Strip mine truck multi-target low-carbon scheduling optimization method, system and equipment based on two-stage evolution selection mechanism and medium

Abstract

The invention discloses a strip mine truck multi-target low-carbon dispatching optimization method, system and device based on a two-stage evolution selection mechanism and a medium, and the method comprises the steps: constructing a strip mine multi-target truck low-carbon dispatching optimization model, and constructing an NSGA-II-TS algorithm based on the two-stage evolution selection mechanism, solving the multi-target low-carbon scheduling model of the strip mine by using an NSGA-II-TS algorithm to obtain a strip mine truck multi-target low-carbon scheduling optimization result based on a two-stage evolution selection mechanism; the system, the equipment and the medium are used for realizing the strip mine truck multi-target low-carbon scheduling optimization method based on the two-stage evolution selection mechanism. According to the method, the transportation distance and the transportation frequency in the truck dispatching process can be effectively reduced, the transportation cost of enterprises is remarkably reduced, the mine production efficiency can be improved on the premise of effectively guaranteeing the ore grade, the economic benefits of mine enterprises are remarkably improved on the basis of reducing the carbon emission, and the economic benefits are increased. And the actual requirements of the current strip mine transportation scheduling can be met.

Description

Technical Field

[0001] This invention belongs to the field of mining systems engineering and mine optimization technology, specifically relating to a multi-objective low-carbon scheduling optimization method, system, equipment and medium for open-pit mine trucks based on a two-stage evolutionary selection mechanism. Background Technology

[0002] In open-pit mining, production plans are implemented through mining and transportation equipment and their scheduling, with mining and transportation costs accounting for over 60% of total open-pit mining production costs. Currently, the country is increasingly emphasizing low-carbon, energy-saving, and green development. Since open-pit mining has a significant environmental impact and transportation consumes substantial amounts of energy, reducing energy consumption and carbon emissions during scheduling and transportation is crucial.

[0003] Patent application number [CN202110764499.3] provides "a method, system, device and storage medium for scheduling mining trucks in open-pit mines". It determines several production tasks in the mine and establishes a real-time scheduling model for mining trucks. Using an improved neighborhood search algorithm, it obtains an initial solution for real-time scheduling of mining trucks based on the several production tasks. From the initial solution, it uses a neighborhood search operator to obtain a new solution. Based on the objective function value of the real-time scheduling model, it obtains the objective function value of the new solution. It then judges the difference between the objective function value of the new solution and the current solution, and outputs the final optimal solution. Therefore, under my country's current dual-carbon policy, this method does not meet the development requirements of green mines and is difficult to realize its value in current mine practice. It suffers from high transportation costs, high carbon emissions, and low production efficiency. Summary of the Invention

[0004] To overcome the shortcomings of the existing technologies, the present invention aims to provide a multi-objective low-carbon scheduling optimization method for open-pit mine trucks based on a two-stage evolutionary selection mechanism. Based on the actual production scheduling of open-pit mines, a multi-objective low-carbon scheduling model is established with the objectives of minimizing total truck waiting time, total ore transportation cost, ore grade deviation, and total transportation carbon emissions. An NSGA-II-TS algorithm based on a two-stage evolutionary selection mechanism is constructed, and finally, the NSGA-II-TS algorithm is used to solve the multi-objective low-carbon scheduling model for open-pit mines. This effectively reduces the transportation distance and number of trips during truck scheduling, thereby significantly reducing the transportation costs of enterprises. While effectively ensuring ore grade, it can improve the production efficiency of the mine and significantly enhance the economic benefits of mining enterprises while reducing carbon emissions. Path planning and traffic flow allocation are implemented during the model construction process, which can meet the actual needs of current open-pit mine transportation scheduling.

[0005] To achieve the above objectives, the technical solution adopted by the present invention is as follows:

[0006] A multi-objective low-carbon scheduling optimization method for open-pit mine trucks based on a two-stage evolutionary selection mechanism includes the following steps:

[0007] Step 1: Based on current low-carbon requirements and mine production needs and indicators, construct a multi-objective truck low-carbon scheduling optimization model for open-pit mines with the objectives of minimizing total truck waiting time, minimizing total ore transportation costs, minimizing ore grade deviation, and minimizing total transportation carbon emissions.

[0008] Step 2: Improve the original NSGA-II algorithm to obtain the NSGA-II-TS algorithm based on a two-stage evolutionary selection mechanism;

[0009] Step 3: Solve the multi-objective low-carbon scheduling model of open-pit mine obtained in Step 1 using the NSGA-II-TS algorithm obtained in Step 2, and obtain an optimization result of multi-objective low-carbon scheduling of open-pit mine trucks based on a two-stage evolutionary selection mechanism.

[0010] The specific structure of the open-pit mine multi-objective low-carbon scheduling model in step 1 is as follows:

[0011] 1) Optimization Objective

[0012] F = Minimize(F1, F2, F3, F4)

[0013] in:

[0014]

[0015]

[0016]

[0017]

[0018] min F1 represents the objective of minimizing total truck waiting time, min F2 represents the objective of minimizing total ore transportation cost, min F3 represents the objective of minimizing ore grade deviation, and min F4 represents the objective of minimizing total transportation carbon emissions. r represents the truck number, r = 1, 2, ..., R, C. r Let x represent the loading capacity of the r-th truck. rij y represents the number of times the r-th truck travels from the i-th loading point to the j-th unloading point. rji d represents the number of times the r-th truck returns from the j-th unloading point to the i-th loading point. ijLet Cr1 represent the distance from the i-th loading point to the j-th unloading point, Cr2 represent the unit cost of the r-th truck under heavy load, Cr3 represent the unit distance cost of the r-th truck under empty load, and T represent the maintenance cost of the r-th truck per unit time. svr Indicates the necessary refueling and charging time, Tz rji Let Tq represent the travel time of the r-th truck from the i-th loading point to the j-th unloading point. rij T represents the empty travel time of the r-th truck from the j-th unloading point to the i-th loading point. L The following represents the working time of each truck shift: ρ1 represents the energy consumption per unit mile when the vehicle is heavily loaded, ρ2 represents the energy consumption per unit mile when the vehicle is unloaded, FCr is the fuel consumption of the r-th diesel vehicle, Fr is the electrical energy consumption of the r-th new energy electric vehicle, μ1 is the fuel emission factor, γ is the proportion of thermal power generation in the total electricity generation, μ2 is the carbon emission coefficient of electricity generation, ω is the unit carbon emission cost, and T C For truck maintenance and inspection time, e i Let represent the ore grade at the i-th loading point, and e represent the ore grade limit;

