Dynamic optimization method and system for strip mine mining and selection collaborative plan coping with weather disturbance

Abstract

The invention provides a strip mine mining and selection collaborative plan dynamic optimization method and system for coping with weather disturbance, and relates to the technical field of strip mine mining and selection collaborative production plan optimization. The method specifically comprises the following steps: for any to-be-planned production cycle, acquiring weather forecast data of all planning periods in the production cycle, and respectively extracting time characteristics and meteorological characteristics of each planning period; constructing a bilevel programming model based on data-driven prediction and optimization decision coupling; and based on the time characteristics and the meteorological characteristics of all the planning periods in the production cycle, generating an optimal production plan of the production cycle by using a bilevel planning model based on data-driven prediction and optimization decision coupling. According to the method, weather changes can be dynamically responded, so that productivity fluctuation caused by weather uncertainty to strip mine production is dealt with, and ore supply stability and production economic benefits are guaranteed.

Description

Technical Field

[0001] This invention relates to the field of open-pit mine mining and beneficiation collaborative production planning optimization technology, and in particular to a dynamic optimization method and system for open-pit mine mining and beneficiation collaborative planning in response to weather disturbances. Background Technology

[0002] In the field of open-pit mine production scheduling optimization, existing technologies mainly focus on uncertainties such as geology, price, and supply. Among these, a typical technical solution for addressing geological uncertainties is the two-stage stochastic integer programming (SIP) method. This approach quantifies uncertain parameters such as geological grade and tonnage, constructs a stochastic optimization model, and incorporates geological uncertainties into the long-term production plan generation process to achieve optimal mining scheduling decisions. For example, Ramazan S, Dimitrakopoulos R. Production scheduling with uncertain supply: a new solution to the open pit mining problem [J]. Optimization and engineering, 2013, 14: 361-380. One implementation process proposed is as follows: first, the probability distribution of uncertain variables such as grade and tonnage is quantified using geological exploration data; then, a two-stage stochastic programming model is constructed. The first stage determines the initial mining plan, and the second stage adjusts the subsequent mining strategy based on feedback from actual geological conditions, ultimately outputting a production plan that balances stability and economy. This scheme uses the probability distribution of parameters such as geological grade and tonnage as core inputs, and generates a mining plan through two-stage planning, which can effectively reduce the impact of geological parameter fluctuations on production targets. However, its core flaw lies in its complete failure to consider the dynamic interference of weather factors on the actual production capacity of open-pit mines, leading to a disconnect between the plan execution and actual production capacity. From a technical perspective, the constraints of this scheme are designed only around static parameters such as geological reserves and theoretical equipment capacity, without incorporating the nonlinear impact of weather factors such as extreme temperatures, rainfall, and strong winds on mining efficiency and transportation safety. When such weather conditions are encountered in actual production, the plan generated by this scheme will present a contradiction of "theoretical production capacity meeting the target but actual production capacity plummeting," leading to problems such as ore supply shortages, equipment idleness, and chaotic spoil heap scheduling, failing to meet the stable production needs of open-pit mines under weather uncertainty.

[0003] To address market price uncertainty, a mature existing technology is a long-term planning method that combines price path simulation with the Imperialist Competitive Algorithm (ICA). This approach simulates metal price fluctuation paths using models such as GARCH and GBM, and uses the Imperialist Competitive Algorithm to solve for the optimal production plan under price uncertainty, thereby achieving cost control and profit maximization. For example, Mokhtarian Asl M, Sattarvand J. Integration of commodity price uncertainty in long-term open pit mine production planning byusing an imperialist competitive algorithm [J]. Journal of the SouthernAfrican Institute of Mining and Metallurgy, 2018, 118 (2): 165-172. One proposed implementation process is as follows: a price prediction model is constructed based on historical price data to generate multiple price fluctuation paths; with the goal of maximizing long-term profits, an optimization model is constructed that includes constraints such as mining volume, transportation routes, and equipment scheduling; and the Imperialist Competitive Algorithm is used to iteratively solve the problem, outputting a production plan adapted to different price scenarios.

[0004] This solution focuses on the impact of market price fluctuations on revenue, achieving cost control through simulated price paths and algorithmic optimization, effectively addressing market price volatility risks. However, it also fails to address the direct impact of weather factors on the production capacity of each stage of open-pit mining, stripping, and transportation, resulting in two major flaws: First, the production capacity parameters are set without considering the actual weather-related scenarios. The solution assumes that the daily production capacity of the open-pit mine is stable at its theoretical maximum, failing to quantify weather-induced production capacity fluctuations. This leads to significant discrepancies between the optimal mining volume and transportation costs calculated based on "stable production capacity" and the actual situation. For example, when prices are high, the solution plans maximum mining volume to increase revenue, but in reality, insufficient production capacity due to heavy rain results in additional costs due to equipment idling and redundant manpower. Second, it lacks a linkage between inventory adjustment mechanisms and weather-based production capacity forecasting. The solution only adjusts production rhythm based on price signals, without considering that short-term production capacity gaps caused by weather can be alleviated through inventory replenishment, nor does it establish a dynamic matching logic of "weather-production capacity-inventory." When extreme weather causes a sharp drop in production capacity, it cannot maintain stable ore supply through inventory adjustments, easily leading to risks such as ore processing plant supply disruptions and contract defaults, making it difficult to balance the dual objectives of price revenue and production stability.

[0005] In summary, while existing technologies have achieved optimized management of uncertainties such as geology and pricing, they have not deeply quantified the nonlinear impact of weather disturbances on open-pit mine production capacity, nor do they possess a systematic approach that combines weather forecasting with dynamic adjustments to production plans. The few studies involving environmental factors remain at the level of qualitative recommendations, failing to develop feasible quantitative models and implementation plans. This results in open-pit mines lacking effective basis for adjusting production plans when facing severe weather such as torrential rains, extreme temperatures, and strong winds, easily leading to problems such as equipment idleness, ore supply interruptions, and soaring costs, making it difficult to meet the actual needs of stable operation of open-pit mines.

[0006] Moreover, in addition to the aforementioned specific deficiencies, existing technologies all suffer from the common problem of "lack of quantitative analysis of weather factors and a gap in the planning adjustment mechanism": existing studies may only qualitatively mention the impact of weather on production, but have not constructed a quantitative model between weather factors and production capacity, and cannot accurately output the range of production capacity fluctuations under different weather conditions; at the same time, a complete closed-loop chain of "weather forecasting - production capacity forecasting - dynamic planning adjustment" has not been formed, and production plans are mostly statically formulated, lacking an execution mechanism to optimize mining site allocation, waste rock discharge, and inventory scheduling in real time based on short-term weather changes. As a result, open-pit mines lack data support for production decisions when facing weather uncertainties, and have weak risk resistance capabilities. Summary of the Invention

[0007] To address the shortcomings of existing technologies, this invention proposes a dynamic optimization method and system for open-pit mining and beneficiation collaborative planning by employing a two-layer model coupling and inventory adjustment mechanism. This aims to solve the problems of insufficient quantification of weather factors, rigid planning, and lack of collaborative optimization in existing technologies. It provides a production planning method that can dynamically respond to weather changes and achieve collaborative optimization of mining and beneficiation, applicable to production processes such as mining, stripping, transportation scheduling, capacity allocation, and inventory management in open-pit mines. The core purpose of this invention is to address the capacity fluctuations caused by weather uncertainty in open-pit mine production, ensuring the stability of ore supply and production economic efficiency. It is suitable for various open-pit mining and beneficiation enterprises affected by natural weather.

