Multi-equipment cooperative mining operation path planning method and system

Abstract

The invention provides a multi-device collaborative mining operation path planning method and system, and relates to the technical field of mining operation, and the method comprises the steps: firstly constructing a mining operation scene portrait containing multi-dimensional features, forming an association network, and generating a multi-device collaborative decision matrix based on the association network; an initial collaborative path scheme is generated according to the multi-device collaborative decision matrix and task target features; establishing a mining operation working condition evolution model, and inputting real-time operation working condition data to obtain a working condition evolution trend prediction result; pre-adapting and adjusting the initial collaborative path scheme in combination with a working condition evolution trend prediction result and a multi-device collaborative decision matrix to obtain a final scheme; and finally, generating a multi-device cooperative execution network and sending the network to the device control terminal to drive operation. According to the invention, efficient and accurate collaborative operation of multiple devices in a complex mining scene is realized, and the mining operation efficiency and safety are improved.

Description

Technical Field

[0001] This invention relates to the field of mining operation technology, and more specifically, to a method and system for multi-equipment collaborative mining operation path planning. Background Technology

[0002] In the field of mining operations, with the continuous expansion of mining scale and the increasing complexity of operations, multi-equipment collaborative operations have become crucial for improving mining efficiency and ensuring operational safety. However, existing mining operation path planning methods have many limitations.

[0003] On the one hand, traditional methods have a rather one-sided understanding of mining operation scenarios, often focusing only on the geological features of the work area while ignoring multi-dimensional information such as equipment coordination history, real-time environmental interference characteristics, and task objective characteristics. This makes it difficult to fully and accurately grasp the operating conditions when facing complex and ever-changing mining scenarios, resulting in a lack of targeted and adaptable path planning.

[0004] On the other hand, existing path planning methods lack scientific decision-making basis and dynamic adjustment mechanism when generating multi-device collaborative paths. They are usually based on fixed rules and experience, and cannot optimize the path plan in a timely manner according to the real-time changes in the working conditions. This can easily lead to problems such as equipment conflicts and poor coordination during the operation, affecting the efficiency and safety of mining operations. Summary of the Invention

[0005] In view of the aforementioned problems, and in conjunction with the first aspect of the present invention, embodiments of the present invention provide a method for multi-device collaborative mining operation path planning, the method comprising: A mining operation scenario profile is constructed, which includes the geological features of the operation area, the historical features of equipment collaboration, the real-time environmental interference features, and the features of the task objectives. The geological features of the operation area, the historical features of equipment collaboration, the real-time environmental interference features, and the features of the task objectives are used to form a scenario feature association network through multi-source data association modeling. A multi-device collaborative decision matrix is ​​generated based on the mining operation scenario profile. The multi-device collaborative decision matrix includes spatial collaborative decision rules, temporal collaborative decision rules, and load collaborative decision rules. Each rule is used to define the basis for the collaborative path selection of equipment under different combinations of scenario features. Based on the multi-device collaborative decision matrix and task objective characteristics, an initial multi-device collaborative path scheme is generated. The initial multi-device collaborative path scheme includes the path node sequence of each device, collaborative interaction nodes between devices, and collaborative operation specifications. A mining operation condition evolution model is established. Real-time collected operation condition data is input into the mining operation condition evolution model to obtain the operation condition evolution trend prediction results. The operation condition evolution trend prediction results include the geological state change trend, equipment operation status change trend, and environmental disturbance change trend within a future preset time period. By combining the prediction results of the operating condition evolution trend with the multi-device collaborative decision matrix, the initial collaborative path scheme of the multi-device is pre-adapted and adjusted to obtain the final collaborative path scheme of the multi-device. A multi-device collaborative execution network is generated based on the final scheme of the multi-device collaborative path. The multi-device collaborative execution network marks the path node timing of each device, the operation instructions of the collaborative interaction nodes, and the contingency plan for changes in working conditions. The multi-device collaborative execution network is sent to the control terminal of each mining device to drive the device to perform mining operations according to the collaborative execution network.

[0006] Furthermore, embodiments of the present invention also provide a multi-device collaborative mining operation path planning system, characterized in that it includes: A processor; a machine-readable storage medium for storing machine-executable instructions of the processor; wherein the processor is configured to execute the above-described multi-device collaborative mining operation path planning method by executing the machine-executable instructions.

[0007] Based on the above, a mining operation scenario profile is constructed, encompassing geological features of the work area, historical features of equipment collaboration, real-time environmental interference features, and task objective features. A scenario feature association network is then formed. Based on this scenario profile, a multi-equipment collaborative decision matrix is ​​generated, covering spatial, temporal, and load-based collaborative decision rules. Then, based on the multi-equipment collaborative decision matrix and task objective features, an initial multi-equipment collaborative path scheme is generated, clarifying the path node sequence of each device, the collaborative interaction nodes between devices, and the collaborative operation specifications. The established mining operation condition evolution model can process real-time collected operation condition data, predict the changing trends of geological conditions, equipment operating conditions, and environmental interference within a preset time period. Combining the operation condition evolution trend prediction results with the multi-equipment collaborative decision matrix, the initial path scheme is pre-adapted and adjusted. The resulting final multi-equipment collaborative path scheme better adapts to changes in actual operation conditions, improving the stability and efficiency of equipment collaborative operations. Finally, the generated multi-equipment collaborative execution network marks the path node sequence, collaborative interaction node operation instructions, and contingency plans for changes in operation conditions for each device. This network is then sent to the control terminals of each mining equipment to drive the equipment to execute operations, achieving efficient and precise multi-equipment collaborative operations and significantly improving the safety of mining operations. Attached Figure Description

[0008] Figure 1This is a schematic diagram of the execution flow of the multi-device collaborative mining operation path planning method provided in the embodiments of the present invention.

[0009] Figure 2 This is a schematic diagram of exemplary hardware and software components of the multi-device collaborative mining operation path planning system provided in an embodiment of the present invention. Detailed Implementation

[0010] The present invention will now be described in detail with reference to the accompanying drawings. Figure 1 This is a flowchart illustrating a multi-device collaborative mining operation path planning method provided in one embodiment of the present invention. The following is a detailed description of this multi-device collaborative mining operation path planning method.

[0011] Step S110: Construct a mining operation scenario profile, which includes geological features of the operation area, historical features of equipment collaboration, real-time environmental interference features, and task target features. The geological features of the operation area, historical features of equipment collaboration, real-time environmental interference features, and task target features are used to form a scenario feature association network through multi-source data association modeling.

[0012] This embodiment uses a multi-equipment collaborative mining scenario in a plateau open-pit coal mine as its application background, involving various large mining equipment such as electric shovels, dump trucks, bulldozers, and graders. The goal of constructing a mining operation scenario profile is to integrate the open-pit coal mine's geological structure information, equipment operation records, environmental perception data, and production plan instructions.

[0013] Step S111: Collect multi-source basic data of the mining operation area. The multi-source basic data includes geological exploration data of the operation area, historical operation data of each mining equipment, historical coordination conflict data between equipment, real-time environmental monitoring data, and current mining task data.

[0014] In this open-pit coal mine scenario, geological exploration data was obtained through a combination of geophysical, geochemical, and drilling methods, covering information such as coal-bearing strata distribution, coal seam thickness, roof and floor lithology, and geological structure development. Historical operational data for each mining device was continuously recorded by onboard terminals and sensor systems installed on the equipment. Examples included the digging cycle, bucket capacity utilization, and hoisting motor current of electric shovels; the average speed, fuel consumption, and braking frequency of dump trucks; and the operating time and blade angle adjustment frequency of bulldozers. Historical data on inter-device coordination conflicts was obtained from the coal mine production scheduling center's historical ledgers, recording events such as equipment congestion, waiting, and cross-operation interference that occurred over the past three years, including the specific context of the conflict, the equipment involved, the duration of the impact, and the handling measures. Real-time environmental monitoring data was collected through equipment deployed in the operating area, such as weather stations, dust monitoring stations, and noise sensors, including parameters such as wind speed, wind direction, air temperature, relative humidity, PM2.5 concentration, and PM10 concentration. The current mining task data is issued by the coal mine production management department based on the annual production plan, specifying the mining area (e.g., mining area A, mining area B), planned output, operation period, quality requirements, and the designated equipment models and quantities for this operation. During the data collection process, precise coordinate information involving equipment positioning is blurred using coordinate offset technology; for equipment operation records that may be associated with specific operators, data anonymization technology is used to remove personally identifiable information, ensuring that data collection complies with relevant data security and privacy protection regulations.

[0015] Step S112: Extract features from the geological exploration data to obtain a set of geological features. The set of geological features includes rock strata distribution features, soil density features, mineral occurrence features, and geological fault features. Each geological feature is labeled with its corresponding spatial distribution range.