[0019] 2) Production capacity constraints at unloading points:

[0020]

[0021] Among them, C r Let q represent the loading capacity of the r-th truck. j x represents the maximum production capacity of the j-th unloading point. rij For 0-1 variables, when x rij When = 1, it means that the r-th mining truck has traveled from the i-th loading point to the j-th unloading point, i∈{1, 2, ..., I}, j∈{1, 2, ..., J}, r∈{1, 2, ..., R};

[0022] 3) Production plan requirements at the unloading point:

[0023]

[0024] Among them, C r f represents the loading capacity of the r-th truck. j x represents the maximum production capacity of the j-th unloading point. rij For 0-1 variables, when x rij When = 1, it means that the r-th mining truck has traveled from the i-th loading point to the j-th unloading point, i∈{1, 2, ..., M}, j∈{1, 2, ..., N}, r∈{1, 2, ..., R};

[0025] 4) Production capacity constraints at loading points:

[0026]

[0027] Among them, C r G represents the loading capacity of the r-th truck. i Let x be the maximum production capacity of the i-th unloading point. rij For 0-1 variables, when x rij When = 1, it means that the r-th mining truck has traveled from the i-th loading point to the j-th unloading point, i∈{1, 2, ..., I}, j∈{1, 2, ..., J}, r∈{1, 2, ..., R};

[0028] 5) Ore grade requirements at the unloading point:

[0029]

[0030] Among them, C r Let x represent the loading capacity of the r-th truck. rij y represents the number of times the r-th truck travels from the i-th loading point to the j-th unloading point. rji E represents the number of times the r-th truck returns from the j-th unloading point to the i-th loading point. i Let represent the ore grade at the i-th loading point, e represent the ore grade limit, and ε represent the maximum permissible error in ore grade.

[0031] 6) Traffic flow continuity constraints:

[0032]

[0033]

[0034] Where, x rij y represents the number of times the r-th truck travels from the i-th loading point to the j-th unloading point. rji This represents the number of times the r-th truck returns from the j-th unloading point to the i-th loading point;

[0035] 7) Vehicle transport frequency requirements between loading and unloading points:

[0036] x rij y rij ∈{0, 1, 2, 3, ...}

[0037] Where, x rij y represents the number of times the r-th truck travels from the i-th loading point to the j-th unloading point. rji This represents the number of times the r-th truck returns from the i-th unloading point to the i-th loading point.

[0038] Step 2 improves the original NSGA-II algorithm to obtain the NSGA-II-TS algorithm based on a two-stage evolutionary selection mechanism, specifically as follows:

[0039] 1) Introduce a neighborhood density mechanism based on angle measurement into the original NSGA-II algorithm, specifically:

[0040] First, the angle between any two individuals in the population is measured. The formula for calculating the angle is shown in (1). The target value is normalized using formula (2), and then the angle is normalized as shown in formula (3). When calculating the neighborhood density of an individual in the population, samples can be taken from the surrounding area, and then the neighborhood density of that individual can be calculated according to formula (4).

[0041]

[0042]

[0043]

[0044]

[0045] Where, angle(x) i x j ) represents individual x i and x j The angle between them, ||·|| represents the 2-norm of the vector, specifically the vector angle(x) between any two individuals. i x j )∈[0,π / 2],F′(x) j ) represents individual x j Normalized target value, ideal point is The lowest point is The normalized value of the angle is represented as angle′. d represents the sum of the angles of the h sampling points that have the smallest x-angle distance from the individual. h (x) is the neighborhood density; the individual represents a solution for multi-objective low-carbon scheduling and optimized transportation of open-pit mine trucks.

[0046] 2) Introduce a two-stage evolutionary selection process into the original NSGA-II algorithm, specifically:

[0047] The first stage of the selection process: individuals are entered into the archive set from low to high according to the non-dominated ranking. When selecting the critical layer, individuals with lower neighborhood density are selected into the archive set. In the second stage of the selection process: when the number of individuals in the first non-dominated layer exceeds the population size N, an elite selection mechanism based on aggregation function and neighborhood density is used to select individuals. By calculating the convergence index and diversity index, the two individuals x and y with the largest convergence index are selected. Then, the individual with the larger neighborhood density is deleted. This deletion strategy is used until the population size of the non-dominated layer is N, and these N individuals are selected into the archive set.

[0048] In step 3, the NSGA-II algorithm based on the two-stage evolutionary selection mechanism is used to solve the multi-objective low-carbon scheduling model of the open-pit mine. The specific solution process is as follows:

[0049] Step 3.1: Set the basic parameters of the algorithm: population size N, maximum number of iterations G. max The initial population is generated using a random initialization method;

[0050] Step 3.2: Perform genetic operations on the parent population, simulating binary crossover and polynomial mutation, to form a new offspring population;

[0051] Step 3.3: Merge the parent population and the offspring population into a single population;

[0052] Step 3.4: Calculate the target value, convergence index, and neighborhood density of all individuals in the merged population after step 3.3.

[0053] Step 3.5: Based on the target values ​​of all individuals in the merged population obtained in Step 3.4, perform non-dominated sorting according to the Pareto dominance relationship.

[0054] Step 3.6: Apply a two-stage evolutionary selection mechanism to the same population obtained in Step 3.3, output the archive set, and proceed to the next iteration;

[0055] Step 3.7: The system checks the algorithm's iteration state to determine if the termination condition has been reached. If yes, it outputs the optimal solution and the target value; otherwise, it returns to step 3.2.

[0056] The term "individual" refers to one solution for multi-objective, low-carbon scheduling and optimized transportation of open-pit mine trucks; "population" refers to multiple solutions for multi-objective, low-carbon scheduling and optimized transportation of open-pit mine trucks.

[0057] The Pareto dominance relationship in step 3.5 is specifically as follows:

[0058] For a minimization problem, given two solutions x and y, the Pareto dominance of solution x over solution y is denoted as x□y if and satisfy:

[0059]

[0060]

[0061] If, for two given solutions x and y, solution x does not satisfy the Pareto dominance condition for solution y, and solution y does not satisfy the Pareto dominance condition for solution x, then solutions x and y are mutually non-dominant and are denoted as x□y, y□x.