[0008] On the one hand, this invention proposes a dynamic optimization method for open-pit mining and beneficiation coordination plans to cope with weather disturbances, which includes the following process:

[0009] For any production cycle to be planned, obtain the weather forecast data for all planned periods within that production cycle, and extract the temporal and meteorological characteristics for each planned period.

[0010] Construct a two-level programming model based on the coupling of data-driven prediction and optimization decision-making;

[0011] Based on the time and weather characteristics of all planning periods within the production cycle, the optimal production plan for the production cycle is generated using a two-level programming model that couples data-driven prediction and optimization decision-making.

[0012] Furthermore, for any production cycle to be planned, the specific content of the time characteristics and meteorological characteristics of all planned periods within that production cycle is extracted as follows:

[0013] For any planned period within the production cycle, obtain the weather forecast data for that planned period;

[0014] Extract the temporal characteristics and raw meteorological characteristics of the planned period from the weather forecast data for that period;

[0015] The time features mentioned therein include: the month and season corresponding to the date;

[0016] The original meteorological characteristics include: maximum temperature, minimum temperature, weather category, and wind force level;

[0017] The extracted raw meteorological features are preprocessed to obtain the preprocessed meteorological features;

[0018] The extracted temporal features and preprocessed meteorological features will be used as the temporal and meteorological features for the planning period.

[0019] Furthermore, the preprocessing includes: numerically encoding the weather category; and standardizing the maximum temperature, minimum temperature, and wind force level.

[0020] Furthermore, the two-layer planning model based on the coupling of data-driven prediction and optimization decision-making includes: an upper-layer capacity prediction model and a lower-layer mining, stripping, and transportation planning model;

[0021] The upper-level capacity prediction model is used to predict production data within the planned production cycle based on the time and meteorological characteristics of the production cycle to be planned.

[0022] The lower-level mining and transportation planning model is used to generate the optimal production plan for the production cycle by taking the production data within the production cycle as constraints and minimizing the weighted sum of the mining and transportation cost deviation and the ore supply deviation.

[0023] Furthermore, the method for constructing the upper-level capacity prediction model is as follows:

[0024] Collect historical time characteristics, historical meteorological characteristics, and historical production data corresponding to several historical planning periods; wherein the historical production data refers to the actual daily rock mining volume.

[0025] The historical meteorological characteristics corresponding to each historical planning period are preprocessed to obtain the preprocessed historical meteorological characteristics.

[0026] Using historical time features and preprocessed historical meteorological features of the same historical planning period as a training sample, and using historical production data of the same historical planning period as prediction labels, a training sample set is constructed.

[0027] Based on the random forest framework, a quantile regression forest model with multiple decision trees is constructed. With the goal of minimizing quantile loss, the quantile regression forest model is trained using the training sample set to obtain the trained quantile regression forest model.

[0028] The trained quantile regression forest model is used as the upper-level productivity prediction model.

[0029] Furthermore, based on the random forest framework, a quantile regression forest model with multiple decision trees is constructed. The model is trained using a training sample set with the objective of minimizing quantile loss. The specific content of the trained quantile regression forest model is as follows:

[0030] Initialize the empty quantile regression forest model;

[0031] The Bootstrap sampling method is used to sample the training sample set to generate several training sub-samples;

[0032] Using each training subsample, independently train one decision tree in the quantile regression forest model;

[0033] For any decision tree to be trained, determine the target quantile of the current decision tree, and use the training subsamples corresponding to the current decision tree as the root node data. Starting from the root node, recursively execute the node splitting process; the specific content of each node splitting process is as follows:

[0034] For any node to be split in the current decision tree, determine whether the node satisfies the preset stopping split condition. If it does, the node to be split is taken as a leaf node, and the trained current decision tree is obtained. If it does not, the optimal split point of the node to be split is obtained by minimizing the quantile loss as the node splitting criterion, and the next level splitting node is generated based on the optimal split point.

[0035] Furthermore, the objective function of the lower-level mining and transportation planning model is to minimize the weighted sum of the mining and transportation cost deviation and the ore supply deviation.

[0036] Furthermore, the constraints of the lower-level mining and transportation planning model include:

[0037] Mining capacity constraints of mining equipment: planning period mining area The amount of rock extracted shall not exceed the planned period. mining area The maximum production capacity limit of the equipment;

[0038] Mining capacity constraints: planning period The total rock extraction volume of all mining areas within the planned period is within the planned period. Within the feasible range of production capacity;

[0039] Stope reserves constraints: Stope The total rock extraction volume during all planned periods within the unplanned production cycle shall not exceed the mining area. The upper limit of rock quantity;

[0040] Stripping ratio constraint: for any planning period mining area The total amount of waste rock transported to the spoil heap is equal to the amount transported from the quarry. Total amount of ore mined and mining sites The product of the stripping ratio; wherein the stripping ratio from the mining area The total amount of ore extracted from the mine is: The amount of ore and from the mining site The sum of the ore shipped to storage;

[0041] Waste disposal capacity constraint: All waste transported to the waste disposal site The total amount of waste rock does not exceed the amount of waste rock dump. Maximum capacity;

[0042] Material balance constraints: for any planning period mining area The amount of rock extracted is equal to the amount extracted from the quarry. ore quantity, mining area The amount of ore shipped to storage and all ore from the mine The total amount of waste rock transported to the spoil heap;

[0043] Mineral processing plant grade constraints: for any metal During any planning period Within the total amount of ore transported from all mines to the concentrator, metals... The content is located in the ore dressing plant for metal The required feasible range for metal content;

[0044] Inventory balance constraint: planning period mining area The inventory level is equal to that of the previous planning period. mining area Inventory levels, plus the planned period mining area The amount of ore shipped to storage, minus the planned period. The ore transported from the mine to the concentrator comes from the stockpile. ore quantity;

[0045] Maximum inventory constraint: planning period The sum of the inventory levels in all mining areas shall not exceed the total inventory level of the mining area. The upper limit of ore storage capacity;

[0046] Inventory reservation constraints: planning period mining area The inventory level is not less than the inventory level for the purchasing site. The lower limit of ore reserves;

[0047] Positive Deviation Constraint on Ore Quantity: Planning Period The positive deviation between the internal ore supply and the demand of the ore processing plant is equal to the planned period. The amount of ore transported from all mining sites to the storage area;

[0048] Negative deviation constraint on ore quantity: This will affect the planning period. The difference between the ore demand of the internal concentrator and the total supply of all mines is compared with zero, and the larger value is taken as the planning period. The negative deviation between the internal ore supply and the demand of the ore processing plant;

[0049] Cost deviation constraints: planning period The total planned cost within the planning period equals the total planned cost during the planning period. Total cost within the planning period The positive deviation between total cost and planned cost, minus the planned period cost. The negative deviation between total internal cost and planned cost;

[0050] Direct ore transport constraint: for any planning period All from the mining area The amount of ore transported to the concentrator and the amount of ore transported from the inventory to the concentrator shall not exceed the maximum daily ore processing capacity of the concentrator.

