has been completely excavated.

[0018] Furthermore, in step S4, by analyzing the key factors affecting the large fluctuations in the grade of the feed ore and the low excavation efficiency and high energy consumption of the loader, an open-pit loader movement planning model for simulated ore blending is constructed, and the optimization objectives and constraints to be satisfied by the loader movement planning model are determined.

[0019] Furthermore, the objectives that need to be optimized in the loader traveling planning model include:

[0020] The primary optimization objective is to minimize the total distance traveled by the loader, F1(X).

[0021] min F1(X)=f1(x1)+f2(x2)+…+f M (xM)

[0022]

[0023] Where X represents the loader travel scheduling scheme variable, M represents the number of piles that explode, and x M f represents the shovel path sequence of the burst pile M. s (x s ) represents the path of the s-th shovel loader, l s Let d be the number of loader loading points for the s-th pile explosion. i,i+1 Let the distance between the i-th and (i+1)-th job points in set D be Euclidean distance. Let x and y represent the x and y coordinates of the i-th point in set D, respectively. Let x and y represent the x and y coordinates of the (i+1)th point in set D, respectively.

[0024] The second optimization objective is to minimize the deviation of ore grade after entering the crushing station, and to minimize F2:

[0025]

[0026] Where v represents the ore-producing point within the burst pile, max(d) represents the maximum dimension among all ore-producing points, and g v This represents the grade value of the v-th ore extraction point. This represents the grade value of the v-th ore extraction point in the k-th blast pile. Let G be the total amount of ore extracted from the v-th ore extraction point at the k-th blast pile, and G be the target ore grade.

[0027] Furthermore, the constraints that the loader traveling planning model must satisfy include:

[0028] First constraint condition, ore-rock boundary constraint:

[0029]

[0030] in, E represents the total amount of ore extracted from the v-th ore-producing point in the k-th blast pile. U x represents the total amount of waste rock. i Let U represent the shovel path sequence of the blast pile i, and let U represent the waste rock region;

[0031] The second constraint, a geometric constraint, states that the size of the two loading points during the loading process must be less than the maximum digging radius of the loader.

[0032]

[0033] Where r represents the digging radius of different models of electric shovels, and T ix ,T iy These represent the x and y coordinates corresponding to the i-th loading point, respectively.

[0034] The third constraint is the ore extraction point constraint, which means that the initial ore extraction point of the loader must be on the edge of the blast zone:

[0035]

[0036] Where w represents the coordinate value of the explosion pile in the y-axis of the two-dimensional plane. This indicates that the initial ore-producing point of the burst pile s is at the lower boundary of the burst pile.

[0037] Furthermore, the specific implementation process of step S5 is as follows:

[0038] S51: By using constraint-default value and niche treatment, the solution of the loader traveling planning model is constructed as a dynamic constrained multi-objective optimization problem with four optimization objectives;

[0039] S52: Based on the actual operation of the loader scheduling, design the encoding of the solution to the dynamic constrained multi-objective optimization problem to obtain the scheduling scheme variable X:

[0040] S53: Solve the dynamic constrained multi-objective optimization problem iteratively based on the DCNSGA-III algorithm, represent the solution of the dynamic constrained multi-objective optimization problem through the above encoding method, and output the shovel-moving scheme.

[0041] Furthermore, the specific implementation process of step S51 is as follows:

[0042] S511: The normalized average degree of constraint violation of the unsolved variable x on all constraints is taken as the target value of the violation value cv(x), as shown in the following formula:

[0043]

[0044] Where h represents the number of constraints, P0 represents the initial population, and G i (x) represents the degree to which x violates the constraint under the i-th constraint condition;

[0045] S512: The niche formula is as follows:

[0046]

[0047]

[0048]

[0049] in, Let represent the set of parent and offspring populations, 2N be the population size, σ be the niche radius, nc(x|U,σ) represent the niche target, and sh(x,x) be the target population. i ) represents a shared function between two different individuals, sh(x1,x2) is The shared function between two individuals, d(x1,x2), is the Euclidean distance between individuals x1 and x2;

[0050] S513: The expression for the dynamically constrained multi-objective optimization problem is:

[0051] min F1(X)=f1(x1)+f2(x2)+…+f M (x M )

[0052]

[0053]

[0054] min F3 = min cv(x)

[0055] min F4 = min nc(x|U, σ)

[0056]

[0057]

[0058] Where M represents the number of heap bursts, x M f represents the shovel path sequence of the burst pile M. s (x s ) represents the path of the s-th shovel loader, l s d represents the number of loader loading points for the explosive pile s. i,i+1Let the distance between the i-th and (i+1)-th job points in set D be Euclidean distance. Let x and y represent the x and y coordinates of the i-th point in set D, respectively. Let x and y represent the x and y coordinates of the (i+1)th point in set D, respectively; v represents the ore-producing point within the ore-producing pile; and max(d) represents the maximum dimension among all ore-producing points. Let V be the grade value of the v-th excavation point in the k-th burst pile. Let G be the total amount of ore extracted at the v-th mining point in the k-th blast pile, and G be the target ore grade. E represents the total amount of material excavated at the v-th excavation point in the k-th burst pile. U The total amount of waste rock is represented by U, the waste rock area is represented by r, and the digging radius of different types of electric shovels is represented by T. ix ,T iy Let x and y represent the x and y coordinates corresponding to the i-th shovel loading point, respectively; w represents the coordinate of the blast pile in the y-axis of the two-dimensional plane. This indicates that the initial ore-producing point of the burst pile s is at the lower boundary of the burst pile; ε (t) Represents the dynamic constraint boundary, ε1 (t) ε2 (t) ε3 (t) Let θ(x) represent the ore-rock boundary constraint (1), geometric constraint (2), and ore-exit point constraint (3), respectively. Let t represent the number of environmental changes and T represent the maximum number of environmental changes, satisfying θ(x)≤ε. (t) A solution that is ε-feasible is called an ε-feasible solution; otherwise, it is called an ε-infeasible solution.