[0062] The non-dominated sorting in step 3.5 is performed as follows:

[0063] First, the non-dominated order of the new population is performed, dividing the entire population into different non-dominated layers (Z1, Z2, ..., Z...). n The critical layer (satisfying the condition |Z1|+|Z2|+...+|Z) will be determined. i |<n,|Z1|+|Z2|+...+|Z i+1 |>n,Z i+1 Individuals before the critical layer are added to the offspring population. Elite selection is performed on individuals in the critical layer. The process stops when the number of individuals in the offspring population reaches N.

[0064] The process of implementing the two-stage evolutionary selection mechanism in step 3.6 is as follows:

[0065] Step 3.6.1: The system obtains the sorting results of the non-dominated layer;

[0066] Step 3.6.2: The system determines whether the size of the first-level population is greater than the fixed size N. If so, during the first-stage selection process, individuals are selected from low to high according to their non-dominated ranking and then enter the archive set. When selecting the critical layer, individuals with better distribution and lower neighborhood density are selected to enter the archive set. If not, in the second stage, an elite selection mechanism based on aggregation function and neighborhood density is used to select individuals. By calculating the convergence index and diversity index, the two individuals x and y with the largest convergence index are selected, and then the individual with the larger neighborhood density is deleted. This deletion strategy is used until the population size of the non-dominated layer is N, and then these N individuals are selected to enter the archive set.

[0067] The specific calculation method for the convergence index based on the target value is as follows:

[0068]

[0069] Among them, f i (x) represents the value of the i-th objective function, and C(x) represents a simple aggregate function;

[0070] Step 3.6.3: The system processes the archive set obtained in step 3.6.2, outputs the archive set, and proceeds to the next iteration.

[0071] The specific requirements for the termination condition in step 3.7 are as follows:

[0072] When solving an engineering problem involving a multi-objective, low-carbon scheduling optimization method for open-pit mine trucks based on a two-stage evolutionary selection mechanism, the stopping condition of the algorithm is set as the maximum number of iterations G. max The algorithm stops when it reaches the maximum number of iterations.

[0073] A multi-objective low-carbon scheduling optimization system for open-pit mine trucks based on a two-stage evolutionary selection mechanism includes:

[0074] Low-carbon scheduling model module: Based on the mine's production needs and indicators, a multi-objective low-carbon truck scheduling optimization model for open-pit mines is constructed with the objectives of minimizing total truck waiting time, minimizing total ore transportation costs, minimizing ore grade deviation, and minimizing total transportation carbon emissions.

[0075] The algorithm module is designed for a multi-objective low-carbon scheduling optimization method for open-pit mine trucks based on a two-stage evolutionary selection mechanism. The NSGA-II-TS algorithm based on the two-stage evolutionary selection mechanism is constructed.

[0076] Decision module: Based on the constructed NSGA-II-TS algorithm, the multi-objective low-carbon scheduling model of open-pit mine obtained by the low-carbon scheduling model module is solved, and an optimization result of multi-objective low-carbon scheduling of open-pit mine trucks based on a two-stage evolutionary selection mechanism is obtained.

[0077] A multi-objective low-carbon scheduling optimization device for open-pit mine trucks based on a two-stage evolutionary selection mechanism includes:

[0078] Memory: Used to store a computer program that implements a multi-objective low-carbon scheduling optimization method for open-pit mine trucks based on a two-stage evolutionary selection mechanism;

[0079] A processor is used to implement the above-mentioned multi-objective low-carbon scheduling optimization method for open-pit mine trucks based on a two-stage evolutionary selection mechanism when executing the computer program.

[0080] A computer-readable storage medium comprising:

[0081] The computer-readable storage medium stores a computer program that, when executed by a processor, enables a multi-objective low-carbon scheduling optimization method for open-pit mine trucks based on a two-stage evolutionary selection mechanism.

[0082] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0083] 1. Compared with the existing technology, the present invention introduces multiple objectives such as total truck waiting time, total ore transportation cost, ore grade deviation and total transportation carbon emissions on the basis of the existing single-objective scheduling optimization model. At the same time, it takes into account the actual traffic capacity of the mining area roads and constructs a multi-objective scheduling optimization model for open-pit mines. Therefore, it has the characteristics of low transportation cost, high production efficiency and low carbon emissions.

[0084] 2. Existing optimization scheduling mainly uses linear programming, integer programming and dynamic programming to optimize scheduling problems. However, the scheduling of open-pit mines is complex and has diverse requirements, so it often cannot obtain practical and effective results. The improved NSGA-II-TS algorithm proposed in this invention is simple in principle and easy to implement. In the solution process, it can effectively guarantee the convergence and diversity of solutions and can quickly solve for the solution set that meets the requirements.

[0085] 3. This invention considers the carbon emissions during the dispatching and transportation of open-pit mine trucks based on existing scheduling systems, and proposes the NSGA-II-TS algorithm for high-dimensional scheduling problems. Therefore, it can establish a complete transportation route and transportation vehicle model according to the actual production needs of the mining area, and can find feasible solutions. This is of great significance for reducing carbon emissions during transportation, improving transportation efficiency, reducing transportation costs, and improving the economic benefits of mining enterprises.

[0086] 4. This invention employs an improved NSGA-II-TS algorithm based on an angle measurement-based neighborhood density mechanism and a two-stage evolutionary selection mechanism to solve the multi-objective low-carbon scheduling optimization model for open-pit mine trucks. This approach yields a scheduling and transportation scheme that better suits the current state of mine production. Based on current Chinese policies and the actual situation of mine production, this invention is of great significance for improving transportation efficiency, reducing transportation costs, decreasing carbon emissions, and enhancing the economic benefits of mining enterprises.

[0087] 5. In the original NSGA-II algorithm, when solving two-dimensional and three-dimensional optimization problems, due to the Pareto dominance relation, the proportion of non-dominated solutions increases sharply in the solution process for high-dimensional multi-objective optimization problems of three dimensions and above. At this time, the non-dominated sorting loses its selection pressure and cannot retain solutions with good convergence from the critical layer selection process, and the diversity of solutions cannot be guaranteed. This invention proposes a two-stage NSGA-II-TS algorithm selection mechanism to balance the convergence and diversity of solutions. Based on the two-stage evolutionary selection mechanism strategy of aggregation function and neighborhood density, it not only considers the convergence of solutions, but also enhances the diversity of solutions.