[0051] Waste dump utilization equilibrium constraint: For any waste dump spoil heap The total amount of waste rock during all planning periods is located in the spoil heap. Within the acceptable range of waste rock volume allowed by the balanced utilization rate;

[0052] Inbound and outbound mutual exclusion constraints: in any planning period Inside, mining area The amount of ore shipped to storage and the amount of ore shipped from storage to the concentrator originating from the mine. The amount of ore cannot be positive at the same time;

[0053] Non-negativity constraint: Planning period Positive and negative deviations between internal ore supply and the demand of the concentrator, and the planning period. Positive and negative deviations between total internal cost and planned cost, planning period mining area Rock mining volume, planned period mining area ore quantity, planning period mining area The amount of ore shipped to storage and the planned period The ore transported from the mine to the concentrator comes from the stockpile. The ore quantities are all non-negative.

[0054] Furthermore, the optimal production plan includes the mining volume of each mining site, transportation routes, inventory inflow and outflow status, and planned costs.

[0055] On the other hand, this invention proposes a dynamic optimization system for open-pit mining and beneficiation coordination planning to cope with weather disturbances, the system comprising:

[0056] The data acquisition module is used to acquire weather forecast data for all planned periods within any production cycle;

[0057] The feature extraction module is used to extract the temporal and meteorological features for each planning period based on the weather forecast data for all planning periods.

[0058] The model building module is used to build a two-level programming model based on the coupling of data-driven prediction and optimization decision-making.

[0059] The plan generation module is used to generate the optimal production plan for the production cycle based on the time and meteorological characteristics of all planning periods within the production cycle, using a two-level planning model that couples data-driven prediction and optimization decision-making.

[0060] The beneficial effects of adopting the above technical solution are as follows:

[0061] This invention innovatively constructs a two-layer model architecture of "data-driven prediction-optimization decision coupling," forming a technical system with deep linkage between the upper and lower layers. This provides a method for optimizing mining and beneficiation collaborative production plans that considers the impact of weather. It overcomes the shortcomings of existing technologies that only address geological or price uncertainties, fail to quantify the dynamic impact of weather on production capacity, lack a "weather-production-plan" linkage adjustment mechanism, and have insufficient adaptability of inventory adjustments to weather conditions. A detailed analysis follows:

[0062] The upper layer of the described two-layer model architecture employs a quantile regression forest method, integrating temporal and meteorological features as input variables. This not only outputs the median predicted daily rock output but also generates a 90% confidence interval to quantify the range of production capacity fluctuations, accurately capturing the nonlinear impact of extreme weather on production capacity. In other words, by constructing a weather-driven production capacity prediction model, it accurately captures the nonlinear impact of extreme temperatures, rainfall, strong winds, and other weather conditions on the daily rock output of open-pit mines, outputting the range of production capacity fluctuations including the median and confidence intervals. This provides a production capacity basis that aligns with actual weather scenarios for production planning, avoiding problems such as ore supply shortages and equipment idleness caused by "theoretical production capacity meeting targets but actual production capacity plummeting."

[0063] The lower-level mining, stripping, and transportation planning model in the described two-layer model architecture aims to minimize the weighted sum of "mining, stripping, and transportation cost deviation + ore supply deviation," using the upper-level production capacity forecast as the core constraint, rather than statically using theoretical production capacity parameters. This achieves dynamic adaptation between "prediction" and "planning," solving the problems of existing geological uncertainty optimization schemes based on two-stage stochastic integer programming that do not incorporate weather factors and are disconnected from actual production capacity. This mining, stripping, and transportation planning model, combined with a dynamic inventory adjustment mechanism, achieves coordination between mine production capacity allocation, waste rock discharge, and ore supply under weather-related production capacity fluctuations. It avoids cost deviations caused by defaulting to "stable production capacity" and alleviates supply gaps caused by extreme weather through inventory replenishment, balancing production economics and stability. This overcomes the shortcomings of price uncertainty optimization schemes combining price path simulation and imperial competition algorithms, where production capacity parameters are detached from actual weather conditions and inventory adjustments are not linked to weather.

[0064] To address the shortcomings of existing technologies that lack linkage between inventory and weather, this invention constructs a multi-dimensional inventory constraint system within the planning model, designing a dynamic inventory adjustment mechanism that adapts to weather and production capacity. This multi-dimensional inventory constraint system includes upper and lower inventory limits, mutual exclusion constraints for inbound and outbound operations, and inventory reservation rules, enabling flexible adjustments to inventory based on weather and production capacity fluctuations. When weather causes a sharp drop in production capacity, the model automatically extracts inventory to fill the supply gap; when production capacity is sufficient and weather is stable, excess ore from low-cost mining areas is prioritized for storage in inventory, preventing idle inventory resources and providing a buffer for subsequent extreme weather events. This forms a dynamic balance logic of "weather-production capacity-inventory," achieving a dual guarantee of short-term stable supply and long-term inventory security.

[0065] This invention comprehensively incorporates actual production constraints such as equipment capacity, stripping ratio, waste dump capacity, and ore dressing plant grade requirements into the planning model. Under the premise of compliance, it achieves coordinated scheduling of mining capacity allocation, waste rock discharge paths, and ore transportation flow through optimized algorithms. Simultaneously, the model supports receiving short-term weather forecast data, updating capacity prediction results in real time, and dynamically adjusting planning parameters, forming a closed-loop chain of "weather forecasting, accurate capacity forecasting, mining and dressing plan optimization, and execution feedback." This replaces the static planning model of existing technologies, enabling production decisions to respond to weather changes in real time, allowing open-pit mine production plans to be adjusted according to short-term weather changes, improving the mine's decision-making ability in response to extreme weather, ensuring a continuous and stable ore supply from the dressing plant, reducing risks such as equipment idling and contract defaults, enhancing the open-pit mine's ability to cope with uncertainty, and achieving high efficiency and anti-interference in open-pit mining and dressing collaborative production. It overcomes the common problems of "lack of quantitative analysis of weather factors and gaps in planning adjustment mechanisms" in existing technologies. Attached Figure Description

[0066] Figure 1 This is a flowchart of a dynamic optimization method for open-pit mining and beneficiation coordination plan in response to weather disturbances, as described in an embodiment of the present invention.