[0059] Furthermore, the expression for the variable X of the loader traveling scheduling scheme is:

[0060] X = {x1, x2, x3, ..., x} M}

[0061] x i ={x i1 x i2 x i3 , ...x iN_i}i∈[1,M]

[0062] Where the optimization variable X represents a solution, M represents the number of heap bursts, and x i Let X represent the sequence of shovel paths for the i-th burst pile, where N_i represents the dimension of the i-th burst pile. That is, a solution X contains M burst piles, each with a different size, resulting in a shovel placement point x. iN_i The number of each type is also different, where i represents the burst pile number.

[0063] Furthermore, the specific implementation process of step S53 is as follows:

[0064] S531: For the aforementioned dynamically constrained multi-objective optimization problem, perform parameter initialization and determine reference points on the hyperplane. The formula for calculating the number of reference points Q is as follows:

[0065]

[0066] Where M represents the dimension of the target vector, H represents the number of parts of the target, and C represents the combination;

[0067] S532: Obtain the initial population P0 using the aforementioned loader walking planning model;

[0068] S533: Using the DCNSGA-III algorithm, in the formation of the parent population P t Subsequently, a tournament selection mechanism was introduced from P t Selecting parent individuals to construct offspring population Q t ;

[0069] S534: Regarding the P t Q t The merged population R t Perform non-dominant hierarchy sorting;

[0070] S535: For R t Adaptive normalization, individual association reference points, and niche preservation operations are performed. After environmental selection, dominant individuals are selected for the next generation population R. t+1 ;

[0071] S536: Repeat steps S533-S535 until the maximum number of iterations is reached, then proceed to step S537.

[0072] S537: The compromise optimal solution is calculated as follows:

[0073] Calculate and obtain a set of Pareto solutions to a dynamically constrained multi-objective optimization problem, where the j-th objective value f of the i-th Pareto solution is... ij Its own membership function h ij The calculation formula is:

[0074]

[0075] Among them, f jmax Let f represent the maximum value of the j-th objective function. jmin Let represent the minimum value of the j-th objective function;

[0076] For the i-th Pareto solution, its standardized membership function h i The calculation formula is:

[0077]

[0078] Where I is the total number of Pareto solutions;

[0079] Choose the membership function h i The solution with the largest value is the compromise optimal solution;

[0080] S538: The loader travel plan corresponding to the compromise optimal solution is sent to the loader through the command system;

[0081] S539: Repeat steps S532-S538 until the cumulative excavation volume of the loader reaches the set transportation target.

[0082] A loader travel planning system for simulated ore blending in open-pit mines includes:

[0083] The actual operating environment information acquisition module is used to acquire actual production environment information, which includes: blast pile information, grade and location information of different blocks within the blast pile, number of different types of loaders, maximum production capacity of different types of loaders, and production plan; the production plan includes the ore output and ore grade information for each shift.

[0084] The mineral processing point subsystem module is used to allocate the blasting pile that minimizes fluctuations in ore grade and the ore output task for each loader per shift.

[0085] The loader movement planning model construction module is used to construct an open-pit mine loader movement planning model for simulated ore blending. The open-pit mine loader movement model is solved by the blasting and ore output task information, constraint processing and DCNSGA-III algorithm to obtain the scheduling scheme.

[0086] The scheduling scheme preference decision module is used to set the focus of DCNSGA-III algorithm optimization based on the user's decision preferences and production indicators. The decision preferences include the selected path, time, and energy consumption.

[0087] The loader dispatching plan modification module allows operators to flexibly adjust the plan and change the loader path based on the actual excavation situation and possible unexpected events, according to their own experience and the requirements of each shift.

[0088] Compared with existing technologies, the beneficial effects of the technical solution provided by this invention are:

[0089] 1. By utilizing the minimum set covering problem, combined with the loader's equipment parameters and the shape of the blast pile, we can find the minimum set of loader loading points that can cover the entire blast pile. By traversing the loading points in the set, unnecessary paths during the excavation process can be reduced, minimizing the loader's travel distance while meeting the ore blending requirements. This reduces energy consumption during loader movement and improves excavation efficiency, bringing economic benefits to the enterprise.

[0090] 2. To address the ore blending problem and reduce energy consumption in enterprises, a multi-objective, multi-constraint model is designed and applied to this problem, taking into account actual development conditions and factors such as equipment quantity, parameters, and waste rock disposal as constraints. Several heuristic search operators are designed based on the problem's characteristics. Finally, multi-objective optimization mechanisms and dynamic optimization mechanisms are introduced into the constraint optimization algorithm, forming a dynamic-constrained multi-objective evolutionary algorithm framework for solving the loader path planning problem for ore blending.