[0088] 6. This invention employs the NSGA-II-TS algorithm with a two-stage evolutionary selection mechanism to solve the model, making the solution of multi-objective models more scientific, reasonable, and feasible.

[0089] 7. The NSGA-II-TS algorithm in this invention introduces a neighborhood density mechanism based on angle measurement to improve the diversity maintenance performance of the algorithm. It can ensure that solutions with better diversity in a certain region participate in the further evolution process, and is committed to finding a set of more representative solutions with better distribution. This algorithm can effectively solve high-dimensional multi-objective problems of three dimensions or above.

[0090] In summary, the purpose of this invention is to provide a multi-objective low-carbon scheduling optimization method for open-pit mine trucks based on a two-stage evolutionary selection mechanism. Based on the actual production scheduling of open-pit mines, a multi-objective low-carbon scheduling model for open-pit mines is constructed with the objectives of minimizing total truck waiting time, total ore transportation cost, ore grade deviation, and total transportation carbon emissions. An NSGA-II-TS algorithm based on a two-stage evolutionary selection mechanism is then constructed, and finally, the NSGA-II-TS algorithm is used to solve the multi-objective low-carbon scheduling model for open-pit mines. This multi-objective scheduling optimization method can effectively reduce the transportation distance and number of trips during truck scheduling, thereby significantly reducing the transportation costs of enterprises. Furthermore, it can improve the production capacity of mines while effectively ensuring ore grade, significantly enhancing the economic benefits of mining enterprises. Moreover, path planning and traffic flow allocation are implemented during the model construction process, which can meet the actual needs of current open-pit mine transportation scheduling. Attached Figure Description

[0091] Figure 1 This is a flowchart of the present invention.

[0092] Figure 2 This is a flowchart of the NSGA-II algorithm for solving the model based on a two-stage evolutionary selection mechanism in this invention.

[0093] Figure 3 This is a schematic diagram of the neighborhood density mechanism based on angle measurement in this invention (the diagram shows sampling two neighbors). Detailed Implementation

[0094] The working principle of the present invention will be further described below with reference to the accompanying drawings.

[0095] See Figure 1 A multi-objective low-carbon scheduling optimization method for open-pit mine trucks based on a two-stage evolutionary selection mechanism includes the following steps:

[0096] Step 1: Based on current low-carbon requirements and mine production needs and indicators, a multi-objective truck low-carbon scheduling optimization model for open-pit mines is constructed with the objectives of minimizing total truck waiting time, total ore transportation cost, ore grade deviation, and total transportation carbon emissions. At the same time, the actual traffic capacity of the mining area roads is considered to construct a multi-objective scheduling optimization model for open-pit mines, thus featuring low transportation costs, high production efficiency, and low carbon emissions.

[0097] The specific structure of the open-pit mine multi-objective low-carbon scheduling model is as follows:

[0098] 1) Optimization Objective

[0099] F=Minimize(F1, F2, F3, F4)

[0100] in:

[0101]

[0102]

[0103]

[0104]

[0105] min F1 represents the objective of minimizing total truck waiting time, min F2 represents the objective of minimizing total ore transportation cost, min F3 represents the objective of minimizing ore grade deviation, and min F4 represents the objective of minimizing total transportation carbon emissions. r represents the truck number, r = 1, 2, ..., R, C r Let x represent the loading capacity of the r-th truck. rij y represents the number of times the r-th truck travels from the i-th loading point to the j-th unloading point. rji d represents the number of times the r-th truck returns from unloading point j to loading point i. ij Cr1 represents the distance from loading point i to unloading point j, Cr2 represents the unit cost of the r-th truck under heavy load, Cr3 represents the unit distance cost of the r-th truck under empty load, and T represents the maintenance cost of the r-th truck per unit time. svr Indicates the necessary refueling and charging time, Tz rji T represents the travel time of the r-th truck from the i-th loading point to the j-th unloading point. qrij T represents the empty travel time of the r-th truck from the j-th unloading point to the i-th loading point. LThe following represents the working time of each truck shift: ρ1 represents the energy consumption per unit mile when the vehicle is heavily loaded, ρ2 represents the energy consumption per unit mile when the vehicle is unloaded, FCr is the fuel consumption of the r-th diesel vehicle, Fr is the electrical energy consumption of the r-th new energy electric vehicle, μ1 is the fuel emission factor, γ is the proportion of thermal power generation in the total electricity generation, μ2 is the carbon emission coefficient of electricity generation, ω is the unit carbon emission cost, and T C For truck maintenance and inspection time, e i represents the ore grade at loading point i, and e represents the ore grade limit;

[0106] 2) Production capacity constraints at unloading points: the total amount of ore transported to each unloading point must not exceed the maximum production capacity of that unloading point.

[0107]

[0108] Among them, C r Let q represent the loading capacity of the r-th truck. j x represents the maximum production capacity of unloading point j. rij For 0-1 variables, when x rij When = 1, it means that the r-th mining truck has traveled from the i-th loading point to the j-th unloading point, i∈{1, 2, ..., I}, j∈{1, 2, ..., J}, r∈{1, 2, ..., R};

[0109] 3) Production plan requirements for unloading points, i.e., the total amount of ore unloaded at each unloading point must at least meet the production needs of that unloading point:

[0110]

[0111] Among them, C r f represents the loading capacity of the r-th truck. j x represents the maximum production capacity of unloading point j. rij For 0-1 variables, when x rij When = 1, it means that the r-th mining truck has traveled from the i-th loading point to the j-th unloading point, i∈{1, 2, ..., M}, j∈{1, 2, ..., N}, r∈{1, 2, ..., R};

[0112] 4) Production capacity constraints at loading points: the total amount of ore loaded at each loading point must not exceed the maximum production capacity of that loading point.

[0113]

[0114] Among them, C r G represents the loading capacity of the r-th truck. i x represents the maximum production capacity of unloading point i. rij For 0-1 variables, when x rijWhen = 1, it means that the r-th mining truck has traveled from the i-th loading point to the j-th unloading point, i∈{1, 2, ..., I}, j∈{1, 2, ..., J}, r∈{1, 2, ..., R};

[0115] 5) Ore grade requirements at unloading points: the ore grade deviation at each unloading point must not exceed the maximum grade deviation.