[0016] Based on the collected geological exploration data, the distribution characteristics of rock strata are extracted through analysis of borehole core descriptions, well logging curve interpretations, and geological profile maps. This identifies the three-dimensional spatial distribution of different lithologies (such as sandstone, shale, mudstone, and coal seams). The spatial distribution range is marked using polygonal regions or three-dimensional grid cells within a mining area's independent coordinate system. For example, a sandstone layer is distributed horizontally within a polygon bounded by coordinate points (Xa, Ya), (Xb, Yb), (Xc, Yc), and (Xd, Yd), and vertically extends from elevation Za to elevation Zb. Soil density characteristics are calculated based on in-situ test data from standard penetration tests and static cone penetration tests, reflecting the soil's resistance to deformation in different areas. Similarly, the spatial distribution range is marked using three-dimensional grid cells, with each grid cell corresponding to a density level or specific value. Mineral occurrence characteristics mainly include the morphology, occurrence, scale, grade, and spatial distribution of the ore body. These are determined through comprehensive analysis of geological mapping, geophysical and geochemical anomalies, and drilling control data. The spatial distribution range is accurate to the boundary line and pinch-out point of the ore body. Geological fault characteristics are determined by analyzing parameters such as the fault's strike, dip, dip angle, displacement, and fracture zone width, combined with its outcrop location in both plan and profile sections, to mark the spatial extension range and influence zone width of the fault.

[0017] Step S113: Extract features from the historical operating data of the equipment to obtain a set of equipment operating features. The set of equipment operating features includes historical moving speed features, historical load change features, historical fault frequency features, and historical operating efficiency features. Each equipment operating feature is labeled with its corresponding time series distribution.

[0018] Taking dump trucks in open-pit coal mines as an example, the historical movement speed characteristics of the equipment are derived from historical operating data, extracting statistics such as average speed, maximum speed, and speed standard deviation for different time periods (e.g., morning shift, afternoon shift, night shift). These statistics are arranged chronologically to form a time series distribution, with the time granularity potentially being hourly, shift-by-shift, or daily. The historical load variation characteristics are constructed based on the loading data recorded by the on-board weighing system for each load, combined with the loading time, to create a sequence showing load changes over time, reflecting load fluctuations at different operational stages. The historical failure frequency characteristics are obtained by statistically analyzing the number of failures, failure types, and failure durations occurring within a past period (e.g., each month, each quarter), calculating the failure frequency per unit time, and forming a failure frequency sequence with time as the horizontal axis. The historical operating efficiency characteristics, for electric shovels, can be the digging volume per unit time, and for dump trucks, it can be the transport volume per unit time. This is calculated as the ratio of work volume to operating time and organized chronologically into a time series distribution to reflect the changing trend of equipment efficiency.

[0019] Step S114: Extract features from historical collaboration conflict data between devices to obtain a collaboration conflict feature set. The collaboration conflict feature set includes the spatial location features of historical conflicts, the time node features of historical conflicts, the equipment combination features involved in historical conflicts, and the features of historical conflict resolution methods. Each collaboration conflict feature is marked with the corresponding conflict severity.

[0020] From historical data on equipment coordination conflicts, the spatial location characteristics of historical conflicts are used to extract the specific coordinates of the conflict location in the mining area coordinate system, and to associate the location with the type of work area (e.g., mining face, transport road, unloading point, meeting point, etc.). The time node characteristics of historical conflicts record the specific year, month, day, hour, and minute of the conflict event, as well as the work shift and stage at that time (e.g., pre-shift inspection, normal operation, post-shift cleanup). The equipment combination characteristics of historical conflicts clearly identify the type, model, and number of all equipment involved in the conflict, such as combinations like "electric shovel D1 with dump trucks Z3 and Z5" or "dump trucks Z2 and Z4 at meeting point 3." The historical conflict resolution characteristics record in detail the specific measures taken to handle the conflict, such as "dispatching dump truck Z3 ​​to pause for 5 minutes to wait for electric shovel D1 to complete its turn" or "directing dump truck Z2 to reverse to the widened section ahead to avoid Z4." The severity of the conflict is comprehensively assessed based on factors such as the impact of the conflict on the work progress, whether it causes equipment damage or personnel injury risk, and the time required to resolve the conflict. It is divided into different levels such as minor, moderate, relatively serious, and serious, and each collaborative conflict characteristic is marked.

[0021] Step S115: Extract features from real-time environmental monitoring data to obtain an environmental interference feature set. The environmental interference feature set includes wind speed change features, dust concentration change features, light intensity change features, and temporary obstacle distribution features in the work area. Each environmental interference feature is labeled with its corresponding real-time change rate.

[0022] For real-time environmental monitoring data, wind speed change characteristics are calculated by continuously collecting wind speed data to determine the change in wind speed per unit time. This is achieved by dividing the difference between the current wind speed and the previous wind speed by the time interval, yielding the real-time rate of change in wind speed, measured in meters per second per minute. Similarly, dust concentration change characteristics are calculated based on PM2.5 or PM10 concentration data collected by a dust concentration monitor, determining the magnitude and direction of concentration change per unit time, resulting in the real-time rate of change in dust concentration, measured in milligrams per cubic meter per minute. Light intensity change characteristics primarily target outdoor working environments. Based on illuminance data collected by a light sensor, the rate of change in light intensity at different times (e.g., after sunrise, before sunset, noon, and on cloudy days) is calculated, measured in lux per hour. The distribution characteristics of temporary obstacles are obtained through video surveillance, lidar, or manual inspection reports installed at the work site. The types of obstacles (such as rolled rocks, spilled materials, malfunctioning equipment, temporary facilities, etc.), approximate size, location coordinates, and time of appearance are recorded. The rate of change can be described by the time span from the appearance of the obstacle to its stable existence or the speed at which the area of ​​influence of the obstacle expands over time.

[0023] Step S116: Extract features from the current mining task data to obtain a task target feature set, which includes the task operation area, the task mineral extraction target, the task operation time limit requirement, and the list of equipment involved in the task.

[0024] From the current mining task data, the task operation area is defined by the boundary coordinates of the designated mining area, forming a closed polygon or polyhedron region, thus clearly defining the spatial scope of this operation. The target mineral output is expressed as a specific ore yield or useful mineral content, such as a planned quantity of raw coal to be mined, or a planned quantity of ore containing a certain percentage of a useful component to be mined. The task operation time limit specifies the total timeframe for completing this mining task, such as the number of working days and the effective working hours per day, and may also include progress requirements for key milestones. The task equipment list details all mining equipment deployed to complete this task, including the type of equipment (e.g., electric shovels, dump trucks, bulldozers), model, unique serial number, and the main tasks or operating area each piece of equipment is responsible for.

[0025] Step S117: Establish a multi-source feature association model. Input the geological feature set, equipment operation feature set, collaborative conflict feature set, environmental interference feature set, and task target feature set into the multi-source feature association model, calculate the association strength between different feature sets, and generate a feature association weight matrix.

[0026] When establishing a multi-source feature association model, the specific feature items in each feature set are first determined. For example, in the geological feature set, "coal seam thickness" and "soil compaction" are considered; in the equipment operation feature set, "electric shovel digging efficiency" and "dump truck speed" are considered. Then, appropriate association analysis algorithms are selected, such as Pearson correlation coefficient, Spearman rank correlation coefficient, and mutual information, to calculate the pairwise correlation strength between feature items in different feature sets. For example, the correlation strength between "soil compaction" and "dump truck speed," the correlation strength between "coal seam thickness" and "electric shovel digging efficiency," and the correlation strength between "wind speed" and "open-pit mine worker visibility" (which indirectly affects equipment coordination) are calculated. All calculated correlation strength values ​​between feature items are organized into a matrix, where the rows and columns correspond to feature items from different feature sets, and the matrix elements represent the correlation strength between corresponding feature items, thus generating a feature association weight matrix.

[0027] Step S118: Based on the feature association weight matrix, perform fusion processing on each feature set, and combine features with association strength higher than a preset threshold into scene feature units, including: Step S1181: Extract the correlation strength values ​​between each feature set from the feature correlation weight matrix and establish a correlation strength comparison table. The correlation strength comparison table includes the feature source set, feature name, associated feature source set, associated feature name, and correlation strength value.

[0028] From the generated feature association weight matrix, iterate through all non-zero or significant association strength values. Record and establish an association strength comparison table according to the format of the feature source set (e.g., "geological feature set", "equipment operation feature set"), the specific feature name in the set (e.g., "soil compaction", "historical moving speed of electric shovel"), the associated feature source set (e.g., "equipment operation feature set", "environmental interference feature set"), the associated feature name (e.g., "historical moving speed of dump truck", "wind speed change feature"), and the specific association strength value.

[0029] Step S1182: Traverse all association strength values ​​in the association strength comparison table, and filter out association combinations whose association strength values ​​are higher than the preset threshold of association strength. Each association combination contains one or more pairs of association features.