[0091] 3. Design a loader ore blending production scheduling subsystem based on actual conditions. The system takes daily ore production plans, equipment parameters, ore data, and user preferences as input information, and automatically generates multiple ore blending schemes using a background optimization algorithm. Operators can also flexibly schedule loader paths and ore blending indicators based on their experience to meet the actual production needs of the day. Simultaneously, they can query employee work status and equipment status online, rationally allocate on-duty personnel, and view the adjusted ore supply and grade changes, providing decision-making assistance to mine operators. Attached Figure Description

[0092] The present invention will be further described below with reference to the accompanying drawings and embodiments. In the accompanying drawings:

[0093] Figure 1 This is a flowchart of an open-pit mine shovel travel planning method for simulated ore blending, as described in an embodiment of the present invention.

[0094] Figure 2 This is a schematic diagram of the shovel-moving sequence generated by the shovel moving within the blast pile in an embodiment of the present invention. Detailed Implementation

[0095] To provide a clearer understanding of the technical features, objectives, and effects of the present invention, specific embodiments of the present invention will now be described in detail with reference to the accompanying drawings.

[0096] This invention provides a method for planning the movement of a loader in an open-pit mine for simulated ore blending. It establishes a loader movement planning model with the objectives of minimizing the excavation path and grade deviation. The solution is encoded based on the problem characteristics, and a heuristic search operator is designed. Furthermore, the model is solved using the DCNSGA-III algorithm framework, taking into account the model's features. This addresses the problems of large ore grade fluctuations, low utilization of lean ore, and high loader energy consumption. The most direct movement planning route is of great significance for improving the utilization rate of lean ore and enhancing the economic benefits of enterprises.

[0097] Please refer to Figure 1 , Figure 1 This is a flowchart of a method for planning the movement of a loader in an open-pit mine for simulated ore blending, as described in an embodiment of the present invention. Specifically, it includes:

[0098] S1: Obtain actual mine operating environment information through the mine simulation ore blending system and background statistical information. The actual mine operating environment information includes: blast pile information, grade and location information of different blocks in the blast pile, number of different types of loaders, maximum production capacity and production plan of different types of loaders, and the production plan includes the ore output and ore grade information for each shift.

[0099] S2: Calculate and obtain the blasting and ore output task information that minimizes ore grade fluctuations based on the actual working environment information;

[0100] The determination of the blast pile and ore output task is achieved by the upper-level ore output point subsystem based on the grade deviation and ore content between different blast piles, combined with the road network structure, through the shovel planning of the middle-level loader and the ore truck flow planning of the lower level, and using the DCNSGA-III algorithm to calculate the appropriate blast pile, and in combination with the optimization objective, allocating the daily task volume to each loader (i.e. calculating the amount of excavation required in each blast pile).

[0101] S3: Combine minimum set coverage and greedy algorithm to obtain the shoveling operation set for each overflow pile obtained in step S2; specifically:

[0102] S31: Based on the height of the blast pile and the load capacity of the mining truck obtained in step S2, divide the blast pile into grids along the two adjacent sides and number them. The center of each grid can be used as the loading point of the loader.

[0103] S32: Based on the loader's loading operation point and the loader's maximum digging radius, the grid that can be covered is taken as the digging area of ​​each operation point;

[0104] S33: Based on the coordinate information of the loader's loading operation point and the blasting area, calculate the effective operating area of ​​each operation location within the excavable area of ​​each operation point, forming a set of effective operating areas; the calculation process is as follows:

[0105] First, the blast pile is divided into a two-dimensional grid. Then, the coordinates of each grid point are determined, such as (0,0) and (0,1). Next, considering the maximum digging radius of the loader, the area covered by the grid point (0,0) is determined. This covered area is then added to the effective working area set (calculated using Euclidean distance; if the distance is less than the loader's maximum digging radius, the loader can dig into the area of ​​that grid point). Similarly, the areas covered by all grid points can be determined sequentially using the above method, thus obtaining the effective working area set for all grid working locations.

[0106] S34: Using a greedy algorithm, select the largest working area from the set of effective working areas (that is, take out the grid point with the most coverage area from the set of effective working areas), put its corresponding working position point into the loader loading work set, and delete the area from the set of effective working areas. Repeat this operation until the set of effective working areas is empty. At this time, the pile excavation is completed.

[0107] The TSP problem seeks to find the path between all cities. In this step, the loading operation set is determined by combining the equipment parameters of the loader and identifying several loader working points (the entire blast pile can be excavated by visiting all the working points in the set). If we treat the loader working points as similar to city points, we can solve the TSP problem and obtain the shortest movement route of the loader. The optimization variables in the optimization objective are established on this basis.

[0108] S4: Construct a plan for the movement of a loader in an open-pit mine for simulated ore blending;

[0109] By analyzing the key factors affecting the large fluctuations in the grade of the feed ore and the low efficiency and high energy consumption of the loader excavation, and taking the minimization of the total travel distance and grade deviation of the loader excavation as the objective, a loader travel planning model for open-pit mines oriented towards simulated ore blending was constructed by building constraints on the ore-rock boundary line, geometric constraints, and ore outlet constraints through actual excavation technology. The optimization objectives and constraints that the loader travel planning model needs to satisfy were also determined.