[0116]

[0117] Among them, C r Let x represent the loading capacity of the r-th truck. rij y represents the number of times the r-th truck travels from the i-th loading point to the j-th unloading point. rji e represents the number of times the r-th truck returns from unloading point j to loading point i. i ε represents the ore grade at loading point i, e represents the ore grade limit, and ε represents the maximum permissible error in ore grade.

[0118] 6) Traffic flow continuity constraint, which ensures that the inbound and outbound traffic flow is equal at each loading / unloading point:

[0119]

[0120]

[0121] Where, x rij y represents the number of times the r-th truck travels from the i-th loading point to the j-th unloading point. rji This represents the number of times the r-th truck returns from the unloading point to the loading point i.

[0122] 7) Vehicle transport frequency requirements between loading and unloading points:

[0123] x rij y rij ∈{0, 1, 2, 3, ...}

[0124] Where, x rij y represents the number of times the r-th truck travels from the i-th loading point to the j-th unloading point. rji This represents the number of times the r-th truck returns from the unloading point to the loading point i.

[0125] Table 1 Explanation of Basic Symbols

[0126]

[0127]

[0128] Step 2: For a multi-objective low-carbon scheduling optimization method for open-pit mine trucks based on a two-stage evolutionary selection mechanism, the NSGA-II-TS algorithm based on the two-stage evolutionary selection mechanism is constructed; the specific process of the NSGA-II-TS algorithm based on the two-stage evolutionary selection mechanism is as follows:

[0129] This problem contains four objective functions and falls under the category of high-dimensional multi-objective optimization problems. Balancing convergence and diversity is extremely difficult in high-dimensional multi-objective optimization problems, and the density of individuals is very difficult to measure in high-dimensional space. For practical decision-makers in engineering, a representative set of solutions is a guarantee of good decision-making. Based on this, an NSGA-II-TS algorithm with a two-stage evolutionary selection mechanism is designed to solve this engineering problem.

[0130] The original NSGA-II algorithm introduces a neighborhood density mechanism based on angle measurement and an elite individual selection mechanism, while retaining the random initial population generation and the fast Pareto dominance relationship non-dominated sorting mechanism. The algorithm first generates a randomly initialized population, which then undergoes an evolutionary process. First, mating selection, binary crossover (SBX), and polynomial mutation (PM) are used to generate offspring populations. Next, populations are merged for environmental selection. Non-dominated ranks are formed through the fast Pareto dominance relationship non-dominated sorting, and then the dominant tier is selected to join the archive set. When the archive set population size is less than N, the neighborhood density mechanism is used to select individuals with good diversity from outside the archive set to enter the archive set, maintaining the population size at N. When the archive set population size is greater than N, the elite individual selection mechanism is used to select elite individuals from the non-dominated tier set to enter the archive set, maintaining the population size at N.

[0131] 1) NSGA-II adopts the traditional crowding distance mechanism, which is based on Euclidean distance calculation. In high-dimensional space, this method has poor accuracy in density measurement and is prone to generating dominance resistance solutions, resulting in poor quality of the final solution. In order to provide decision-makers with a set of representative solutions with good diversity, this paper improves the solution diversity maintenance mechanism and proposes a neighborhood density mechanism based on angle measurement. First, the angle between any two individuals in the population is measured. The angle measurement calculation formula is as shown in (1). Then, the angle is normalized as shown in formula (3). When calculating the neighborhood density of an individual in the population, the surrounding area can be sampled, and then the neighborhood density of the individual can be calculated according to formula (4):

[0132]

[0133]

[0134]

[0135]

[0136] Where, angle(x) i x j ) represents individual x i and x j The angle between them, ||·|| represents the 2-norm of the vector, specifically the vector angle(x) between any two individuals. i x j )∈[0,π / 2]. F′(x) j ) represents individual x j Normalized target value, ideal point is The lowest point is The target normalization value is f i ′(x j The normalized value of the angle is represented as angle′. d represents the sum of the angles of the h sampling points that have the smallest x-angle distance from the individual. h (x) is the neighborhood density; the individual represents a solution for multi-objective low-carbon scheduling and optimized transportation of open-pit mine trucks.

[0137] In the original NSGA-II algorithm, when solving two-dimensional and three-dimensional optimization problems, the Pareto dominance relation effectively selects solutions with good convergence based on the non-dominated layer. However, for high-dimensional multi-objective optimization problems (three dimensions and above), the proportion of non-dominated solutions increases sharply during the solution process. At this point, the non-dominated sorting loses its selection pressure and cannot retain solutions with good convergence from the critical layer selection process, nor can the diversity of solutions be guaranteed. To better solve this engineering problem, a two-stage selection mechanism is proposed to balance the convergence and diversity of solutions. Based on the aggregation function and neighborhood density, the two-stage evolutionary selection mechanism strategy not only considers the convergence of solutions but also enhances the diversity of solutions.

[0138] Specifically, the convergence of the population is one of the important factors to consider in the design of multi-objective evolutionary algorithms. The convergence index of each solution can be calculated by summing the objective values. The specific calculation method is as follows:

[0139]

[0140] Among them, f i (x) represents the value of the i-th objective function. C(x) is a simple aggregation function. The smaller the value of C(x), the better the convergence of the solution.

[0141] 2) In the first stage of selection: individuals are ranked from lowest to highest according to their non-dominated order and enter the archive set. When selecting from the critical layer, individuals with lower neighborhood density are selected for the archive set. In the second stage of selection, when the number of individuals in the first non-dominated layer exceeds the population size N, the Pareto dominance relationship loses its evolutionary selection pressure. An elite selection mechanism based on aggregation functions and neighborhood density is then used to select individuals. h The smaller the value of (x), the smaller the neighborhood density of the individual. The smaller the convergence index C(x), the better the convergence of the individual. By calculating the convergence index and the diversity index, the two individuals x and y with the largest convergence index are selected, and then the individual with the larger neighborhood density is deleted. This deletion strategy is used one by one until the population size of the non-dominated layer is N, and these N individuals are selected to enter the archive set.

[0142] See Figure 2 Step 3: Solve the multi-objective scheduling model using the NSGA-II-TS algorithm based on the two-stage evolutionary selection mechanism. The specific solution process is as follows:

[0143] Step 3.1: Set the basic parameters of the algorithm: population size N, maximum number of iterations G. max The initial population is generated using a random initialization method;

[0144] Step 3.2: Perform genetic operations on the parent population, simulating binary crossover and polynomial mutation, to form a new offspring population;

[0145] Step 3.3: Merge the parent population and the offspring population into a single population;

[0146] Step 3.4: Calculate the target value, convergence index, and neighborhood density of all individuals in the merged population after step 3.3.