[0030] A preset threshold for association strength is set, which can be determined based on domain knowledge and the actual data distribution. Then, each association strength value in the association strength comparison table is traversed, and association combinations with association strength values ​​greater than the preset threshold are filtered out. The above association combinations may be a single feature pair, that is, a feature in one feature set is associated with a feature in another feature set; or they may be a combination of multiple feature pairs, that is, a feature in one feature set is highly associated with multiple features in multiple other feature sets, or multiple features in multiple feature sets are highly associated with each other.

[0031] Step S1183: For each selected association combination, extract the specific content of the association features in the association combination.

[0032] For each selected association combination, specific descriptive information and data content of these features are extracted from the original feature sets to which each associated feature belongs. For example, for the association combination of "soil compaction (from the geological feature set) and dump truck historical movement speed (from the equipment operation feature set)," the specific numerical distribution or grade classification of soil compaction in different regions, as well as the specific content such as the historical movement speed data sequence and statistical characteristics of dump trucks in the corresponding regions or similar soil conditions, are extracted.

[0033] Step S1184: Calculate the comprehensive association strength of each selected association combination. If the association combination contains multiple pairs of association features, take the average value of the association strength values ​​between each association feature as the comprehensive association strength.

[0034] For an association combination containing only one pair of related features, its overall association strength is the numerical value of the association strength between that pair of features. For an association combination containing multiple pairs of related features, such as a combination containing three pairs of related features: feature A and feature B, feature A and feature C, and feature B and feature C, and their association strength values ​​are S1, S2, and S3 respectively, then the overall association strength of the combination is (S1 + S2 + S3) divided by 3, that is, taking the arithmetic mean of the association strength values ​​of all related feature pairs as the overall association strength of the combination.

[0035] Step S1185: Analyze the mining operation scenarios corresponding to each selected association combination to determine the applicable scenarios for the features.

[0036] By combining the actual processes and common scenarios of mining operations, we analyze the specific mining operation scenarios that are most likely to occur or have the greatest impact on each selected correlation combination. For example, the correlation combination of "soil compaction and historical moving speed of dump trucks" is likely applicable to the scenario of "dump trucks traveling on internal transport roads within the mining area"; the correlation combination of "wind speed variation characteristics and electric shovel operating efficiency" is likely applicable to the scenario of "open-pit electric shovels digging in high wind speed environments". Through the above analysis, we clarify the specific operational context applicable to each correlation combination.

[0037] Step S1186: Integrate the associated feature content, comprehensive association strength, and applicable scenarios of the features into a structured data unit, which is defined as a scenario feature unit. Each scenario feature unit is assigned a unique unit identifier.

[0038] The specific content of all the associated features contained in each selected association combination, the calculated comprehensive association strength of the combination, and the applicable scenarios of the features determined by analysis are integrated together according to a certain structured format (such as JSON, XML, or a custom structure) to form an independent data unit, which is defined as a scene feature unit. For ease of management and reference, each scene feature unit is assigned a unique unit identifier, which can be a string, a number sequence, or a code.

[0039] Step S119: Structure and integrate all scene feature units, classify them according to geological dimension, equipment dimension, collaboration dimension, environmental dimension and task dimension to form an initial mining operation scene profile.

[0040] All generated scene feature units are collected and then categorized according to the dimensions to which the core features or main influencing factors involved in each scene feature unit belong. Under the geological dimension, scene feature units are categorized based on geology-related factors such as rock layer distribution, soil density, mineral occurrence, and geological faults; under the equipment dimension, scene feature units are categorized based on equipment operation-related factors such as equipment movement speed, load changes, failure frequency, and operational efficiency; under the collaboration dimension, scene feature units are categorized based on equipment collaboration-related factors such as conflict location, conflict time, equipment combination, and resolution methods; under the environmental dimension, scene feature units are categorized based on environmental interference factors such as wind speed, dust, lighting, and temporary obstacles; and under the task dimension, scene feature units are categorized based on task objectives such as work area, mining volume, work time limit, and equipment list. Through this multi-dimensional structured integration, the scattered scene feature units are organized into a clear and comprehensive initial mining operation scene profile.

[0041] Step S1110: Collect real-time scene feedback data from the mining operation site, compare the real-time scene feedback data with the scene feature units in the initial mining operation scene profile, correct scene feature units whose correlation strength deviates from the actual scene, update the content and applicable scene of the scene feature units, and obtain the final mining operation scene profile.

[0042] Real-time scenario feedback data reflecting the current actual operational situation is continuously collected through various channels, including sensors and monitoring equipment deployed at the mining site, as well as inspection records and work logs from on-site management personnel. For example, actual operating parameters of the equipment are obtained through vehicle-mounted terminals, the latest environmental data is acquired through environmental monitoring stations, and changes in temporary obstacles are assessed through on-site video. This real-time scenario feedback data is then compared item by item with the corresponding scenario feature units in the initial mining operation scenario profile to check whether the associated feature content in the feature unit matches the actual situation and whether the association strength accurately reflects the current degree of association. If a significant deviation is found in the association strength of a scenario feature unit from the actual scenario—for example, the actual impact of soil compaction on dump truck speed is lower than the association strength in the feature unit—the association strength of that scenario feature unit is corrected. Simultaneously, the specific content of the associated features is updated based on real-time data, such as updating the current value of soil compaction and the latest operating efficiency of the equipment, and the applicable scenario description of the scenario feature unit is adjusted according to the actual situation. After checking, correcting, and updating all scenario feature units, a final mining operation scenario profile that accurately reflects the current mining operation situation is formed.

[0043] Step S120: Generate a multi-device collaborative decision matrix based on the mining operation scenario profile. The multi-device collaborative decision matrix includes spatial collaborative decision rules, temporal collaborative decision rules, and load collaborative decision rules. Each rule is used to define the basis for collaborative path selection of equipment under different combinations of scenario features.

[0044] Step S121: Extract spatial dimension-related features from the mining operation scene profile. The spatial dimension-related features include the spatial distribution range in the geological feature set, the distribution features of temporary obstacles in the environmental disturbance feature set, and the spatial location features of historical conflicts in the collaborative conflict feature set.

[0045] From the mining operation scenario profile, features related to spatial location and distribution are specifically extracted. These include: spatial distribution information of all geological features (such as rock strata, soil density, minerals, and faults) in the geological feature set, which defines the spatial location and extent of different geological conditions within the operation area; distribution characteristics of temporary obstacles in the environmental disturbance feature set, clarifying the spatial location, size, and shape of various obstacles in the current operation environment; and spatial location characteristics of historical conflicts in the collaborative conflict feature set, recording the specific locations where past equipment collaborative conflicts occurred. These spatial dimension-related features collectively constitute the spatial information basis for formulating spatial collaborative decision-making rules.

[0046] Step S122: Establish spatial collaborative decision-making rules based on spatial dimension-related features. The spatial collaborative decision-making rules define the path avoidance distance, path detour priority, and spatial interaction area division standards of devices under different combinations of spatial features.

[0047] By comprehensively analyzing the extracted spatial dimension features, corresponding spatial collaborative decision-making rules are formulated for different combinations of spatial features. For example, for the spatial feature combination of "equipment paths passing through geological fault areas and the presence of temporary obstacles nearby," the rule stipulates that the path avoidance distance between equipment should not be less than a certain value to ensure safety. For the situation of "multiple possible detour paths, each passing through areas with different soil density and different numbers of temporary obstacles," the detour priority is set based on factors such as path length, passage difficulty, and number of obstacles, clarifying which path is preferred. For "different functional areas such as mining areas, transportation roads, loading points, unloading points, and maintenance areas," spatial interaction area division standards are formulated, clarifying the boundaries of each area, the types of equipment allowed to enter, the speed limits of equipment within the area, and passing rules, so as to regulate the collaborative behavior of equipment in space.

[0048] Step S123: Extract time-related features from the mining operation scenario profile. The time-related features include the time series distribution in the equipment operation feature set, the historical conflict time node features in the collaborative conflict feature set, and the task operation time limit requirements in the task target feature set.

[0049] Features related to time factors are extracted from the mining operation scenario profile. The time series distribution in the equipment operation feature set, such as the equipment's movement speed sequence, load change sequence, and fault occurrence time sequence at different time periods, reflects the changing patterns of equipment operation status over time. The historical conflict time node features in the collaborative conflict feature set record the specific times of past equipment collaborative conflicts, helping to analyze the time patterns and regularities of conflict occurrence. The task operation time limit requirements in the task objective feature set clarify the start time, end time, and key stage time nodes of the entire mining task, imposing constraints on the time arrangement of equipment operations. These time-related features are key inputs for constructing time-series collaborative decision-making rules.