[0110] The objectives that need to be optimized in the loader travel planning model include:

[0111] The primary optimization objective is to minimize the total distance traveled by the loader, F1(X).

[0112] min F1(X)=f1(x1)+f2(x2)+…+f M (x M (1)

[0113]

[0114] Where X represents the loader travel scheduling scheme variable, M represents the number of piles that explode, and x M f represents the shovel path sequence of the burst pile M. s (x s ) represents the path of the s-th shovel loader, l s Let d be the number of loader loading points for the s-th pile explosion. i，i+1 Let the distance between the i-th and (i+1)-th job points in set D be Euclidean distance. Let x and y represent the x and y coordinates of the i-th point in set D, respectively. Let x and y represent the x and y coordinates of the (i+1)th point in set D, respectively.

[0115] The second optimization objective is to minimize the deviation of ore grade after entering the crushing station, and to minimize F2:

[0116]

[0117] Where v represents the ore-producing point within the burst pile, max(d) represents the maximum dimension among all ore-producing points, and g v This represents the grade value of the v-th ore extraction point. This represents the grade value of the v-th ore extraction point in the k-th blast pile. Let G be the total amount of ore extracted from the v-th ore extraction point at the k-th blast pile, and G be the target ore grade.

[0118] The constraints that the loader traveling planning model must satisfy include:

[0119] First constraint condition, ore-rock boundary constraint:

[0120]

[0121] in, E represents the total amount of ore extracted from the v-th ore extraction point at the k-th blast pile (only ore is counted, excluding waste rock). U x represents the total amount of waste rock. i Let U represent the shovel path sequence for pile i, and let U represent the waste rock region. x represents i This sequence does not belong to the waste rock area;

[0122] The second constraint, a geometric constraint, states that the size of the two loading points during the loading process must be less than the maximum digging radius of the loader.

[0123]

[0124] Where r represents the digging radius of different models of electric shovels, and T ix， T iy These represent the x and y coordinates corresponding to the i-th loading point, respectively;

[0125] The third constraint is the ore extraction point constraint, which means that the initial ore extraction point of the loader must be on the edge of the blast zone:

[0126]

[0127] Where w represents the coordinate value of the explosion pile in the y-axis of the two-dimensional plane. This indicates that the initial ore-producing point of the burst pile s is at the lower boundary of the burst pile.

[0128] S5: The loader's movement planning model is solved using the information on the blast pile and ore output task, constraint processing, and the DCNSGA-III algorithm to obtain the loader's movement route. The specific process is as follows:

[0129] S51: By using constraint-default value and niche treatment, the solution of the loader traveling planning model is constructed as a dynamic constrained multi-objective optimization problem with four optimization objectives;

[0130] S52: Based on the actual operation of the loader scheduling, design the encoding of the solution to the dynamic constrained multi-objective optimization problem to obtain the scheduling scheme variable X:

[0131] S53: Solve the dynamic constrained multi-objective optimization problem iteratively based on the DCNSGA-III algorithm, represent the solution of the dynamic constrained multi-objective optimization problem through the above encoding method, and output the shovel-moving scheme.

[0132] The specific implementation process of step S51 is as follows:

[0133] S511: The normalized average degree of constraint violation of the unsolved variable x on all constraints is used as the target value of the violation value, as shown in the following formula:

[0134]

[0135] Where h represents the number of constraints, P0 represents the initial population, and G i (x) represents the degree to which x violates the constraint under the i-th constraint condition;

[0136] S512: The niche formula is as follows:

[0137]

[0138]

[0139]

[0140] in, Let represent the set of parent and offspring populations, 2N be the population size, σ be the niche radius, nc(x|U, σ) represent the niche target, and sh(x, x) be the target value. i ) represents a shared function between two different individuals, sh(x1, x2) is The shared function between two individuals, d(x1, x2), is the Euclidean distance between individuals x1 and x2;

[0141] S513: The expression for the dynamically constrained multi-objective optimization problem is:

[0142] min F1(X)=f1(x1)+f2(x2)+…+fM (x M )

[0143]

[0144]

[0145] min F3=mincv(x) (11)

[0146] min F4=min nc(x|U,σ) (12)

[0147]

[0148]

[0149] Where M represents the number of heap bursts, x M f represents the shovel path sequence of the burst pile M. s (x s ) represents the path of the s-th shovel loader, l s d represents the number of loader loading points for the explosive pile s. i，i+1 Let the distance between the i-th and (i+1)-th job points in set D be Euclidean distance. Let x and y represent the x and y coordinates of the i-th point in set D, respectively. Let x and y represent the x and y coordinates of the (i+1)th point in set D, respectively; v represents the ore-producing point within the burst pile; and max(d) represents the maximum dimension among all ore-producing points. Let V be the grade value of the v-th excavation point in the k-th burst pile. Let G be the total amount of ore extracted at the v-th mining point in the k-th blast pile, and G be the target ore grade. E represents the total amount of material excavated at the v-th excavation point in the k-th burst pile. U The total amount of waste rock is represented by U, the waste rock area is represented by r, and the digging radius of different types of electric shovels is represented by T. ix T iy Let x and y represent the x and y coordinates corresponding to the i-th shovel loading point, respectively; w represents the coordinate of the blast pile in the y-axis of the two-dimensional plane. This indicates that the initial ore-producing point of the burst pile s is at the lower boundary of the burst pile; ε (t) Denotes the dynamic constraint boundary, ε1 (t) ε2 (t) ε3 (t) Let represent the ore-rock boundary constraint, geometric constraint, and ore-extraction point constraint, respectively; t represent the number of environmental changes; and T represent the maximum number of environmental changes, satisfying θ(x)≤ε. (t) A solution that is ε-feasible is called an ε-feasible solution; otherwise, it is called an ε-infeasible solution.