[0147] Step 3.5: Based on the target values ​​of all individuals in the merged population obtained in Step 3.4, perform non-dominated sorting according to the Pareto dominance relationship.

[0148] Step 3.6: Apply a two-stage evolutionary selection mechanism to the same population obtained in Step 3.3, output the archive set, and proceed to the next iteration;

[0149] Step 3.7: The system checks whether the algorithm iteration state has reached the termination condition. If yes, it outputs the optimal solution and the target value; otherwise, it returns to step 3.2.

[0150] The Pareto dominance relationship in step 3.5 is specifically as follows:

[0151] For a minimization problem, given two solutions x and y, the Pareto dominance of solution x over solution y is denoted as x□y if and satisfies:

[0152]

[0153]

[0154] If, for two given solutions x and y, solution x does not satisfy the Pareto dominance condition for solution y, and solution y does not satisfy the Pareto dominance condition for solution x, then solutions x and y are mutually non-dominant and are denoted as...

[0155] The non-dominated sorting in step 3.5 is performed as follows:

[0156] First, the non-dominated order of the new population is performed, dividing the entire population into different non-dominated layers (Z1, Z2, ..., Z...). n The critical layer (satisfying the condition |Z1|+|Z2|+...+|Z) will be determined. i |<n,|Z1|+|Z2|+...+|Z i+1 |>n,Z i+1 Individuals before the critical layer are added to the offspring population. Elite selection is performed on individuals in the critical layer. The process stops when the number of individuals in the offspring population reaches N.

[0157] The process of implementing the two-stage evolutionary selection mechanism in step 3.6 is as follows:

[0158] Step 3.6.1: The system obtains the sorting results of the non-dominated layer;

[0159] Step 3.6.2: The system assesses the first-level population size to determine if it exceeds a fixed size N. If so, individuals are selected for the archive set in the first stage based on their non-dominated hierarchy from low to high. Individuals with better distribution at the critical level are selected based on neighborhood density. If not, in the second stage, an aggregation function and neighborhood density index are used to select elite individuals for the archive set. During the first stage selection, individuals are selected for the archive set from low to high based on their non-dominated ranking. When selecting at the critical level, individuals with lower neighborhood density are selected. During the second stage selection, when the number of individuals in the first non-dominated hierarchy exceeds the population size N, the Pareto dominance relationship loses its evolutionary selection pressure. An elite selection mechanism based on the aggregation function and neighborhood density is then used to select individuals. hThe smaller the value of (x), the smaller the neighborhood density of the individual. The smaller the convergence index C(x), the better the convergence of the individual. By calculating the convergence index and the diversity index, the two individuals x and y with the largest convergence index are selected. Then, the individual with the larger neighborhood density is deleted. This deletion strategy is used one by one until the population size of the non-dominated layer is N. These N individuals are then selected to enter the archive set.

[0160] The specific calculation method for the convergence index based on the target value is as follows:

[0161]

[0162] Among them, f i C(x) represents the value of the i-th objective function, and C(x) represents a simple aggregate function.

[0163] Step 3.6.3: The system processes the archive set obtained in step 3.6.2, outputs the archive set, and proceeds to the next iteration;

[0164] The individual represents one solution for multi-objective low-carbon scheduling and optimization of open-pit mine trucks; the population represents multiple solutions for multi-objective low-carbon scheduling and optimization of open-pit mine trucks. In the process of solving the multi-objective low-carbon scheduling model of open-pit mines using the NSGA-II-TS algorithm, the population simultaneously outputs multiple mine scheduling solutions, which provide mine production decision-makers with a variety of choices.

[0165] The specific requirements for the termination condition in step 3.7 are as follows:

[0166] When solving an engineering problem involving a multi-objective, low-carbon scheduling optimization method for open-pit mine trucks based on a two-stage evolutionary selection mechanism, the stopping condition of the algorithm is set as the maximum number of iterations G. max The algorithm stops when it reaches the maximum number of iterations.

[0167] See Figure 2 The schematic diagram of the neighborhood density mechanism based on angle measurement in this invention (showing sampling of two neighbors) illustrates the situation under two-dimensional objectives. The horizontal axis represents the target value of the first objective, the vertical axis represents the target value of the second objective, the green curve represents the true Pareto dominance front of the problem, the star represents the origin of angle measurement, the red solid dot represents the individual currently being calculated, and the black solid dot represents the solution in the current population. It can be seen that the neighborhood density mechanism based on angle measurement can ensure that solutions with good diversity within a certain area participate in the further evolutionary process.

[0168] A multi-objective low-carbon scheduling optimization system for open-pit mine trucks based on a two-stage evolutionary selection mechanism includes:

[0169] Low-carbon scheduling model module: Based on the mine's production needs and indicators, and with the objectives of minimizing total truck waiting time, total ore transportation cost, ore grade deviation, and total transportation carbon emissions, a multi-objective low-carbon scheduling optimization model for open-pit mine trucks is constructed; this model is used to implement step 1 of a multi-objective low-carbon scheduling optimization method for open-pit mine trucks based on a two-stage evolutionary selection mechanism.

[0170] The algorithm module is designed to implement step 2 in a multi-objective low-carbon scheduling optimization method for open-pit mine trucks based on a two-stage evolutionary selection mechanism.

[0171] Decision module: Solve the multi-objective low-carbon scheduling model of open-pit mine obtained by the low-carbon scheduling model module according to the constructed NSGA-II-TS algorithm, and obtain a multi-objective low-carbon scheduling optimization result of open-pit mine trucks based on a two-stage evolutionary selection mechanism; used to implement step 3 in the multi-objective low-carbon scheduling optimization method of open-pit mine trucks based on a two-stage evolutionary selection mechanism.

[0172] A multi-objective low-carbon scheduling optimization device for open-pit mine trucks based on a two-stage evolutionary selection mechanism includes:

[0173] Memory: Used to store a computer program that implements a multi-objective low-carbon scheduling optimization method for open-pit mine trucks based on a two-stage evolutionary selection mechanism;

[0174] A processor is used to implement the above-mentioned multi-objective low-carbon scheduling optimization method for open-pit mine trucks based on a two-stage evolutionary selection mechanism when executing the computer program.