[0050] Step S124: Establish time-series collaborative decision-making rules based on time-related features. The time-series collaborative decision-making rules define the work start-up sequence, work time interval, and time window of collaborative interaction nodes of the equipment under different combinations of time features.

[0051] Based on the extracted time-related features, time-series collaborative decision-making rules are established for different combinations of time features. For example, considering the time-series distribution of historical equipment failure frequencies, if a certain type of equipment has a high failure frequency during specific time periods each day (such as immediately after startup or after several hours of continuous operation), the rules for setting the work startup sequence can be designed to avoid starting this type of equipment during these high-risk periods, or to conduct more frequent status monitoring after startup. Referring to the characteristics of historical conflict time nodes, if equipment is prone to collaborative conflicts before and after shift changes or during meal times, the work interval rules can be appropriately extended during these periods, or the work plan can be adjusted to avoid peak times. Based on the task's time limit requirements and the equipment's operational efficiency time series, time windows are set for collaborative interaction nodes (such as loading and unloading interactions at loading points, and handover of transport vehicles), clearly defining the time range for each piece of equipment to arrive at the node, begin interaction, complete interaction, and leave the node, ensuring orderly collaboration of equipment in time.

[0052] Step S125: Extract load dimension related features from the mining operation scenario profile. The load dimension related features include historical load change features in the equipment operation feature set, task mineral mining volume targets in the task target feature set, and mineral occurrence features in the geological feature set.

[0053] Features related to equipment load and task load are extracted from the mining operation scenario profile. The historical load variation features in the equipment operation feature set, such as the load rate, maximum load, average load, and load fluctuation under different operating conditions, reflect the equipment's load-bearing capacity and historical load level. The task mineral extraction target in the task target feature set specifies the total operational load requirement. The mineral occurrence features in the geological feature set, such as the grade, thickness, and distribution density of the ore body, affect the difficulty of mining operations and the load output efficiency per unit time.

[0054] Step S126: Establish load coordination decision rules based on load dimension related features. The load coordination decision rules define the load allocation ratio, load adjustment frequency and coordination transfer scheme when the load is overloaded for equipment under different load feature combinations.

[0055] Based on the extracted load dimension-related features, load coordination decision rules are formulated for different combinations of load features. For example, based on the historical load change characteristics of each piece of equipment (such as maximum safe load and average effective load) and the target mineral extraction volume, a reasonable load distribution ratio is determined among different pieces of equipment to ensure a relatively balanced load and avoid some equipment being overloaded while others are underloaded. Considering changes in mineral occurrence characteristics (such as changes in ore grade leading to changes in the amount of metal per unit of excavation) and fluctuations in equipment operating status, a load adjustment frequency is set, such as per shift, per day, or when the load deviation reaches a certain threshold, to readjust the load distribution of the equipment. Simultaneously, a coordinated transfer plan for overload situations is formulated. When a piece of equipment becomes overloaded due to a sudden situation or task change, it is clear how to safely and efficiently transfer part of its load to other coordinated equipment with lower loads or surplus capacity, and the transfer process, priority, and constraints are specified.

[0056] Step S127: Collect spatial collaborative decision-making rules, temporal collaborative decision-making rules, and load collaborative decision-making rules, and assign rule priority to each decision-making rule. The rule priority is determined based on the importance of the task in the task objective feature set and the severity of the conflict in the collaborative conflict feature set.

[0057] The previously established spatial collaborative decision-making rules, temporal collaborative decision-making rules, and load collaborative decision-making rules are systematically collected and organized to form a complete set of decision rules. Then, a rule priority is assigned to each decision rule in the set to resolve potential conflicts between different rules. The determination of rule priorities is mainly based on two aspects: first, the importance of each clearly defined task in the task objective feature set, assigning higher priority to rules that ensure the completion of critical tasks; second, the severity of historical conflicts recorded in the collaborative conflict feature set, assigning higher priority to rules that can prevent or resolve severe conflicts. The importance of tasks and the severity of conflicts can be quantified into numerical values, and the final priority ranking of each rule can be determined through weighted summation or other comprehensive evaluation methods.

[0058] Step S128: Establish a decision rule conflict resolution mechanism. When decision rules of different dimensions contradict each other under the same combination of scenario features, determine the decision rule to be executed first according to the rule priority. If the priorities are the same, select the decision rule with higher correlation to the scenario by combining the feature association weight matrix in the mining operation scenario profile.

[0059] When decision rules from different dimensions such as space, time, and load are applied to the same combination of scenario features, conflicting requirements may arise. For example, a spatial rule might require device A to prioritize a certain path, while a temporal rule might require device B to prioritize that path to meet a time window. To address this, a decision rule conflict resolution mechanism is established. When a conflict occurs, the priorities of the conflicting decision rules are first compared, with the rule with the higher priority receiving priority execution. If the conflicting rules have the same priority, the feature association weight matrix in the mining operation scenario profile is queried. The sum or average of the association strength between each conflicting rule and each feature in the current scenario feature combination is calculated, and the decision rule with the higher overall association strength with the current scenario is selected as the final execution rule to ensure that the decision better meets the actual needs of the current scenario.

[0060] Step S129: Import the spatial collaborative decision-making rules, temporal collaborative decision-making rules, load collaborative decision-making rules, and decision rule conflict resolution mechanism into the matrix generation tool to construct the initial multi-device collaborative decision-making matrix.

[0061] Choose a suitable matrix generation tool that supports custom matrix structures, rule input, and conflict resolution strategy configuration. Organize and code the previously defined spatial collaborative decision-making rules, temporal collaborative decision-making rules, load collaborative decision-making rules, and the established decision rule conflict resolution mechanism according to the format and specifications required by the matrix generation tool. Then, import these rules and mechanisms into the matrix generation tool. The tool automatically constructs an initial multi-device collaborative decision-making matrix based on the preset matrix structure (e.g., rows represent scene feature combinations, columns represent decision dimensions or device types, and elements represent specific rules). This multi-device collaborative decision-making matrix can output corresponding collaborative decision-making rules based on the input scene feature combinations.

[0062] Step S1210: Verify the initial multi-device collaborative decision matrix using historical collaborative conflict data in the mining operation scenario profile, and calculate the matching rate between the decision results of the multi-device collaborative decision matrix and the optimal solution for resolving historical conflicts. If the matching rate is lower than a preset threshold, adjust the content or priority of the decision rules and verify again until the matching rate reaches the preset threshold to obtain the final multi-device collaborative decision matrix.

[0063] Step S12101: Extract historical collaborative conflict data from the mining operation scene profile. The historical collaborative conflict data includes the scene feature combination when the conflict occurred, the conflict type, the scope of the conflict's impact, the actual solution adopted, and the evaluation of the solution effect.

[0064] Detailed data on historical equipment coordination conflict events are extracted from the set of collaborative conflict features in mining operation scenario profiles. For each conflict event, the following are recorded: the combination of scenario features at the time of the conflict, i.e., the combination of features such as geological conditions, environmental conditions, equipment status, and task stage; the conflict type, such as path conflict, waiting conflict, resource contention conflict, and load allocation conflict; the scope of the conflict's impact, including the number of affected equipment, the work area, and the degree of delay to the task schedule; the actual resolution method adopted, i.e., the specific conflict handling measures taken by on-site dispatch or the system at the time; and the evaluation of the resolution effect, which is assessed by experts or dispatchers based on the timeliness, effectiveness, and whether secondary problems were caused, categorized as excellent, good, average, or poor.

[0065] Step S12102: Filter historical collaborative conflict data with excellent resolution effects, determine the resolution methods corresponding to the filtered historical collaborative conflict data with excellent resolution effects as the optimal solutions for historical conflict resolution, and form a set of optimal solutions for historical conflict resolution. Each entry in the set of optimal solutions for historical conflict resolution includes a combination of conflict scenario features, the optimal solution for historical conflict resolution, and resolution effect parameters.

[0066] The extracted historical collaborative conflict data is filtered, retaining only those conflict event records where the resolution effect is evaluated as "excellent". The "actually adopted resolution method" from the filtered records is determined as the optimal historical conflict resolution solution for that type of conflict scenario. A data entry is created for each of these optimal solutions, containing: a combination of conflict scenario characteristics, detailing the various scenario characteristics leading to the conflict; the optimal historical conflict resolution solution, specifically elaborating on the optimal resolution measures and steps; and resolution effect parameters, such as quantifiable indicators like conflict resolution time, reduced delays, and avoided economic losses, forming a dataset of optimal historical conflict resolution solutions.

[0067] Step S12103: Extract the decision rules corresponding to all scene feature combinations from the initial multi-device collaborative decision matrix to form a set of decision rules for the multi-device collaborative decision matrix.