[0067] Figure 2 This is a schematic diagram of the model framework for a dynamic optimization method for open-pit mining and beneficiation coordination plan in response to weather disturbances, as described in an embodiment of the present invention.

[0068] Figure 3 This is a schematic diagram of the capacity prediction model in an embodiment of the present invention;

[0069] Figure 4 This is a structural diagram of a dynamic optimization system for open-pit mining and beneficiation coordination planning in response to weather disturbances, as described in an embodiment of the present invention. Detailed Implementation

[0070] To facilitate understanding of this application, specific embodiments of the present invention will be described in further detail below with reference to the accompanying drawings and embodiments. The following embodiments are illustrative of the invention but are not intended to limit its scope. Rather, these embodiments are provided to provide a more thorough and complete understanding of the disclosure of this application.

[0071] Example 1:

[0072] This embodiment presents a dynamic optimization method for open-pit mining and beneficiation coordination plans to cope with weather disturbances, such as... Figure 1 As shown, the method includes the following steps:

[0073] For any production cycle to be planned, obtain the weather forecast data for all planned periods within that production cycle, and extract the temporal and meteorological characteristics for each planned period.

[0074] For any production cycle to be planned, the weather forecast data for all planned periods within that production cycle is obtained, and the specific content of the time characteristics and meteorological characteristics for each planned period is extracted as follows:

[0075] For any planned period within the production cycle, obtain the weather forecast data for that planned period.

[0076] Extract the temporal and raw meteorological characteristics of the planned period from the weather forecast data for that period.

[0077] The time features mentioned therein include the month and season corresponding to the date.

[0078] The original meteorological characteristics include: maximum temperature, minimum temperature, weather category, and wind force level.

[0079] The extracted raw meteorological features are preprocessed to obtain the preprocessed meteorological features.

[0080] The preprocessing includes: numerically encoding the weather category; and standardizing the maximum temperature, minimum temperature, and wind force level.

[0081] In this embodiment, weather categories are numerically coded as follows: "Sunny" is set as 1, "Cloudy, Overcast, Light Rain, Light Snow" as 2, "Showers, Thunderstorms, Snow Showers" as 3, "Thunderstorms with Hail, Sleet, Moderate Rain, Moderate Snow" as 4, and "Heavy Rain, Torrential Rain, Extremely Heavy Rain, Freezing Rain, Heavy Snow, Blizzard, Extremely Heavy Rain" as 5. Numerical characteristics such as temperature and wind force are standardized.

[0082] The extracted temporal features and preprocessed meteorological features will be used as the temporal and meteorological features for the planning period.

[0083] Construct a two-level programming model based on the coupling of data-driven prediction and optimization decision-making.

[0084] The two-layer planning model based on the coupling of data-driven prediction and optimization decision-making includes: an upper-layer capacity prediction model and a lower-layer mining, stripping and transportation planning model.

[0085] In this embodiment, as Figure 2 As shown, the upper-level production capacity prediction model uses a fusion of time and meteorological characteristics as input variables, outputting the median predicted value of daily rock mining volume and quantifying the range of production capacity fluctuations to accurately capture the nonlinear impact of extreme weather on production capacity. The lower-level mining, stripping, and transportation planning model uses the upper-level production capacity prediction results as the core constraint to achieve dynamic adaptation between "prediction" and "planning".

[0086] The upper-level capacity prediction model is used to predict production data within the planned production cycle based on the time and meteorological characteristics of the production cycle.

[0087] The method for constructing the upper-level capacity prediction model is as follows:

[0088] Collect historical time characteristics, historical meteorological characteristics, and historical production data corresponding to several historical planning periods; wherein the historical production data refers to the actual daily rock mining volume.

[0089] The historical meteorological characteristics corresponding to each historical planning period are preprocessed to obtain the preprocessed historical meteorological characteristics.

[0090] Using historical time features and preprocessed historical meteorological features of the same historical planning period as a training sample, and using historical production data of the same historical planning period as prediction labels, a training sample set is constructed.

[0091] In this embodiment, as Figure 3 As shown, for the upper-level production capacity prediction model, historical time features and historical meteorological features are used as input features, and the actual daily rock mining volume of the corresponding historical period is used as the prediction label to form a model training sample set. Among them, it is necessary to numerically encode the weather category and standardize the numerical features such as temperature and wind force level to ensure that the sample set meets the model training requirements.

[0092] Based on the random forest framework, a quantile regression forest model with multiple decision trees is constructed. With the goal of minimizing quantile loss, the quantile regression forest model is trained using the training sample set to obtain the trained quantile regression forest model.

[0093] In this embodiment, the training objective is to construct an integrated model with multiple decision trees, enabling the model to accurately fit the changing patterns of daily rock production through input features, and to quantify the production capacity fluctuations caused by weather disturbances through quantile output.

[0094] The method described above involves constructing a quantile regression forest model based on a random forest framework, integrating multiple decision trees, and aiming to minimize the quantile loss. The model is trained using a training sample set, and the specific content of the trained quantile regression forest model is as follows:

[0095] Initialize the empty quantile regression forest model.

[0096] The Bootstrap sampling method is used to sample the training sample set to generate several training sub-samples.

[0097] In this embodiment, the training objective is to construct an ensemble model with multiple decision trees, enabling the model to accurately fit the daily rock extraction volume variation pattern through input features and to quantify the production capacity fluctuations caused by weather disturbances through quantile output. Specifically, based on the random forest framework, hyperparameters such as the number of decision trees, the maximum depth of a single tree, and the minimum number of samples required for node splitting are set to initialize an empty ensemble model. Then, the Bootstrap sampling method is used on the training sample set to extract independent training subsamples for each decision tree.

[0098] Using each training subsample, independently train one decision tree in the quantile regression forest model.

[0099] For any decision tree to be trained, determine the target quantile of the current decision tree, and use the training subsamples corresponding to the current decision tree as the root node data. Starting from the root node, recursively execute the node splitting process; the specific content of each node splitting process is as follows:

[0100] For any node to be split in the current decision tree, determine whether the node satisfies the preset stopping split condition. If it does, the node to be split is taken as a leaf node, and the trained current decision tree is obtained. If it does not, the optimal split point of the node to be split is obtained by minimizing the quantile loss as the node splitting criterion, and the next level splitting node is generated based on the optimal split point.

[0101] In this embodiment, for each decision tree, the optimal split point is found by traversing the input features, with the minimum quantile loss as the node splitting criterion, thus completing the structure construction of a single tree.