[0150] refer to Figure 2In this embodiment, during the mining process, due to blasting, each blast pile has a different shape, and the loader moves through the blast pile to perform excavation operations. Combining step S3 above, the minimum set covering problem and a greedy algorithm are used to find the loading operation set. By traversing the loading operation points in the set, the entire blast pile can be excavated. Because the size of the blast pile varies, the number of loading points will also vary. The loading points are represented by numbers 1, 2, ..., N_i, and different combinations can generate different loading paths x. i For example, the sequence of explosions 1, 5, 2, 4, 3, or 5, 1, 4, 2, 3. X represents a solution containing the sequence of all loader paths, and M represents the number of explosions.

[0151] The expression for the variable X in the traveling loader scheduling scheme is:

[0152] X = {x1, x2, x3, ..., x} M} (13)

[0153] x i ={x i1 x i2 x i3 , ...x iN_i}i∈[1,M] (14)

[0154] Where the optimization variable x represents a solution, M represents the number of burst heaps, and x i Let X represent the sequence of shovel paths for the i-th burst pile, where N_i represents the dimension of the i-th burst pile. That is, a solution X contains M burst piles, each with a different size, resulting in a shovel placement point x. iN_i The number of each type is also different, with i representing the burst pile number.

[0155] The specific implementation process of step S53 is as follows:

[0156] S531: For the aforementioned dynamically constrained multi-objective optimization problem, perform parameter initialization and determine reference points on the hyperplane. The formula for calculating the number of reference points Q is as follows:

[0157]

[0158] Where M represents the dimension of the target vector, H represents the number of parts of the target, and C represents the combination;

[0159] S532: Obtain the initial population P0 using the aforementioned loader walking planning model;

[0160] During the initial run, an initial population P0 is generated based on the established open-pit mine loader movement planning model for simulated ore blending. Subsequent runs generate the same initial population P0 based on the execution of scheduling schemes by all loaders.

[0161] S533: Using the DCNSGA-III algorithm, in the formation of the parent population P t Subsequently, a tournament selection mechanism was introduced from P t Selecting parent individuals to construct offspring population Q t ;

[0162] In the specific implementation of this step, based on the characteristics of the problem, the following four search operators are designed for tournament selection: LS1-LS3 are non-heuristic search operators, and LS4 is a heuristic search operator:

[0163] LS1: Any random cross-cutting of the ore extraction order of a single shovel (excluding the initial ore extraction point) can generate a new solution;

[0164] LS2: Randomly swap two shovel loading points. For example, if the original path is 1, 2, 3, 4, 5, 6, 7, when 2 and 5 are swapped, ensure that 3 is still adjacent to 2. The sequence will then become 1, 5, 4, 3, 2, 6, 7. After the sequence changes, compare the path sum with the original sequence. If the path sum is shorter, update the path. Repeat the above steps, setting the number of loops. Once the number of loops is reached, exit the loop.

[0165] LS3: First, randomly select two individuals P1 and P2 from the parent generation; then select different cut points, which must be different in both position and quantity. Copy the cut points in p1 to O1 with the positions unchanged, and fill the remaining numbers in p2 (excluding the cut points) into the remaining positions in O1 in order; the same applies to O2.

[0166] LS4: Combining path distance and grade deviation for variation, after the first few schemes are determined, the last scheme is calculated based on the grade deviation to determine the movement of the last loader. If the scheme does not meet the requirements, the remaining position points of the current scheme are determined according to the shortest distance, and the movement schemes of other blast piles are modified again until the scheme is modified.

[0167] S534: Regarding the P t Q t The merged population R t Perform non-dominant hierarchy sorting;

[0168] S535: For R t Adaptive normalization, individual association reference points, and niche preservation operations are performed. After environmental selection, dominant individuals are selected for the next generation population R. t+1 ;

[0169] The adaptive normalization process is as follows:

[0170] The minimum values ​​of the four established objective functions are calculated. Let's assume the minimum value obtained on objective axis i is... and The set is the ideal point set mentioned in the DCNSGA-III algorithm, and then the objective function value is transformed using equation (16);

[0171]

[0172] To find the extreme points, a scalarization function (ASF) as shown in equation (17) is required.

[0173]

[0174] Where e i The target axis f i If the axis direction is such that 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 To replace; traverse each objective function, find the individual with the lowest ASF value, and form the extreme point. The extreme point and the origin (ideal point) form three lines, which can form a hyperplane. The intersection of the hyperplane and the coordinate axes is the intercept to be found. Then, normalization is performed by equation (18). It is the normalized objective function value;

[0175]

[0176] The process for linking individual reference points is as follows:

[0177] After normalization, individuals need to be associated with reference points. A line formed by the reference point and the origin is used as a baseline. All reference lines are then traversed to find the reference line closest to each individual, and the corresponding reference point and shortest distance are recorded. The number of individuals associated with each reference point is then calculated.