[0068] Iterate through all possible input items of the initial multi-device collaborative decision matrix, i.e., all possible combinations of scene features. For each combination of scene features, extract the corresponding spatial collaborative decision rules, temporal collaborative decision rules, load collaborative decision rules, conflict resolution mechanisms to be adopted when rules conflict, and the final decision result from the matrix. Organize the extracted rules and results according to the combination of scene features to form a set of decision rules for the multi-device collaborative decision matrix.

[0069] Step S12104: Traverse each entry in the dataset of optimal solutions for historical conflict resolution, input the conflict scenario features corresponding to that entry into the initial multi-device collaborative decision matrix, and obtain the decision results output by the initial multi-device collaborative decision matrix.

[0070] Each entry in the dataset of historical optimal conflict resolution solutions is processed one by one. For the currently processed entry, its "conflict scenario feature combination" is used as input and substituted into the initial multi-device collaborative decision matrix. The matrix analyzes and matches the input scenario feature combination according to its internal decision rules and conflict resolution mechanisms, and outputs a decision result for that conflict scenario, which includes the suggested collaborative decision rules and specific conflict resolution solutions.

[0071] Step S12105: Compare the decision result output by the initial multi-device collaborative decision matrix with the historical optimal conflict resolution solution corresponding to the entry, and determine whether the two are consistent. If the decision rule content, rule priority and conflict resolution solution are the same, it is determined to be a match; otherwise, it is determined to be a mismatch.

[0072] The decision result output by the initial multi-device collaborative decision matrix for the current item's conflict scenario characteristics is compared in detail with the corresponding "best historical conflict resolution solution". The comparison includes: whether the decision rules used in the decision result are consistent with the best solution; whether the priority order of the decision rules is the same; and whether the final conflict resolution solution (such as equipment adjustment measures, route change suggestions, schedule modifications, etc.) is completely consistent. Only when all three aspects are completely identical is the decision result considered a match with the best historical conflict resolution solution; if any aspect differs, it is considered a mismatch.

[0073] Step S12106: Calculate the ratio of the number of matching entries in all entries of the historical conflict resolution optimal solution data set to the total number of entries, and obtain the matching rate between the decision result of the multi-device collaborative decision matrix and the historical conflict resolution optimal solution.

[0074] After processing and comparing all entries in the dataset of optimal historical conflict resolution solutions, the number of entries judged as matches is counted. Then, the number of matched entries is divided by the total number of entries in the dataset to obtain a ratio. This ratio is the matching rate between the decision result of the initial multi-device collaborative decision matrix and the optimal historical conflict resolution solution, reflecting the quality of the matrix decision result.

[0075] Step S12107: If the matching rate reaches the preset threshold, the initial multi-device collaborative decision matrix is ​​verified and used as the final multi-device collaborative decision matrix; if the matching rate is lower than the preset threshold, the mismatched entries in the historical conflict resolution optimal solution data set are analyzed to determine the reasons for the mismatch.

[0076] The preset threshold for the matching rate is determined based on the accuracy requirements of the decision matrix. If the calculated matching rate is greater than or equal to the preset threshold, it indicates that the decision results of the initial multi-device collaborative decision matrix are highly consistent with the historical optimal solution, and can effectively guide device collaborative decision-making. This verifies the matrix and can be used as the final multi-device collaborative decision matrix. If the matching rate is lower than the preset threshold, a thorough analysis of all entries judged as mismatched in the historical conflict resolution optimal solution dataset is required to identify the specific reasons for the mismatch. Possible reasons include: incomplete or incorrect decision rules in the matrix, failing to cover or accurately handle the conflict scenario; unreasonable rule priority settings, leading to the selection of suboptimal rules during conflict; and a lack of specificity or errors in the conflict resolution solution.

[0077] Step S12108: Adjust the initial multi-device collaborative decision matrix according to the reasons for mismatch. If the decision rule content is inconsistent, modify the decision rule content of the corresponding scenario feature combination; if the rule priority is unreasonable, adjust the rule priority order; if the conflict solution is missing, supplement the corresponding conflict solution.

[0078] Based on the identified causes of mismatch, the initial multi-device collaborative decision-making matrix is ​​adjusted accordingly. If the mismatch is due to a discrepancy between the decision rules and the optimal solution, the decision rules for the corresponding scenario feature combinations in the matrix are identified, and the specific clauses and content of the rules are modified based on the approaches and measures of historical conflict resolution optimal solutions. If the decision error is caused by unreasonable rule priority settings, the priorities of the relevant rules are reassessed, and the priority order is adjusted to ensure that high-priority rules can be correctly applied in similar scenarios. If the matrix lacks conflict solutions for a specific conflict scenario, detailed conflict solutions for that scenario feature combination are supplemented by referring to historical conflict resolution optimal solutions, including specific operational steps and parameters.

[0079] Step S12109: After the adjustment is completed, the modified multi-device collaborative decision matrix is ​​re-verified. The above steps of traversing the data set of historical conflict resolution optimal solutions, comparing the decision results with the historical conflict resolution optimal solutions, and calculating the matching rate are repeated to calculate the new matching rate.

[0080] After adjusting the initial multi-device collaborative decision matrix, the modified matrix is ​​re-verified according to steps S12104 to S12106. This involves iterating through each entry of the historical conflict resolution optimal solution dataset, inputting the conflict scenario features into the modified matrix, obtaining new decision results, comparing them with historical conflict resolution optimal solutions, and calculating the new number of matches and the matching rate.

[0081] Step S121010: If the new matching rate reaches the preset matching rate threshold, the multi-device collaborative decision matrix verification is confirmed to be successful; if it still does not reach the threshold, continue to analyze the reasons for the mismatch and make adjustments until the matching rate reaches the preset matching rate threshold, and obtain the final multi-device collaborative decision matrix.

[0082] Check whether the new matching rate of the adjusted multi-device collaborative decision-making matrix has reached the preset threshold. If it has, the verification is successful, and the modified matrix is ​​the final multi-device collaborative decision-making matrix. If the new matching rate is still lower than the preset threshold, repeat steps S12107 to S12109 to continue analyzing the reasons for the remaining mismatched entries, further adjusting and optimizing the matrix, and then verifying again until the calculated matching rate reaches or exceeds the preset threshold, finally determining the multi-device collaborative decision-making matrix that can accurately guide multi-device collaborative decision-making.

[0083] Step S130: Based on the multi-device collaborative decision matrix and task target characteristics, generate a multi-device initial collaborative path scheme. The multi-device initial collaborative path scheme includes the path node sequence of each device, collaborative interaction nodes between devices, and collaborative operation specifications.

[0084] After obtaining the final multi-equipment collaborative decision matrix, and combining the core task requirements clearly defined in the task objective characteristics, such as the scope of the work area, the target mining volume, and the time limit, the path planning process is initiated. First, the starting point (e.g., initial parking location, maintenance area) and ending point (e.g., work area, unloading point) are determined for each piece of equipment participating in the operation. Then, based on the geological features (e.g., avoiding faults, selecting paths with suitable soil density), environmental features (e.g., avoiding temporary obstacles), and spatial collaborative decision rules (e.g., path avoidance distance, area division) in the mining operation scenario profile, an initial path from the starting point to the ending point is planned for each piece of equipment using path search algorithms (e.g., A* algorithm, Dijkstra's algorithm, or genetic algorithm). This initial path consists of a series of ordered path nodes (coordinate points), forming a sequence of path nodes for the equipment. Simultaneously, according to the equipment's operational flow and task division, key locations where interaction between equipment is required are determined, i.e., inter-equipment collaborative interaction nodes, such as loading points, unloading points, meeting points, and material handover points. At each collaborative interaction node, detailed collaborative operation specifications are formulated based on the timing collaborative decision rules and load collaborative decision rules in the multi-device collaborative decision matrix. These specifications include a list of participating devices, the operation sequence of each device, operational action requirements, signal interaction methods, waiting time limits, and load transfer standards. The path node sequences of all devices, collaborative interaction node information, and collaborative operation specifications are then integrated to form an initial multi-device collaborative path scheme.

[0085] Step S140: Establish a mining operation condition evolution model. Input the real-time collected operation condition data into the mining operation condition evolution model to obtain the operation condition evolution trend prediction result. The operation condition evolution trend prediction result includes the geological state change trend, equipment operation status change trend, and environmental disturbance change trend within a future preset time period.

[0086] Step S141: Determine the core influencing factors of mining operation conditions. The core influencing factors include geological conditions, equipment operating conditions, environmental interference, and operation progress. Each core influencing factor corresponds to multiple specific operating condition parameters.