[0102] The loss function for the quantile loss is defined as follows:

[0103]

[0104] in, This indicates the actual daily rock extraction volume; This indicates the predicted daily rock extraction volume; In this embodiment, the target quantile is represented as... We take 0.05, 0.5, and 0.95, which correspond to the lower limit, median, and upper limit of the interval, respectively. Indicates the actual daily rock extraction volume Compared with the predicted daily rock mining volume The quantile loss value.

[0105] In this embodiment, the loss function of quantile loss adapts to the nonlinear effects of weather disturbances through asymmetric penalty: when extreme weather (strong winds, blizzards) causes the actual rock mining volume to be lower than the predicted value, the weighted loss function is applied. Strengthen penalties for "underestimation bias"; when suitable weather leads to actual rock extraction exceeding the predicted value, a weighted penalty will be imposed. Strengthening the penalty for "overestimation bias" enables individual trees to accurately learn the production capacity correlation rules under different weather scenarios. Finally, repeat the above training steps to complete the training of all decision trees, and combine all decision trees to form the final quantile regression forest model.

[0106] The trained quantile regression forest model is used as the upper-level productivity prediction model.

[0107] In this embodiment, the trained quantile regression forest model is used as the upper-level production capacity prediction model. The feature data of the future planning period is input to predict the future rock mining volume. The upper and lower limits of the confidence interval of the prediction result are used as the upper and lower limits of the production capacity constraint of the lower-level model to quantify the production capacity fluctuation risk under weather disturbances.

[0108] The lower-level mining and transportation planning model is used to generate the optimal production plan for the production cycle by taking the production data within the production cycle as constraints and minimizing the weighted sum of the mining and transportation cost deviation and the ore supply deviation.

[0109] The underlying mining and transportation planning model is as follows:

[0110] The objective function for constructing the lower-level mining, stripping, and transportation planning model is to minimize the weighted sum of the mining, stripping, and transportation cost deviation and the ore supply deviation.

[0111] The objective function of the lower-level mining and transportation planning model is expressed as:

[0112]

[0113] in Weighting for ore supply deviation; Cost deviation weight; and respectively planning period The positive and negative deviations between the internal ore supply and the demand of the ore processing plant; For the planning period set; and respectively planning period The positive or negative deviation between total internal cost and planned cost.

[0114] Define the constraints of the lower-level mining and transportation planning model, including:

[0115] (1) Mining capacity constraints of mining equipment in the mining area: planning period mining area The amount of rock extracted shall not exceed the planned period. mining area The maximum production capacity limit of the equipment.

[0116] The mining capacity constraint of the mining equipment is expressed as follows:

[0117]

[0118] in, Indicates the planning period mining area The amount of rock mined; Indicates the planning period mining area The maximum production capacity limit of the equipment; This indicates the assembly at the mining site.

[0119] (2) Mine capacity constraints: planning period The total rock extraction volume of all mining areas within the planned period is within the planned period. Within the feasible range of production capacity.

[0120] The mine capacity constraint is expressed as follows:

[0121]

[0122] in, and They represent the planning period. The minimum production capacity and the maximum production capacity.

[0123] (3) Stope reserves constraints: Stope The total rock extraction volume during all planned periods within the unplanned production cycle shall not exceed the mining area. The upper limit of the amount of rock.

[0124] The reserve constraint in the stope is expressed as follows:

[0125]

[0126] in, Indicates mining area The upper limit of the amount of rock.

[0127] (4) Stripping ratio constraint: for any planning period mining area The total amount of waste rock transported to the spoil heap is equal to the amount transported from the quarry. Total amount of ore mined and mining sites The product of the stripping ratio; wherein the stripping ratio from the mining area The total amount of ore extracted from the mine is: The amount of ore and from the mining site The sum of the amount of ore shipped to the warehouse.

[0128] The stripping ratio constraint is expressed as follows:

[0129]

[0130] in, Indicates the planning period From the mining site Transported to the spoil heap The amount of waste rock; Indicates the number of spoil heaps; Indicates the planning period mining area ore quantity; Indicates the planning period mining area The amount of ore shipped to storage; For mining area The stripping ratio.

[0131] (5) Waste disposal capacity constraints: All waste transported to the waste disposal site The total amount of waste rock does not exceed the amount of waste rock dump. The maximum capacity.

[0132]

[0133] in, Indicates spoil heap The maximum capacity.

[0134] (6) Material balance constraints: for any planning period mining area The amount of rock extracted is equal to the amount extracted from the quarry. ore quantity, mining area The amount of ore shipped to storage and all ore from the mine The total amount of waste rock transported to the spoil heap.

[0135] The material balance constraint is expressed as follows:

[0136]

[0137] (7) Grade constraints of ore dressing plants: for any metal During any planning period Within the total amount of ore transported from all mines to the concentrator, metals... The content is located in the ore dressing plant for metal The required feasible range for metal content.

[0138] The grade constraint of the ore dressing plant is expressed as follows:

[0139]

[0140] in, and These represent the ore dressing plant's treatment of metals. The lower limit and upper limit of taste requirements; Indicates the planning period The ore transported from the mine to the concentrator comes from the stockpile. ore quantity; Indicates mining area Metal Taste; This represents a set of metal types.

[0141] (8) Inventory balance constraint: planning period mining area The inventory level is equal to that of the previous planning period. mining area Inventory levels, plus the planned period mining area The amount of ore shipped to storage, minus the planned period. The ore transported from the mine to the concentrator comes from the stockpile. The amount of ore.

[0142] The inventory balance constraint is expressed as:

[0143]

[0144] in, Indicates the planning period mining area Inventory levels; Indicates the previous planning period mining area Inventory levels.

[0145] (9) Maximum inventory constraint: planning period The sum of the inventory levels in all mining areas shall not exceed the total inventory level of the mining area. The upper limit of ore storage.

[0146] The maximum inventory constraint is expressed as:

[0147]

[0148] in, This indicates that inventory is relevant to the purchasing market. The upper limit of ore storage.

[0149] (10) Inventory reservation constraints: planning period mining area The inventory level is not less than the inventory level for the purchasing site. The lower limit of ore reserves.

[0150] The inventory reservation constraint is expressed as follows:

[0151]

[0152] in, This indicates that inventory is relevant to the purchasing market. The lower limit of ore reserves.

[0153] (11) Positive deviation constraint of ore quantity: planning period The positive deviation between the internal ore supply and the demand of the ore processing plant is equal to the planned period. The amount of ore transported from all mining sites to the storage area.

[0154] The positive deviation constraint on the ore quantity is expressed as follows:

[0155]

[0156] (12) Negative deviation constraint on ore quantity: the planning period The difference between the ore demand of the internal concentrator and the total supply of all mines is compared with zero, and the larger value is taken as the planning period. The internal supply of ore has a negative deviation from the demand of the ore processing plant.

[0157] Among them, the planning period The total supply of all mining areas within the planned period The sum of the supply from all mining areas, and the supply from each mining area is: (Planning period) mining area Rock extraction volume minus the amount extracted from the quarry The total amount of waste rock transported to the spoil heap, plus the amount of ore from the mine transported from the storage to the concentrator. The amount of ore.