[0178] The procedure for microhabitat preservation is as follows:

[0179] Define S t+1 Given a set containing all individuals from non-dominated level 1 to non-dominated level L, then iterate through each reference point, checking if the reference point is dominated by S. t+1 The number of times ρ is referenced by individuals other than the dominant hierarchy L. j First, determine if any individual is associated with this reference point; if not, find a different reference point. If a reference point is associated with it, then determine ρ. j If ρ j=0, then the solution with the minimum distance from the non-dominated level L to the reference point j is added to P. t+1 If ρ j If the value is greater than 1, then a solution randomly selected and associated with the non-dominated level L reference point will be added to P. t+1 Middle; until P t+1 The number of individuals is equal to the size of the original population;

[0180] S536: Repeat steps S533-S535 until the maximum number of iterations is reached, then proceed to step S537.

[0181] S537: The compromise optimal solution is calculated as follows:

[0182] Calculate and obtain a set of Pareto solutions to a dynamically constrained multi-objective optimization problem, where the j-th objective value f of the i-th Pareto solution is... ij Its own membership function h ij The calculation formula is:

[0183]

[0184] Among them, f jmax Let f represent the maximum value of the j-th objective function. jmin Let represent the minimum value of the j-th objective function;

[0185] For the i-th Pareto solution, its standardized membership function h i The calculation formula is:

[0186]

[0187] Where I is the total number of Pareto solutions;

[0188] Choose the membership function h i The solution with the largest value is the compromise optimal solution;

[0189] S538: The loader travel plan corresponding to the compromise optimal solution is sent to the loader through the command system;

[0190] S539: Repeat steps S532-S538 until the cumulative excavation volume of the loader reaches the set transportation target.

[0191] A loader travel planning system for simulated ore blending in open-pit mines includes:

[0192] The actual working environment information acquisition module is used to acquire actual production environment information;

[0193] The mineral processing point subsystem module is used to allocate the blasting pile that minimizes fluctuations in ore grade and the ore output task for each loader per shift.

[0194] The loader movement planning model construction module is used to construct a scheduling scheme generation module for the open-pit loader movement planning model oriented to simulate ore blending. It is used to solve the open-pit loader movement model by using the blasting and ore output task information, constraint processing and DCNSGA-III algorithm to obtain the scheduling scheme.

[0195] The scheduling scheme preference decision module is used to set the focus of DCNSGA-III algorithm optimization based on the user's decision preferences and production indicators. Preference selection includes path, time, and energy consumption, etc.

[0196] In this module, the decision preference ratio can be manually adjusted on the system interface, and the parameter ratio is passed into the DCNSGA-III algorithm. By allocating weight to different optimization objectives, the final optimization of the loader walking scheme with decision preference is obtained.

[0197] The loader dispatching plan modification module allows operators to flexibly adjust the plan and change the loader path based on the actual excavation situation and potential unforeseen events, according to their experience and the requirements of each shift.

[0198] Emergency warnings can be set on the system interface. When the current fuel level is insufficient to travel to all loading points or the current loading point is unmovable, the staff can manually adjust the loader's path.

[0199] The beneficial effects of this invention are:

[0200] 1. By utilizing the minimum set covering problem, combined with the loader's equipment parameters and the shape of the blast pile, we can find the minimum set of loader loading points that can cover the entire blast pile. By traversing the loading points in the set, unnecessary paths during the excavation process can be reduced, minimizing the loader's travel distance while meeting the ore blending requirements. This reduces energy consumption during loader movement and improves excavation efficiency, bringing economic benefits to the enterprise.

[0201] 2. To address the ore blending problem and reduce energy consumption in enterprises, a multi-objective, multi-constraint model is designed and applied to this problem, taking into account actual development conditions and factors such as equipment quantity, parameters, and waste rock disposal as constraints. Several heuristic search operators are designed based on the problem's characteristics. Finally, multi-objective optimization mechanisms and dynamic optimization mechanisms are introduced into the constraint optimization algorithm, forming a dynamic-constrained multi-objective evolutionary algorithm framework for solving the loader path planning problem for ore blending.

[0202] 3. Design a loader ore blending production scheduling subsystem based on actual conditions. The system takes daily ore production plans, equipment parameters, ore data, and user preferences as input information, and automatically generates multiple ore blending schemes using a background optimization algorithm. Operators can also flexibly schedule loader paths and ore blending indicators based on their experience to meet the actual production needs of the day. Simultaneously, they can query employee work status and equipment status online, rationally allocate on-duty personnel, and view the adjusted ore supply and grade changes, providing decision-making assistance to mine operators.