[0175] A computer-readable storage medium comprising:

[0176] The computer-readable storage medium stores a computer program that, when executed by a processor, enables a multi-objective low-carbon scheduling optimization method for open-pit mine trucks based on a two-stage evolutionary selection mechanism.

Claims

1. A multi-objective low-carbon scheduling optimization method for open-pit mine trucks based on a two-stage evolutionary selection mechanism, characterized in that: The following steps are involved: Step 1: Based on the current low-carbon requirements and mine production needs and indicators, a multi-objective truck low-carbon scheduling optimization model for open-pit mines is constructed with the goals of minimizing the total truck waiting time, minimizing the total ore transportation cost, minimizing the ore grade deviation, and minimizing the total transportation carbon emissions; Step 2, improve the original NSGA-II algorithm to obtain the NSGA-II-TS algorithm based on the two-stage evolutionary selection mechanism; Step 3, solving the multi-objective low-carbon scheduling model of the open-pit mine obtained in step 1 according to the NSGA-II-TS algorithm obtained in step 2, and obtaining a multi-objective low-carbon scheduling optimization result of open-pit mine trucks based on a two-stage evolutionary selection mechanism.

2. The multi-objective low-carbon scheduling optimization method for open-pit mine trucks based on a two-stage evolutionary selection mechanism according to claim 1 is characterized in that: The specific structure of the open-pit mine multi-objective low-carbon scheduling model in step 1 is as follows: 1) Optimization goals F=Minimize(F1,F2,F3,F4) in: min F1 represents the minimum total waiting time target for trucks, min F2 represents the minimum total transportation cost target for ore, min F3 represents the minimum ore grade deviation target, min F4 represents the minimum total transportation carbon emission target, r represents the truck number, r=1,2,...,R,C r represents the loading capacity of the rth truck, x rij represents the number of times the rth truck goes from the ith loading point to the jth unloading point, y rji represents the number of times the r-th truck returns from the j-th unloading point to the i-th loading point, d ij represents the distance from the i-th loading point to the j-th unloading point, Cr1 represents the unit cost of the r-th truck under heavy-load operation, Cr2 represents the unit distance cost of the r-th truck when it is empty, Cr3 represents the maintenance cost of the r-th truck per unit time, T svr Indicates the necessary refueling and charging time, Tz rji Tq represents the heavy vehicle running time of the rth truck from the ith loading point to the jth unloading point, rij represents the empty truck running time of the rth truck from the jth unloading point to the ith loading point, T L represents the working time of the truck in each shift, ρ1 represents the energy consumption per unit mileage when the vehicle is heavily loaded, ρ2 represents the energy consumption per unit mileage when the vehicle is unloaded, FCr is the fuel consumption of the rth diesel vehicle, Fr is the electricity consumption of the rth new energy electric vehicle, μ1 is the fuel emission factor, γ is the proportion of thermal power generation in the total power generation, μ2 is the carbon emission coefficient generated by electricity, ω is the unit carbon emission cost, T C Maintenance inspection time for trucks, e i represents the ore grade of the i-th loading point, and e represents the grade limit of the ore; 2) Production capacity constraints at unloading points: Among them, C r represents the loading capacity of the rth truck, q j is the maximum production capacity of the jth unloading point, x rij is a 0-1 variable. When x rij =1, it means that the r-th mining truck has traveled from the i-th loading point to the j-th unloading point, i∈{1,2,...,I}, j∈{1,2,…,J}, r∈{1,2,…,R}; 3) Production planning requirements for unloading points: Among them, C r represents the loading capacity of the rth truck, f j is the maximum production capacity of the jth unloading point, x rij is a 0-1 variable. When x rij =1, it means that the rth mining truck has traveled from the ith loading point to the jth unloading point, i∈{1,2,...,M}, j∈{1,2,...,N}, r∈{1,2,...,R}; 4) Production capacity constraints of loading points: Among them, C r represents the loading capacity of the rth truck, g i is the maximum production capacity of the i-th unloading point, x rij is a 0-1 variable. When x rij =1, it means that the r-th mining truck has traveled from the i-th loading point to the j-th unloading point, i∈{1,2,...,I}, j∈{1,2,…,J}, r∈{1,2,…,R}; 5) Ore grade requirements at unloading point: Among them, C r represents the loading capacity of the rth truck, x rij represents the number of times the rth truck goes from the ith loading point to the jth unloading point, y rji represents the number of times the r-th truck returns from the j-th unloading point to the i-th loading point, e i represents the ore grade of the i-th loading point, e represents the grade limit of the ore, and ε represents the maximum allowable error of the ore grade; 6) Traffic flow continuity constraints: Among them, x rij represents the number of times the rth truck goes from the ith loading point to the jth unloading point, y rji represents the number of times the r-th truck returns from the j-th unloading point to the i-th loading point; 7) Vehicle transportation frequency requirements between loading and unloading points: x rij ,and rij ∈{0,1,2,3,…} Among them, x rij represents the number of times the rth truck goes from the ith loading point to the jth unloading point, y rji It represents the number of times the r-th truck returns from the j-th unloading point to the i-th loading point.

3. The multi-objective low-carbon scheduling optimization method for open-pit mine trucks based on a two-stage evolutionary selection mechanism according to claim 1 is characterized in that: The step 2 improves the original NSGA-II algorithm to obtain the NSGA-II-TS algorithm based on the two-stage evolutionary selection mechanism, specifically: 1) Introduce the neighborhood density mechanism based on angle measurement into the original NSGA-II algorithm, specifically: First, measure the angle between any two individuals in the population. The angle measurement calculation formula is as follows (1). Use formula (2) to normalize the target value, and then normalize the angle, see formula (3). When calculating the neighborhood density of an individual in the population, sample the surroundings, and then calculate the neighborhood density of the individual according to formula (4): Among them, angle(x i , x j ) represents individual x i and x j The angle value between them, ‖·‖ represents the 2-norm of the calculated vector. In particular, the vector angle angle(x i , x j )∈[0,π / 2],F′(x j ) represents individual x j Normalized target value, the ideal point is The lowest point is The normalized angle value is expressed as angle′. represents the sum of the angles of the h sampling points with the smallest angle from the individual x, d h (x) is the neighborhood density; the individual represents a solution for multi-objective low-carbon scheduling optimization of open-pit mine truck transportation; 2) A two-stage evolutionary selection mechanism is introduced into the original NSGA-II algorithm, specifically: The selection process in the first stage: enter the archive set according to the level of non-dominated sorting from low to high. When selecting the critical layer, select the individuals with smaller domain density to enter the archive set; in the selection process of the second stage: when the number of individuals in the first non-dominated layer exceeds the population size N, an elite selection mechanism based on aggregation function and neighborhood density is used to select individuals. By calculating the convergence index and diversity index, the two individuals x and y with the largest convergence index are selected, and then the individual with larger neighborhood density is deleted. This strategy of deletion one by one is used until the population size of the non-dominated layer is N, and these N individuals are selected to enter the archive set.