[0158] The negative deviation constraint on ore quantity is expressed as follows:

[0159]

[0160] in, Indicates the planning period The ore demand of the ore processing plant.

[0161] (13) Cost deviation constraint: planning period The total planned cost within the planning period equals the total planned cost during the planning period. Total cost within the planning period The positive deviation between total cost and planned cost, minus the planned period cost. The negative deviation between total internal cost and planned cost.

[0162] The cost deviation constraint is expressed as follows:

[0163]

[0164] in, Indicates the planning period Total planned cost within the period; Indicates the planning period The total cost within, and we have:

[0165]

[0166] in, Indicates mining area The unit mining cost; Indicates mining area Unit transportation cost to the ore processing plant; Indicates mining area Transportation distance to the ore processing plant; Indicates mining area Unit transportation cost to inventory; Indicates mining area Transportation distance to the inventory; This represents the unit transportation cost from inventory to the ore dressing plant; The transportation distance from the inventory to the ore processing plant; mining area To the spoil heap The unit transportation cost; Indicates mining area To the spoil heap The transportation distance.

[0167] (14) Direct ore transport constraint: for any planning period All from the mining area The amount of ore transported to the concentrator and the amount of ore transported from the inventory to the concentrator shall not exceed the maximum daily ore processing capacity of the concentrator.

[0168] The direct ore transport constraint is expressed as follows:

[0169]

[0170] in, This indicates the maximum daily ore processing capacity of the ore processing plant.

[0171] (15) Waste dump utilization equilibrium constraint: For any waste dump spoil heap The total amount of waste rock during all planning periods is located in the spoil heap. Within the acceptable range of waste rock volume for balanced utilization.

[0172] The constraint on the utilization rate equilibrium of the spoil heap is expressed as follows:

[0173]

[0174]

[0175]

[0176] in, The allowable deviation for the balanced utilization rate of the spoil heap; This indicates the total amount of waste rock from all spoil heaps during the planning period; spoil heap The percentage of capacity.

[0177] (16) Mutual exclusion constraint for inbound and outbound operations: During any planning period Inside, mining area The amount of ore shipped to storage and the amount of ore shipped from storage to the concentrator originating from the mine. The amount of ore cannot be positive at the same time.

[0178] The mutual exclusion constraint for inbound and outbound operations is expressed as follows:

[0179]

[0180] in, , For binary variables, Used to indicate a mining area During the planning period Whether to ship to inventory.

[0181] (17) Non-negativity constraint: planning period Positive and negative deviations between internal ore supply and the demand of the concentrator, and the planning period. Positive and negative deviations between total internal cost and planned cost, planning period mining area Rock mining volume, planned period mining area ore quantity, planning period mining area The amount of ore shipped to storage and the planned period The ore transported from the mine to the concentrator comes from the stockpile. The ore quantities are all non-negative.

[0182] The nonnegativity constraint is expressed as:

[0183]

[0184] In this embodiment, Python is used to call the CPLEX solver to solve the lower-level mining and transportation plan model to obtain the optimal production plan, including the mining volume of each mining area, transportation route, inventory inflow and outflow status, and planned cost.

[0185] Based on the time and weather characteristics of all planning periods within the production cycle, the optimal production plan for the production cycle is generated using a two-level programming model that couples data-driven prediction and optimization decision-making.

[0186] Existing technologies often assume that open-pit mine production capacity is stable at a theoretical value, which cannot account for dynamic fluctuations caused by weather. Therefore, existing technologies suffer from the problem of "production capacity deviating from actual weather conditions." This embodiment, however, uses a quantile regression forest model to achieve accurate predictions of weather scenarios in different seasons. The specific results are as follows:

[0187] In the experimental verification during the 15-day planning period, the median fluctuation range of the predicted daily rock production under extreme low temperature scenarios in winter was 132,000-199,000 tons, and the 90% confidence interval width was controlled within 20,000-50,000 tons. The actual production capacity all fell within the confidence interval, accurately reflecting the inhibitory effect of low temperature on equipment efficiency and avoiding the planning failure caused by the existing technology "overestimating the low temperature production capacity".

[0188] In the context of high temperatures and heavy rainfall during the summer, the confidence interval narrows sharply to 349,900-350,000 tons, clearly defining the production capacity constraint boundary under severe weather conditions. Compared with the existing technology's production capacity setting that "ignores the impact of rainfall," the prediction bias is reduced. In the context of variable weather in autumn, the median prediction reaches as high as 490,000 tons under sunny weather conditions, and remains stable at around 350,000 tons under rainy weather conditions. The model can automatically adjust the prediction results according to the weather type, and its adaptability is significantly better than the existing static production capacity setting method, thereby achieving a significant improvement in production capacity prediction accuracy and scenario adaptability.

[0189] To address the existing "ore supply-cost" dilemma, this embodiment utilizes a two-layer model and an inventory adjustment mechanism to achieve balanced control of ore supply and costs across multiple experimental scenarios. Specific comparative results are as follows:

[0190] (1) Stability of ore supply: In the scenario of sufficient inventory, when the initial inventory of each mining site increases, even if the daily ore production fluctuates due to weather, the model extracts inventory that is completely matched with the supply and demand difference every day, so that the supply and demand of the concentrator are consistent throughout the 15 days, and the negative deviation of the ore supply is 0, which completely solves the risk of "ore supply interruption" under extreme weather in the existing technology; In the scenario of normal inventory, when considering weather production capacity constraints, the total negative deviation of the ore supply within 15 days is controlled within a reasonable range. Compared with the scenario where weather is not considered and production capacity is reduced, the risk of ore supply interruption is reduced.

[0191] (2) The cost control experiment was a comparative experiment during a 15-day planning period. Under the scenario of weather-related production capacity constraints, the total production cost was RMB47.42 million and the average daily production cost was RMB3.16 million. Compared with the scenario of not considering weather but reducing the production capacity of the mining area, the total production cost was reduced by 18.9% and the average daily cost was reduced by 19.0%. By optimizing the waste rock discharge path, the utilization rate of the three waste rock dumps was only 0.96% at most, and the others were 0.75% and 0.53% respectively, all of which were far below the capacity limit. This avoided the problem of "waste rock accumulation leading to additional transportation costs" in the existing technology, and at the same time reduced the waste of idle resources in the waste rock dumps.

[0192] In addition, the model prioritizes low-cost mining sites for extraction and inventory replenishment, which further reduces mining costs compared to the existing technology's "indiscriminate mining of high-cost sites," achieving a dual optimization of ore supply stability and cost control.