[0203] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A method for planning the movement of a loader in an open-pit mine for simulated ore blending, characterized in that: include: S1: Obtain actual mine operating environment information, which includes: blast pile information, grade and location information of different blocks within the blast pile, number of different types of loaders, maximum production capacity and production plan of different types of loaders, and the production plan includes the ore output and ore grade information for each shift. S2: Calculate and obtain the blasting and ore output task information that minimizes ore grade fluctuations based on the actual working environment information; S3: Combine minimum set coverage and greedy algorithm to obtain the shovel operation set for each blast pile obtained in step S2 that can be covered; S4: Construct a plan for the movement of a loader in an open-pit mine for simulated ore blending; S5: Based on the shovel loading operation set, the shovel loader movement planning model is solved by the blasting pile and ore output task information, constraint processing and DCNSGA-III algorithm to obtain the shovel loader's movement route; The specific implementation process of step S5 is as follows: S51: By using constraint-default value and niche treatment, the solution of the loader traveling planning model is constructed as a dynamic constrained multi-objective optimization problem with four optimization objectives; S52: Based on the actual operation of the loader dispatching, design the encoding of the solution to the dynamic constrained multi-objective optimization problem to obtain the loader moving and dispatching scheme variable X: S53: Solve the dynamic constrained multi-objective optimization problem iteratively based on the DCNSGA-III algorithm, represent the solution of the dynamic constrained multi-objective optimization problem through the above encoding method, and output the scheduling scheme; The specific implementation process of step S53 is as follows: S531: For the aforementioned dynamically constrained multi-objective optimization problem, perform parameter initialization and determine reference points on the hyperplane. The formula for calculating the number of reference points Q is as follows: Where M represents the dimension of the target vector, H represents the number of parts of the target, and C represents the combination; S532: Obtain the initial population P0 using the aforementioned loader walking planning model; S533: Using the DCNSGA-III algorithm, in the formation of the parent population P t Subsequently, a tournament selection mechanism was introduced from P t Selecting parent individuals to construct offspring population Q t ; S534: Regarding the P t Q t The merged population R t Perform non-dominant hierarchy sorting; S535: For R t Adaptive normalization, individual association reference points, and niche preservation operations are performed. After environmental selection, dominant individuals are selected for the next generation population R. t+1 ; S536: Repeat steps S533-S535 until the maximum number of iterations is reached, then proceed to step S537. S537: The compromise optimal solution is calculated as follows: Calculate and obtain a set of Pareto solutions to a dynamically constrained multi-objective optimization problem, where the j-th objective value f of the i-th Pareto solution is... ij membership function h ij The calculation formula is: Among them, f jmax Let f represent the maximum value of the j-th objective function. jmin Let represent the minimum value of the j-th 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; Choose the membership function h i The solution with the largest value is the compromise optimal solution; S538: The loader travel plan corresponding to the compromise optimal solution is sent to the loader through the command system; S539: Repeat steps S532-S538 until the cumulative excavation volume of the loader reaches the set transportation target.

2. The method for planning the movement of a loader in an open-pit mine for simulated ore blending as described in claim 1, characterized in that: The specific implementation process of step S3 is as follows: S31: Based on the height of the blast pile and the load capacity of the mining truck obtained in step S2, divide the blast pile into grids along the two adjacent sides and number them. The center of each grid can be used as the loading point of the loader. S32: Based on the loader's loading operation point and the loader's maximum digging radius, the grid that can be covered is taken as the digging area of ​​each operation point; S33: Based on the coordinate information of the loader's loading operation point and the blasting area, calculate the effective operating area of ​​each operation location point within the excavable area of ​​each operation point, and form a set of effective operating areas; S34: Using a greedy algorithm, select the largest working area from the set of effective working areas, add its corresponding working location point to the loader loading work set, and delete the area from the set of effective working areas. Repeat step S34 until the set of effective working areas is empty. At this point, the blast pile has been completely excavated.

3. The method for planning the movement of a loader in an open-pit mine for simulated ore blending as described in claim 1, characterized in that: In step S4, by analyzing the key factors affecting the large fluctuation of the grade of the feed ore and the low excavation efficiency and high energy consumption of the loader, a loader movement planning model for simulated ore blending in open-pit mines is constructed, and the optimization objectives and constraints to be satisfied by the loader movement planning model are determined.

4. The method for planning the movement of a loader in an open-pit mine for simulated ore blending as described in claim 3, characterized in that: The objectives that need to be optimized in the loader travel planning model include: The primary optimization objective is to minimize the total distance traveled by the loader, F1(X). min F1(X)=f1(x1)+f2(x2)+…+f M (x M ) Where X represents the loader travel scheduling scheme variable, M represents the number of piles that explode, and x M f represents the shovel path sequence of the burst pile M. s (x s ) represents the path of the s-th shovel loader, l s Let d be the number of loader loading points for the s-th pile explosion. i,i+1 Let the distance between the i-th and (i+1)-th job points in set D be Euclidean distance. Let x and y represent the x and y coordinates of the i-th point in set D, respectively. Let x and y represent the x and y coordinates of the (i+1)th point in set D, respectively. The second optimization objective is to minimize the ore grade deviation F2 after entering the crushing station. Where v represents the ore-producing point within the burst pile, max(d) represents the maximum dimension among all ore-producing points, and g v This represents the grade value of the v-th ore extraction point. This represents the grade value of the v-th ore extraction point in the k-th blast pile. Let G be the total amount of ore extracted from the v-th ore extraction point at the k-th blast pile, and G be the target ore grade.