[0087] A comprehensive analysis of the characteristics and patterns of mining operations identifies the core influencing factors that play a decisive role in the operational status of mining operations. Geological conditions are fundamental, including specific parameters such as rock mass stability, soil moisture content, ore grade variations, and fault activity. These parameters directly affect mining difficulty and operational safety. Equipment operating conditions include specific parameters such as equipment temperature, vibration, pressure, fuel consumption, wear of key components, motor current, and hydraulic system pressure, reflecting the health and performance of the equipment. Environmental disturbances encompass specific parameters such as wind speed, wind direction, air humidity, dust concentration, rainfall (open-pit mines), and underground water inflow (underground mines), significantly impacting the working environment and equipment operation. Operational progress factors include specific parameters such as completed mining volume, remaining mining volume, equipment utilization rate, and process completion rate, reflecting the progress of the task and resource allocation efficiency. Each core influencing factor is quantified through its multiple specific operating parameters.

[0088] Step S142: Collect historical working condition data sequence of the mining operation site. The historical working condition data sequence includes records of the changes in working condition parameters corresponding to each core influencing factor over time within a preset time period. Each record is marked with a corresponding timestamp and scene feature label.

[0089] The data acquisition system collects operational data from the mining site over a relatively long, pre-defined period (such as a year, a quarter, or a month, determined based on data volume and cycle). This data is recorded continuously in chronological order, forming a historical operational data sequence. Each data record includes detailed descriptions of the values ​​or states of all specific operational parameters under each core influencing factor at that specific moment. Simultaneously, each record is labeled with a precise timestamp, including year, month, day, hour, minute, and second, along with a corresponding scene feature label. The scene feature label describes the macro-level operational scenario at that moment, such as "normal mining scenario," "equipment maintenance scenario," "severe weather scenario," "shift handover scenario," or "operation near a fault," facilitating subsequent classification and model training of data under different scenarios.

[0090] Step S143: Preprocess the historical operating condition data sequence, classify the preprocessed historical operating condition data sequence according to the scene feature label, and form a historical operating condition data set under different scenes.

[0091] The collected raw historical operating condition data sequences are preprocessed to improve data quality and applicability. Preprocessing operations typically include data cleaning (removing outliers, filling missing values, and correcting data errors), data standardization or normalization (converting operating condition parameters of different dimensions and orders of magnitude to a uniform numerical range, such as [0,1] or [-1,1]), data smoothing (removing noise interference, such as using moving averages), and feature selection or dimensionality reduction (preserving key information and reducing data redundancy). After preprocessing, based on the scene feature labels labeled for each data record, the historical operating condition data sequences are divided into different subsets, each subset corresponding to a specific scene feature label, thus forming historical operating condition data sets under different scenarios, such as "historical operating condition data set for normal mining scenarios" and "historical operating condition data set for severe weather scenarios," etc., so as to train more accurate prediction models for different scenarios.

[0092] Step S144: For each core influencing factor, select the corresponding time series prediction algorithm, input the classified historical operating condition data set into the corresponding time series prediction algorithm, train the single-factor operating condition prediction model for each core influencing factor, and adjust the algorithm parameters to make the error between the prediction result of the single-factor operating condition prediction model and the historical actual data lower than the preset error threshold.

[0093] For the four core influencing factors—geological conditions, equipment operating conditions, environmental disturbances, and work progress—appropriate time series forecasting algorithms are selected based on the characteristics of their respective operating condition parameter time series (e.g., trend, periodicity, randomness, stationarity) and forecasting requirements. For example, for environmental disturbances with obvious trends and seasonality (e.g., wind speed and rainfall in open-pit mines), ARIMA (Autoregressive Integral Moving Average) or SARIMA (Seasonal ARIMA) models can be selected; for equipment operating conditions with nonlinear and complex dynamic characteristics, deep learning algorithms such as LSTM (Long Short-Term Memory) or GRU (Gated Recurrent Unit) can be selected. The historical operating condition data sets under different scenarios, after classification, are input into the time series forecasting algorithms selected for the corresponding core influencing factors. For each core influencing factor, a portion of its historical operating condition data set (e.g., 70%-80%) is used as the training set to train a single-factor operating condition forecasting model. During training, the hyperparameters of the algorithm (such as the number of hidden layer neurons, learning rate, and number of iterations for LSTM, and the p, d, and q orders for ARIMA) are adjusted, and the model performance is evaluated using a validation set (such as 10%-15% of the data). The error between the prediction results and historical actual data (such as mean squared error (MSE) and mean absolute error (MAE)) is calculated. The parameters are continuously adjusted until the prediction error of the model on the validation set is lower than the preset error threshold. At this point, the model training is complete, and a single-factor working condition prediction model for each core influencing factor under different scenarios is obtained.

[0094] Step S145: Establish a multi-factor coupling model, take the prediction results output by the single-factor working condition prediction model of each core influencing factor as input, analyze the coupling relationship between different core influencing factors, and generate a coupling relationship coefficient matrix.

[0095] Step S1451: Determine the type of potential coupling relationship between the core influencing factors.

[0096] A thorough analysis of the interactions and influence mechanisms among the core influencing factors in the mining operation system is crucial to identifying the types of potential coupling relationships that may exist between them. For example, "decreased rock mass stability" in geological conditions may lead to "increased vibration" and "increased energy consumption" in equipment operation conditions (causal coupling); "high dust concentration" in environmental disturbance factors may interact with "increased motor temperature" in equipment operation conditions (bidirectional coupling); and "increased mining volume" in operation progress factors may exacerbate "rock mass disturbance" in geological conditions (indirect coupling). Common coupling relationship types also include linear coupling, nonlinear coupling, time-varying coupling, strong coupling, and weak coupling, which need to be identified and defined according to the specific characteristics of the mining operation.

[0097] Step S1452: Extract the time series of operating parameters corresponding to each core influencing factor from the historical operating condition data series.

[0098] From the preprocessed historical operating condition data series, time series data of all specific operating condition parameters contained in each core influencing factor are extracted according to the division of core influencing factors. For example, time series of "rock mass displacement" and "soil moisture content" under geological condition factors are extracted separately from historical data; time series of "spindle temperature" and "hydraulic system pressure" under equipment operating condition factors are extracted separately. This ensures that each core influencing factor has its corresponding complete set of operating condition parameter time series.

[0099] Step S1453: Perform correlation analysis on the time series of operating parameters for each pair of core influencing factors, and use the correlation analysis algorithm to calculate the correlation coefficient between the two time series of operating parameters.

[0100] For each pair of core influencing factors (such as geological condition factors and equipment operating condition factors, equipment operating condition factors and environmental disturbance factors, etc., all possible pairwise combinations), representative time series of operating condition parameters for each are selected (or principal component analysis is performed on all time series of operating condition parameters under that factor to obtain a comprehensive series). Then, appropriate correlation analysis algorithms, such as Pearson correlation coefficient, Spearman rank correlation coefficient, mutual information, etc., are used to calculate the correlation coefficient between the two time series of operating condition parameters. The magnitude of the correlation coefficient reflects the strength of the linear or nonlinear correlation between the two core influencing factors.

[0101] Step S1454: Determine the coupling strength level based on the absolute value of the correlation coefficient, and record the coupling relationship type, correlation coefficient, and coupling strength level of each pair of core influencing factors as coupling relationship entries.

[0102] Based on the absolute value of the calculated correlation coefficient, and combined with domain knowledge and practical engineering experience, the coupling strength is classified into levels. For example, a correlation coefficient absolute value of 0.8 or above is considered "extremely strong coupling," 0.6-0.8 is "strong coupling," 0.4-0.6 is "moderate coupling," 0.2-0.4 is "weak coupling," and below 0.2 is "extremely weak coupling" or "no coupling." The coupling relationship type (e.g., causal coupling, bidirectional coupling), the specific value of the calculated correlation coefficient, and the determined coupling strength level for each pair of core influencing factors are recorded together as a coupling relationship entry, clearly describing the coupling characteristics between the pair of factors.

[0103] Step S1455: Collect all coupling relationship entries between core influencing factors and construct a coupling relationship entry list, which includes coupling factor pairs, coupling type, correlation coefficient, and coupling strength level.

[0104] All the core influencing factors identified in the analysis were systematically collected and organized into a list of coupling relationships, arranged in a specific order (e.g., by the importance of the coupling factor pair or by the level of coupling strength). Each entry in the list clearly displays: the coupling factor pair (identifying which two core influencing factors are being coupled), the coupling type (e.g., causal coupling, bidirectional coupling, etc.), the correlation coefficient (the specific numerical value), and the level of coupling strength (e.g., strong coupling, moderate coupling, etc.), facilitating subsequent reference and matrix construction.

[0105] Step S1456: Construct a coupling relationship coefficient matrix based on the list of coupling relationship entries. The rows and columns of the coupling relationship coefficient matrix are all core influencing factors. The elements of the coupling relationship coefficient matrix are the correlation coefficients of the corresponding factor pairs. If two factors are not coupled, the corresponding element of the coupling relationship coefficient matrix is ​​zero.