4. The multi-objective low-carbon scheduling optimization method for open-pit mine trucks based on a two-stage evolutionary selection mechanism according to claim 1 is characterized in that: In step 3, the NSGA-II algorithm based on the two-stage evolutionary selection mechanism solves the multi-objective low-carbon scheduling model of the open-pit mine. The specific solution process is as follows: Step 3.1, set the basic parameters of the algorithm, population size N, maximum number of iterations G max , the initial population is generated by random initialization; Step 3.2, the parent population is subjected to genetic operations simulating binary crossover and polynomial mutation to form a new offspring population; Step 3.3, merge the parent population and the offspring population into the same population; Step 3.4, calculate the target value, convergence index, and neighborhood density of all individuals in the population after merging in step 3.3; Step 3.5, according to the target values ​​of all individuals in the merged population obtained in step 3.4, non-dominated sorting is performed according to the distance Pareto dominance relationship; Step 3.6, perform a two-stage evolutionary selection mechanism on the same population obtained in step 3.3, output the archive set, and enter the next iteration; Step 3.7, the system determines whether the algorithm iteration state has reached the termination condition. If so, it outputs the optimal solution and target value; if not, it returns to step 3.2; The individual represents a solution for multi-objective low-carbon scheduling and optimization of transportation for open-pit mine trucks; the population represents multiple solutions for multi-objective low-carbon scheduling and optimization of transportation for open-pit mine trucks.

5. The multi-objective low-carbon scheduling optimization method for open-pit mine trucks based on a two-stage evolutionary selection mechanism according to claim 1 is characterized in that: The Pareto dominance relationship in step 3.5 is specifically: For a minimization problem, given two solutions x and y, solution x Pareto dominates solution y, and is marked as x < y, if and satisfies: If for two given solutions x and y, if solution x does not satisfy the Pareto domination condition of solution y, and solution y does not satisfy the Pareto domination condition of solution x, then solutions x and y are not mutually dominated and are marked as The specific process of non-dominated sorting in step 3.5 is as follows: First, sort the new population into non-dominated layers (Z1, Z2, .., Z n ), the critical layer (satisfying the conditions |Z1|+|Z2|+...+|Z i | <n,|Z1|+|Z2|+...+|Z i+1 |>n,Z i+1 The individuals before the critical layer are added to the offspring population, and elite selection is performed on the individuals in the critical layer. The process stops when the number of individuals in the offspring population reaches N.

6. The multi-objective low-carbon scheduling optimization method for open-pit mine trucks based on a two-stage evolutionary selection mechanism according to claim 1 is characterized in that: The process of performing the two-stage evolutionary selection mechanism in step 3.6 is: Step 3.6.1, the system obtains the ranking result of the non-dominated layer; Step 3.6.2, the system determines whether the population size of the first level is greater than the fixed size N. If so, in the first stage of selection, the individuals are sorted from low to high according to the non-dominated level to enter the archive set. When selecting the critical layer, individuals with better distribution and smaller domain density are selected to enter the archive set; if not, in the second stage, the elite selection mechanism based on aggregation function and neighborhood density is used to select individuals. By calculating the convergence index and diversity index, the two individuals x and y with the largest convergence index are selected, and then the individual with the larger neighborhood density is deleted. This strategy of deletion is used until the population size of the non-dominated layer is N, and these N individuals are selected to enter the archive set; The convergence index is calculated according to the target value, and the specific calculation method is as follows: Among them, f i (x) represents the value of the i-th objective function, and C(x) represents a simple aggregation function; In step 3.6.3, the system processes the archive set obtained in step 3.6.2, outputs the archive set, and enters the next iteration.

7. The multi-objective low-carbon scheduling optimization method for open-pit mine trucks based on a two-stage evolutionary selection mechanism according to claim 1 is characterized in that: The specific requirements for the termination conditions in step 3.7 are as follows: When seeking a multi-objective low-carbon scheduling optimization method for open-pit mine trucks based on a two-stage evolutionary selection mechanism, the stopping condition of the algorithm is set to the maximum number of iterations of the algorithm G max , when the number of iterations of the algorithm reaches the maximum number of iterations, the algorithm stops.

8. A multi-objective low-carbon scheduling optimization system for open-pit mine trucks based on a two-stage evolutionary selection mechanism, characterized in that: include: Low-carbon scheduling model module: Based on the production needs and indicators of the mine, a multi-objective low-carbon scheduling optimization model for open-pit mine trucks is constructed with the goals of minimizing the total waiting time of trucks, minimizing the total ore transportation cost, minimizing the ore grade deviation, and minimizing the total carbon emissions of transportation; Solution algorithm module: for a multi-objective low-carbon scheduling optimization method for open-pit mine trucks based on a two-stage evolutionary selection mechanism, an NSGA-II-TS algorithm based on a two-stage evolutionary selection mechanism is constructed; Decision-making module: The multi-objective low-carbon scheduling model of the open-pit mine obtained by the low-carbon scheduling model module is solved according to the constructed NSGA-II-TS algorithm, and a multi-objective low-carbon scheduling optimization result of open-pit mine trucks based on a two-stage evolutionary selection mechanism is obtained.

9. A multi-objective low-carbon scheduling optimization device for open-pit mine trucks based on a two-stage evolutionary selection mechanism, characterized in that: include: Memory: used to store a computer program for implementing the multi-objective low-carbon scheduling optimization method for open-pit mine trucks based on a two-stage evolutionary selection mechanism according to claims 1-7; A processor is used to implement the multi-objective low-carbon scheduling optimization method for open-pit mine trucks based on a two-stage evolutionary selection mechanism according to claims 1-7 when executing the computer program.

10. A computer-readable storage medium, characterized in that: include: The computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, it can implement claims 1-7, a multi-objective low-carbon scheduling optimization method for open-pit mine trucks based on a two-stage evolutionary selection mechanism.

Images