5. The method for planning the movement of a loader in an open-pit mine for simulated ore blending as described in claim 3, characterized in that: The constraints that the loader traveling planning model must satisfy include: First constraint condition, ore-rock boundary constraint: in, E represents the total amount of ore extracted from the v-th ore-producing point in the k-th blast pile. U x represents the total amount of waste rock. i Let U represent the shovel path sequence of the blast pile i, and let U represent the waste rock region; The second constraint, a geometric constraint, states that the size of the two loading points during the loading process must be less than the maximum digging radius of the loader. Where r represents the digging radius of different models of electric shovels, and T ix ,T iy These represent the x and y coordinates corresponding to the i-th loading point, respectively. The third constraint is the ore extraction point constraint, which means that the initial ore extraction point of the loader must be on the edge of the blast zone: Where w represents the coordinate value of the explosion pile in the y-axis of the two-dimensional plane. This indicates that the initial ore-producing point of the burst pile s is at the lower boundary of the burst pile.

6. The method for planning the movement of a loader in an open-pit mine for simulated ore blending as described in claim 1, characterized in that: The specific implementation process of step S51 is as follows: S511: The normalized average degree of constraint violation of the unsolved variable x on all constraints is taken as the target value of the violation value cv(x), as shown in the following formula: Where h represents the number of constraints, P0 represents the initial population, and G i (x) represents the degree to which x violates the constraint under the i-th constraint condition; S512: The niche formula is as follows: in, Let represent the set of parent and offspring populations, 2N be the population size, σ be the niche radius, nc(x|U,σ) represent the niche target, and sh(x,x) be the target population. i ) represents a shared function between two different individuals, sh(x1,x2) is The shared function between two individuals, d(x1,x2), is the Euclidean distance between individuals x1 and x2; S513: The expression for the dynamically constrained multi-objective optimization problem is: min F1(X)=f1(x1)+f2(x2)+…+f M (x M ) min F3 = min cv(x) min F4=min nc(x|U,σ) Where M represents the number of heap bursts, x M f represents the shovel path sequence of the burst pile M. s (x s ) represents the path of the s-th shovel loader, l s d represents the number of loader loading points for the explosive pile s. i,i+1 Let the distance between the i-th and (i+1)-th job points in set D be Euclidean distance. Let x and y represent the x and y coordinates of the i-th point in set D, respectively. Let x and y represent the x and y coordinates of the (i+1)th point in set D, respectively; v represents the ore-producing point within the ore-producing pile; and max(d) represents the maximum dimension among all ore-producing points. Let V be the grade value of the v-th excavation point in the k-th burst pile. Let G be the total amount of ore extracted at the v-th mining point in the k-th blast pile, and G be the target ore grade. E represents the total amount of material excavated at the v-th excavation point in the k-th burst pile. U The total amount of waste rock is represented by U, the waste rock area is represented by r, and the digging radius of different types of electric shovels is represented by T. ix ,T iy Let x and y represent the x and y coordinates corresponding to the i-th shovel loading point, respectively; w represents the coordinate of the blast pile in the y-axis of the two-dimensional plane. This indicates that the initial ore-producing point of the burst pile s is at the lower boundary of the burst pile; ε (t) Represents the dynamic constraint boundary, ε1 (t) ε2 (t) ε3 (t) Let represent the constraints of the ore-rock boundary line (1), geometric constraints (2), and ore extraction point constraints (3), respectively. t represents the number of environmental changes, and T represents the maximum number of environmental changes.

7. The method for planning the movement of a loader in an open-pit mine for simulated ore blending as described in claim 4, characterized in that: The expression for variable X in the loader traveling dispatching scheme is: X={x1,x2,x3,…,x M } x i ={x i1 ,x i2 ,x i3 ,…x iN_i }i∈[1,M] Where the optimization variable X represents a solution, M represents the number of heap bursts, and x i Let X represent the sequence of shovel paths for the i-th burst pile, where N_i represents the dimension of the i-th burst pile. That is, a solution X contains M burst piles, each with a different size, resulting in a shovel placement point x. iN_i The number of each type is also different, where i represents the burst pile number.

8. A loader movement planning system for simulated ore blending in open-pit mines, used to implement the loader movement planning method for simulated ore blending in open-pit mines as described in any one of claims 1-7, characterized in that, include: The actual mine operating environment information acquisition module is used to acquire actual mine operating environment information, which includes: blast pile information, grade and location information of different blocks within the blast pile, number of different types of loaders, maximum production capacity and production plan of different types of loaders; the production plan includes the ore output and ore grade information for each shift. The mineral processing point subsystem module is used to allocate the blasting pile that minimizes fluctuations in ore grade and the ore output task for each loader per shift. The loader movement planning model construction module is used to construct an open-pit mine loader movement planning model for simulated ore blending. The open-pit mine loader movement planning model is solved by the blasting and ore output task information, constraint processing and DCNSGA-III algorithm to obtain the scheduling scheme. The scheduling scheme preference decision module is used to set the focus of DCNSGA-III algorithm optimization based on the user's decision preferences and production indicators. The decision preferences include the selected path, time, and energy consumption. The loader dispatching plan modification module allows operators to flexibly adjust the plan and change the loader path based on the actual excavation situation and possible unexpected events, according to their own experience and the requirements of each shift.

Images