[0106] A square matrix, the coupling coefficient matrix, is constructed using the core influencing factors as row and column indices. The row and column headings of the matrix are each one of the four core influencing factors: geological condition factors, equipment operating condition factors, environmental interference factors, and work progress factors. For each pair of coupling factors in the coupling relationship entry list, the corresponding row and column intersection position is found in the matrix, and the correlation coefficient of that entry is filled in as the value of the matrix element. If two core influencing factors are determined to be uncoupled based on the coupling relationship entry list (e.g., extremely weak coupling or an absolute correlation coefficient below a certain threshold), the element value at the corresponding row and column intersection position in the matrix is ​​set to zero. In this way, the coupling coefficient matrix visually represents the coupling strength between the core influencing factors in numerical form.

[0107] Step S1457: Normalize the coupling coefficient matrix to adjust the numerical range of the elements of the coupling coefficient matrix to a preset range.

[0108] To ensure uniformity and comparability of elements in the coupling coefficient matrix, facilitating subsequent model calculations and analysis, the matrix needs to be normalized. A suitable normalization method, such as min-max normalization, linearly maps the numerical range of all elements in the matrix to a predefined interval, typically [0, 1]. The normalization formula is: Normalized value = (Original value - Minimum value) / (Maximum value - Minimum value), where the maximum and minimum values ​​refer to the original maximum and minimum values ​​of all elements in the matrix. After normalization, the values ​​of the matrix elements are all between 0 and 1, which is more conducive to the quantitative calculation of multi-factor coupling effects.

[0109] Step S1458: Verify the coupling relationship coefficient matrix using historical operating condition data sequences. Select a portion of historical timestamps, substitute the actual parameter changes of each core influencing factor into the coupling relationship coefficient matrix, calculate the predicted value of coupling impact, compare the predicted value of coupling impact with the actual coupling impact result, and determine whether the error is within the preset range.

[0110] Multiple representative historical time stamps are randomly selected from the historical operating condition data sequence or selected at certain intervals as verification points. For each verification point, the actual operating condition parameter changes (changes relative to the previous moment) of each core influencing factor corresponding to that time stamp are obtained. These actual parameter changes are substituted into the coupling relationship coefficient matrix. Based on the coupling strength represented by the matrix, the predicted coupling impact of each core influencing factor on other core influencing factors is calculated using a pre-defined coupling impact calculation model (such as linear superposition, weighted summation, etc.). Simultaneously, the actual coupling impact result corresponding to that time stamp (i.e., the actual change in other core influencing factors due to the change in that factor) is extracted from the historical data. The error (e.g., root mean square error RMSE) between the predicted coupling impact value and the actual coupling impact result is calculated, and it is determined whether this error is within the pre-defined acceptable error range.

[0111] Step S1459: If the error is within the preset range, the coupling relationship coefficient matrix is ​​confirmed to be completed; if the error exceeds the preset range, the correlation of the factor pair is re-analyzed, the correlation coefficient value is adjusted, and the verification is repeated until the error meets the requirements, and the final coupling relationship coefficient matrix is ​​generated.

[0112] If the calculated error is within the preset range during the verification process, it indicates that the coupling coefficient matrix can accurately reflect the actual coupling relationship between the core influencing factors, and the matrix construction is complete. If the error exceeds the preset range, for the core influencing factor pairs with larger errors, their operating parameter time series data should be re-examined to check for data quality issues or omissions of important coupling mechanisms. Then, a more suitable correlation analysis method or adjustment of the data processing procedure should be used to recalculate the correlation coefficient values ​​between them and update the coupling coefficient matrix. Afterward, the same verification points and methods should be used to perform verification again and calculate the new error. The above analysis, adjustment, and verification steps are repeated until the calculated error meets the preset requirements, and finally, the coupling coefficient matrix that can accurately characterize the coupling relationship between each core influencing factor is determined.

[0113] Step S146: Based on the multi-factor coupling model and the coupling relationship coefficient matrix, integrate the single-factor working condition prediction model of each core influencing factor, and construct the mining operation working condition evolution model. The mining operation working condition evolution model includes an input layer, a coupling analysis layer and an output layer. The input layer receives real-time working condition data, the coupling analysis layer calculates the coupling influence of multiple factors, and the output layer outputs the working condition evolution trend.

[0114] A mining operation condition evolution model is constructed by using single-factor operating condition prediction models of each core influencing factor as basic components, combined with a multi-factor coupling model and a determined coupling coefficient matrix. This mining operation condition evolution model adopts a layered architecture, including an input layer, a coupling analysis layer, and an output layer. The input layer receives real-time collected current mining operation condition data, including the current operating condition parameter values ​​of each core influencing factor, providing initial input to the model. The coupling analysis layer first calls the single-factor operating condition prediction models of each core influencing factor, predicting the preliminary change trend of each factor in the near future based on the current input and historical data. Then, using the coupling coefficient matrix, it analyzes and calculates the mutual coupling effects between different core influencing factors, correcting and adjusting the single-factor prediction results to consider the combined effect of multiple factors. The output layer organizes and formats the results processed by the coupling analysis layer, finally outputting the complete operating condition evolution trend prediction results for the future preset time period.

[0115] Step S147: Collect real-time operating condition data of the mining operation site. The real-time operating condition data includes the operating condition parameters corresponding to each core influencing factor at the current time point. Input the real-time operating condition data into the input layer of the mining operation condition evolution model.

[0116] By deploying various sensors, smart meters, and equipment monitoring systems at the mining site, real-time values ​​of all specific operating parameters corresponding to each core influencing factor are acquired at the current time point (i.e., the prediction start time). For example, geological factors include current rock mass displacement monitoring values ​​and soil moisture content; equipment operating status factors include current equipment temperature, vibration values, and fuel consumption; environmental interference factors include current wind speed and dust concentration; and operational progress factors include current completed mining volume and current equipment utilization rate. After necessary format conversion and preprocessing (such as removing obvious noise and standardizing units) of the above-mentioned real-time acquired operational data, it is input into the model's input layer according to the data format and interface specifications required by the mining operational condition evolution model input layer, serving as the initial conditions for model prediction.

[0117] Step S148: Process the real-time operation data through the coupling analysis layer of the mining operation condition evolution model, and calculate the parameter change trend of each core influencing factor in the future preset time period by combining the change pattern of the historical operation data set and the coupling relationship coefficient matrix.

[0118] After receiving real-time operational data from the input layer, the coupling analysis layer of the mining operation condition evolution model first inputs this data into the corresponding single-factor operational condition prediction models for each core influencing factor. Each single-factor prediction model uses real-time data as initial values ​​and, combined with the changes in historical operational condition data sets (such as trends, periods, and fluctuations) of the core influencing factor learned during training, independently predicts the preliminary parameter change trend of the core influencing factor within a preset time period (such as the next 1 hour, 2 hours, or one shift), obtaining a series of predicted values ​​for future moments. Then, the coupling analysis layer calls the coupling relationship coefficient matrix and, based on the coupling strength and type between each core influencing factor in the matrix, performs multi-factor coupling influence calculations on the preliminary predicted values. For example, the predicted change of the geological state factor is multiplied by the corresponding coupling coefficient and superimposed on the prediction result of the equipment operating status factor to correct the predicted value of the equipment status. Through the above method, comprehensively considering the interaction between all core influencing factors, the detailed parameter change trend of each core influencing factor after coupling correction is finally calculated within the preset time period.

[0119] Step S149: Integrate the parameter change trends of each core influencing factor by timestamp, label the predicted value and prediction confidence of the operating condition parameter corresponding to each timestamp, and generate the prediction result of the operating condition evolution trend. If the prediction confidence of a certain core influencing factor is lower than the preset confidence threshold, then readjust the single-factor operating condition prediction model parameters of that core influencing factor, and re-input real-time data for prediction, until the prediction confidence of all core influencing factors reaches the preset confidence threshold.

[0120] The parameter change trends of each core influencing factor output by the coupling analysis layer over a preset time period are integrated in chronological order (e.g., every 5, 10, or 30 minutes). For each timestamp, the predicted values ​​of all specific operating condition parameters under each core influencing factor are recorded. Simultaneously, when outputting predicted values, the single-factor operating condition prediction model also calculates and provides a prediction confidence index, reflecting the model's reliability or uncertainty regarding the predicted value. This index is typically calculated based on the prediction error distribution during model training or the bootstrap method. The predicted operating condition parameter values ​​and their corresponding prediction confidence scores for each timestamp are labeled together. If the prediction confidence score of a core influencing factor (which could be the average confidence score of all parameters under that factor or the confidence score of key parameters) is lower than a preset confidence threshold, the parameters of the single-factor operating condition prediction model for that core influencing factor need to be adjusted (e.g., by increasing training data, adjusting model structure, or hyperparameters). Then, the same real-time operating condition data is input again into the mining operation condition evolution model for prediction. Repeat this process until the prediction confidence of all core influencing factors reaches or exceeds the preset confidence threshold, at which point the final prediction result of the working condition evolution trend is generated.