[0193] To address the problems of static planning failing to respond to weather changes, decision-making lags under extreme weather conditions, and high equipment idle rates in existing technologies, this embodiment employs a closed-loop decision-making process across the entire chain, enabling production plans to adapt to weather changes in real time. The specific effects are as follows:

[0194] Taking the variable weather scenario in September as an example, in sunny weather, the model increases the median daily ore production forecast to 468,000-490,000 tons, while simultaneously reducing inventory withdrawals, prioritizing the use of current capacity to meet demand, reducing inventory consumption, and improving equipment utilization. In rainy weather, the median forecast stabilizes at around 350,000 tons, and the model increases inventory withdrawals to maintain stable ore supply through inventory replenishment, avoiding equipment idling due to insufficient capacity and reducing equipment idle rate. Compared with existing technologies, this invention improves the mine's decision-making response speed to weather changes, reduces the risk of ore supply interruption at the concentrator, and allows for flexible adjustment of model inputs according to production parameters of different climate zones and different mineral types, enhancing the risk resistance and adaptability of production decisions, improving the resilience of open-pit mine production, and providing strong support for the stable operation of open-pit mines in complex weather environments.

[0195] Example 2:

[0196] This embodiment presents a dynamic optimization system for open-pit mining and beneficiation coordination planning to cope with weather disturbances, such as... Figure 4 As shown, the system includes:

[0197] The data acquisition module is used to acquire weather forecast data for all planned periods within any production cycle.

[0198] The feature extraction module is used to extract the temporal and meteorological features for each planning period based on the weather forecast data for all planning periods.

[0199] The model building module is used to build a two-level programming model based on the coupling of data-driven prediction and optimization decision-making.

[0200] The plan generation module is used to generate the optimal production plan for the production cycle based on the time and meteorological characteristics of all planning periods within the production cycle, using a two-level planning model that couples data-driven prediction and optimization decision-making.

[0201] Example 3:

[0202] This embodiment proposes an electronic device, including: one or more processors, and a memory, wherein the memory is used to store instructions, and when the instructions are executed by the one or more processors, the one or more processors execute the dynamic optimization method for open-pit mining and beneficiation coordination plan in response to weather disturbances.

[0203] The electronic device may be a mobile phone, computer, or tablet computer, etc., and includes a memory and a processor. The memory stores a computer program, which, when executed by the processor, implements the dynamic optimization method for open-pit mining and beneficiation coordination plan in response to weather disturbances as described in the embodiments. It is understood that the electronic device may also include input / output (I / O) interfaces and communication components.

[0204] The processor is used to execute all or part of the steps in the dynamic optimization method for coordinated open-pit mining and beneficiation planning in response to weather disturbances, as described in the above embodiments. The memory is used to store various types of data, which may include, for example, instructions for any application or method in the electronic device, as well as application-related data.

[0205] The processor can be implemented as an Application Specific Integrated Circuit (ASIC), Digital Signal Processor (DSP), Programmable Logic Device (PLD), Field Programmable Gate Array (FPGA), controller, microcontroller, microprocessor, or other electronic components, and is used to execute the dynamic optimization method for open-pit mining and beneficiation coordination plan in response to weather disturbances described in the above embodiments.

[0206] Example 4:

[0207] This embodiment proposes a computer-readable storage medium that stores executable instructions. When these instructions are executed, if they are implemented as software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium.

[0208] The computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the dynamic optimization method for open-pit mining and beneficiation coordination plan in response to weather disturbances as described in the various embodiments of this application.

[0209] The aforementioned storage media include: flash memory, hard disks, multimedia cards, card-type memory (e.g., SD (Secure Digital Memory Card) or DX (Memory Data Register, MDR) memory), random access memory (RAM), static random-access memory (SRAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), programmable read-only memory (PROM), magnetic storage, disks, optical discs, servers, APP (Application) app stores, and other media capable of storing program verification codes. These media store computer programs, which, when executed by a processor, can implement the various steps of the aforementioned dynamic optimization method for open-pit mining and beneficiation coordination plan in response to weather disturbances.

[0210] Example 5:

[0211] This embodiment proposes a computer program product, including a computer program or instructions, which, when executed by a processor, implements the aforementioned dynamic optimization method for open-pit mining and beneficiation coordination plan in response to weather disturbances.

[0212] Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or part of the technical solution, can be embodied in the form of a computer program product.

[0213] The various embodiments in this application are described in a progressive manner. The same or similar parts between the various embodiments can be referred to each other. Each embodiment focuses on describing the differences from other embodiments.

[0214] The scope of protection of this application is not limited to the embodiments described above. Obviously, those skilled in the art can make various modifications and variations to this disclosure without departing from the scope and spirit of this disclosure. If such modifications and variations fall within the scope of this disclosure and its equivalents, then the intent of this disclosure also includes these modifications and variations.

Claims

1. A dynamic optimization method for open-pit mining and beneficiation coordination plans in response to weather disturbances, characterized in that, This method includes the following steps: For any production cycle to be planned, obtain the weather forecast data for all planned periods within that production cycle, and extract the temporal and meteorological characteristics for each planned period. Construct a two-level programming model based on the coupling of data-driven prediction and optimization decision-making; Based on the time and weather characteristics of all planning periods within the production cycle, the optimal production plan for the production cycle is generated using a two-level programming model that couples data-driven prediction and optimization decision-making.

2. The method for dynamic optimization of open-pit mining and beneficiation coordination plan in response to weather disturbances as described in claim 1, characterized in that, For any production cycle to be planned, the weather forecast data for all planned periods within that production cycle is obtained, and the specific content of the time characteristics and meteorological characteristics for each planned period is extracted as follows: For any planned period within the production cycle, obtain the weather forecast data for that planned period; Extract the temporal characteristics and raw meteorological characteristics of the planned period from the weather forecast data for that period; The time features mentioned therein include: the month and season corresponding to the date; The original meteorological characteristics include: maximum temperature, minimum temperature, weather category, and wind force level; The extracted raw meteorological features are preprocessed to obtain the preprocessed meteorological features; The extracted temporal features and preprocessed meteorological features will be used as the temporal and meteorological features for the planning period.

3. The method for dynamic optimization of open-pit mining and beneficiation coordination plan in response to weather disturbances as described in claim 2, characterized in that, The preprocessing includes: numerically encoding the weather category; and standardizing the maximum temperature, minimum temperature, and wind force level.

4. The method for dynamic optimization of open-pit mining and beneficiation coordination plan in response to weather disturbances as described in claim 3, characterized in that, The two-layer planning model based on the coupling of data-driven prediction and optimization decision-making includes: an upper-layer capacity prediction model and a lower-layer mining, stripping and transportation planning model. The upper-level capacity prediction model is used to predict production data within the planned production cycle based on the time and meteorological characteristics of the production cycle to be planned. The lower-level mining and transportation planning model is used to generate the optimal production plan for the production cycle by taking the production data within the production cycle as constraints and minimizing the weighted sum of the mining and transportation cost deviation and the ore supply deviation.

5. The method for dynamic optimization of open-pit mining and beneficiation coordination plan in response to weather disturbances as described in claim 4, characterized in that, The method for constructing the upper-level capacity prediction model is as follows: Collect historical time characteristics, historical meteorological characteristics, and historical production data corresponding to several historical planning periods; wherein the historical production data refers to the actual daily rock mining volume. The historical meteorological characteristics corresponding to each historical planning period are preprocessed to obtain the preprocessed historical meteorological characteristics. Using historical time features and preprocessed historical meteorological features of the same historical planning period as a training sample, and using historical production data of the same historical planning period as prediction labels, a training sample set is constructed. Based on the random forest framework, a quantile regression forest model with multiple decision trees is constructed. With the goal of minimizing quantile loss, the quantile regression forest model is trained using the training sample set to obtain the trained quantile regression forest model. The trained quantile regression forest model is used as the upper-level productivity prediction model.

6. The method for dynamic optimization of open-pit mining and beneficiation coordination plan in response to weather disturbances as described in claim 5, characterized in that, The method described above involves constructing a quantile regression forest model based on a random forest framework, integrating multiple decision trees, and aiming to minimize the quantile loss. The model is trained using a training sample set, and the specific content of the trained quantile regression forest model is as follows: Initialize the empty quantile regression forest model; The Bootstrap sampling method is used to sample the training sample set to generate several training sub-samples; Using each training subsample, independently train one decision tree in the quantile regression forest model; For any decision tree to be trained, determine the target quantile of the current decision tree, and use the training subsamples corresponding to the current decision tree as the root node data. Starting from the root node, recursively execute the node splitting process; the specific content of each node splitting process is as follows: For any node to be split in the current decision tree, determine whether the node satisfies the preset stopping split condition. If it does, the node to be split is taken as a leaf node, and the trained current decision tree is obtained. If it does not, the optimal split point of the node to be split is obtained by minimizing the quantile loss as the node splitting criterion, and the next level splitting node is generated based on the optimal split point.

7. The method for dynamic optimization of open-pit mining and beneficiation coordination plan in response to weather disturbances as described in claim 4, characterized in that, The objective function of the lower-level mining and transportation planning model is to minimize the weighted sum of the mining and transportation cost deviation and the ore supply deviation.

8. The method for dynamic optimization of open-pit mining and beneficiation coordination plan in response to weather disturbances as described in claim 4, characterized in that, The constraints of the lower-level mining and transportation planning model include: Mining capacity constraints of mining equipment: planning period mining area The amount of rock extracted shall not exceed the planned period. mining area The maximum production capacity limit of the equipment; Mining capacity constraints: planning period The total rock extraction volume of all mining areas within the planned period is within the planned period. Within the feasible range of production capacity; Stope reserves constraints: Stope The total rock extraction volume during all planned periods within the unplanned production cycle shall not exceed the mining area. The upper limit of rock quantity; Stripping ratio constraint: for any planning period mining area The total amount of waste rock transported to the spoil heap is equal to the amount transported from the quarry. Total amount of ore mined and mining sites The product of the stripping ratio; wherein the stripping ratio from the mining area The total amount of ore extracted from the mine is: The amount of ore and from the mining site The sum of the ore shipped to storage; Waste disposal capacity constraint: All waste transported to the waste disposal site The total amount of waste rock does not exceed the amount of waste rock dump. Maximum capacity; Material balance constraints: for any planning period mining area The amount of rock extracted is equal to the amount extracted from the quarry. ore quantity, mining area The amount of ore shipped to storage and all ore from the mine The total amount of waste rock transported to the spoil heap; Mineral processing plant grade constraints: for any metal During any planning period Within the total amount of ore transported from all mines to the concentrator, metals... The content is located in the ore dressing plant for metal The required feasible range for metal content; Inventory balance constraint: planning period mining area The inventory level is equal to that of the previous planning period. mining area Inventory levels, plus the planned period mining area The amount of ore shipped to storage, minus the planned period. The ore transported from the mine to the concentrator comes from the stockpile. ore quantity; Maximum inventory constraint: planning period The sum of the inventory levels in all mining areas shall not exceed the total inventory level of the mining area. The upper limit of ore storage capacity; Inventory reservation constraints: planning period mining area The inventory level is not less than the inventory level for the purchasing site. The lower limit of ore reserves; Positive Deviation Constraint on Ore Quantity: Planning Period The positive deviation between the internal ore supply and the demand of the ore processing plant is equal to the planned period. The amount of ore transported from all mining sites to the storage area; Negative deviation constraint on ore quantity: This will affect the planning period. The difference between the ore demand of the internal concentrator and the total supply of all mines is compared with zero, and the larger value is taken as the planning period. The negative deviation between the internal ore supply and the demand of the ore processing plant; Cost deviation constraints: planning period The total planned cost within the planning period equals the total planned cost during the planning period. Total cost within the planning period The positive deviation between total cost and planned cost, minus the planned period cost. The negative deviation between total internal cost and planned cost; Direct ore transport constraint: for any planning period All from the mining area The amount of ore transported to the concentrator and the amount of ore transported from the inventory to the concentrator shall not exceed the maximum daily ore processing capacity of the concentrator. Waste dump utilization equilibrium constraint: For any waste dump spoil heap The total amount of waste rock during all planning periods is located in the spoil heap. Within the acceptable range of waste rock volume allowed by the balanced utilization rate; Inbound and outbound mutual exclusion constraints: in any planning period Inside, mining area The amount of ore shipped to storage and the amount of ore shipped from storage to the concentrator originating from the mine. The amount of ore cannot be positive at the same time; Non-negativity constraint: Planning period Positive and negative deviations between internal ore supply and the demand of the concentrator, and the planning period. Positive and negative deviations between total internal cost and planned cost, planning period mining area Rock mining volume, planned period mining area ore quantity, planning period mining area The amount of ore shipped to storage and the planned period The ore transported from the mine to the concentrator comes from the stockpile. The ore quantities are all non-negative.

9. The method for dynamic optimization of open-pit mining and beneficiation coordination plan in response to weather disturbances as described in claim 1, characterized in that, The optimal production plan includes the mining volume, transportation route, inventory inflow and outflow status, and planned cost for each mining site.

10. A dynamic optimization system for open-pit mining and beneficiation coordination planning in response to weather disturbances, used to implement the dynamic optimization method for open-pit mining and beneficiation coordination planning in response to weather disturbances as described in any one of claims 1-9, characterized in that, The system includes: The data acquisition module is used to acquire weather forecast data for all planned periods within any production cycle; The feature extraction module is used to extract the temporal and meteorological features for each planning period based on the weather forecast data for all planning periods. The model building module is used to build a two-level programming model based on the coupling of data-driven prediction and optimization decision-making. The plan generation module is used to generate the optimal production plan for the production cycle based on the time and meteorological characteristics of all planning periods within the production cycle, using a two-level planning model that couples data-driven prediction and optimization decision-making.

Images