[0121] Step S150: Combining the prediction results of the operating condition evolution trend with the multi-device collaborative decision matrix, pre-adapt and adjust the initial multi-device collaborative path scheme to obtain the final multi-device collaborative path scheme.

[0122] Step S151: Extract the predicted values ​​of operating condition parameters corresponding to each time stamp within the future preset time period from the operating condition evolution trend prediction results, and form an operating condition prediction sequence according to the core influencing factors. Each operating condition prediction sequence contains the parameter changes of any core influencing factor at different time stamps.

[0123] From the generated operating condition evolution trend prediction results, extract all predicted operating condition parameters for each time point (e.g., t1, t2, ..., tn) within a preset future time period. Then, classify and organize these predicted values ​​according to the categories of core influencing factors (geological state factors, equipment operating state factors, environmental disturbance factors, and work progress factors). Arrange the predicted values ​​of all operating condition parameters belonging to the same core influencing factor at different time points in chronological order to form the operating condition prediction sequence for that core influencing factor. For example, the operating condition prediction sequence for geological state factors includes the predicted value sequence of parameters such as rock mass stability and soil moisture content at time points t1, t2, ..., tn; the operating condition prediction sequence for environmental disturbance factors includes the predicted value sequence of parameters such as wind speed and dust concentration at each time point. Each operating condition prediction sequence is a multidimensional parameter sequence that changes over time.

[0124] Step S152: Compare the operating condition prediction sequence with the scene feature combination in the multi-device collaborative decision matrix, identify the scene feature combination that matches the predicted value of the operating condition parameter at each time stamp, and determine the collaborative decision rule and rule priority corresponding to each time stamp.

[0125] For each timestamp within a pre-defined future time period, the predicted parameter values ​​from the operating condition prediction sequences of all core influencing factors corresponding to that timestamp are combined to form a comprehensive scenario feature combination describing the expected scenario conditions at that timestamp. Then, this comprehensive scenario feature combination is compared and matched one by one with various possible scenario feature combinations defined in the rows or columns of the multi-device collaborative decision-making matrix. The matching process can be based on similarity calculations to find the scenario feature combination in the matrix that is most similar to the current timestamp's scenario feature combination. Once a match is successful, the spatial collaborative decision-making rules, temporal collaborative decision-making rules, and load collaborative decision-making rules corresponding to that scenario feature combination, along with their respective rule priorities, are extracted from the multi-device collaborative decision-making matrix. In this way, the collaborative decision-making rules to be followed under the expected operating conditions at that time and their priority order are determined for each timestamp.

[0126] Step S153: Extract the path node sequence and collaborative interaction node of each device from the initial collaborative path scheme of multiple devices, and split them according to timestamp to obtain the path segment and collaborative operation plan corresponding to each timestamp.

[0127] From the initial multi-device collaborative path plan, a complete path node sequence for each participating device is extracted. This complete path node sequence consists of a series of ordered path node coordinates. Simultaneously, all inter-device collaborative interaction nodes defined in the initial multi-device collaborative path plan are extracted. Based on the preliminary planning of each device's travel speed and task time arrangement in the initial multi-device collaborative path plan, and combined with the distance between path nodes, the estimated timestamps for each device's arrival at and departure from each path node are calculated. Based on this timestamp information, the entire path node sequence of the device is divided into multiple consecutive path segments, each path segment corresponding to a specific timestamp interval; that is, each timestamp corresponds to a path segment that the device is about to travel or is currently traveling. Similarly, the collaborative operation plans at collaborative interaction nodes are also broken down by timestamp, clarifying which collaborative operations the device needs to perform at which collaborative interaction nodes within each timestamp, forming a collaborative operation plan corresponding to each timestamp.

[0128] Step S154: For each timestamp corresponding to the path segment and collaborative operation plan, determine whether the path segment meets the requirements of the collaborative decision rule by combining the collaborative decision rule matched by the timestamp.

[0129] For each timestamp within a preset future time period, the attributes of the path segment corresponding to each device (such as the spatial location, length, expected travel speed, and geological environment of the area traversed), as well as the content of the collaborative operation plan corresponding to that timestamp (such as participating devices, operation sequence, time allocation, and load requirements), are compared and checked in detail with the collaborative decision rules (including spatial, temporal, and load rules) obtained through matching in step S152. For example, it checks whether the path segment meets the path avoidance distance requirements in the spatial collaborative decision rules and whether it has entered prohibited or restricted areas; it checks whether the timing of the collaborative operation plan conforms to the operation interval and time window requirements in the temporal collaborative decision rules; and it checks whether the expected load of the device on that path segment conforms to the load allocation ratio and overload limit in the load collaborative decision rules.

[0130] Step S155: If the path segment or collaborative operation plan does not meet the requirements of the collaborative decision-making rules, then determine the content that needs to be adjusted.

[0131] If, during the judgment process, it is found that a path segment of a certain device at a certain timestamp does not meet the requirements of the corresponding collaborative decision-making rule, for example, the predicted soil density value of the area traversed by the path segment is too low, and the device should detour according to the rule, but the current path does not detour; or the device startup sequence in the collaborative operation plan violates the timing collaborative decision-making rule, leading to possible conflicts; or the load allocation exceeds the upper limit of the load collaborative decision-making rule, then the specific content of the non-compliance with the rule and the rule clauses violated should be recorded in detail, clearly indicating the specific path nodes, driving speed, and traversed areas that need to be adjusted in the path segment, or the device operation sequence, time arrangement, load allocation, and other content that needs to be adjusted in the collaborative operation plan.

[0132] Step S156: Based on the rule priority and the changing trend of the work condition prediction sequence, formulate a pre-adaptation adjustment plan. The pre-adaptation adjustment plan should ensure that the adjusted path segments and collaborative operation plans conform to the collaborative decision-making rules of the current timestamp and adapt to the work condition evolution trend of subsequent timestamps.

[0133] When developing a pre-adaptation adjustment plan, the priority of collaborative decision-making rules should be considered first, prioritizing rules with higher priority. For example, if spatial and temporal rules conflict in some aspect and the spatial rules have higher priority, the adjustment should prioritize the spatial rules. Simultaneously, the changing trends of the predicted operating conditions sequence over subsequent time points should be analyzed, such as whether wind speed will continue to increase, soil moisture content will rise, or equipment temperature will show an upward trend in the future. The adjustment plan should not only ensure that the path segment and collaborative operation plan at the current time point conform to the collaborative decision-making rules for that time point, but also adapt the adjusted content to the evolving operating conditions at subsequent time points as much as possible, avoiding frequent adjustments due to changes in operating conditions. For example, if future wind speed is predicted to increase, routes with less wind impact can be pre-selected when adjusting the current path, or the equipment spacing can be appropriately increased. The pre-adaptation adjustment plan should specifically list the equipment identifiers to be adjusted, timestamps, path node modification suggestions, and changes to the operation plan.

[0134] Step S157: Modify the path segments and collaborative operation plans with corresponding timestamps in the initial collaborative path scheme of multiple devices according to the pre-adaptation adjustment scheme to generate intermediate collaborative paths.

[0135] Based on the established pre-adaptation adjustment plan, the device path segments at the corresponding timestamps in the initial multi-device collaborative path plan are modified. For example, the coordinates of path nodes are replanned, the start and end points of the path are adjusted, and the expected travel speed is modified. Simultaneously, parts of the collaborative operation plan that do not conform to the rules are modified, such as adjusting the order of participating devices, changing the operation time of collaborative interaction nodes, and modifying the quantity and timing of load transfer. All modifications are integrated into the initial multi-device collaborative path plan to form a preliminary adjusted intermediate collaborative path plan.

[0136] Step S158: Perform full-time working condition adaptation verification on the intermediate collaborative path, and compare the path segments and collaborative operation plans of all timestamps in the intermediate collaborative path with the corresponding working condition prediction sequences and collaborative decision rules.

[0137] The generated intermediate collaborative path scheme is comprehensively validated, covering all timestamps within the preset future time period. For each timestamp, the path segments and collaborative operation plans of all devices corresponding to that timestamp in the intermediate collaborative path are compared one by one and in detail with the predicted operating condition sequence (i.e., the expected operating condition parameters at that moment) and the collaborative decision rules obtained through matching in step S152. It is checked whether the adjusted path segments meet the requirements of the collaborative decision rules in all aspects, whether the collaborative operation plans are adapted to the current and predicted operating conditions, and whether there are any potential conflicts or non-compliance with the rules.

[0138] Step S159: If there are cases that do not meet the requirements of the collaborative decision-making rules, re-analyze the matching relationship between the working condition prediction sequence and the collaborative decision-making rules, adjust the pre-adaptation adjustment scheme, modify the intermediate collaborative path again and verify it, until the content of all timestamps meets the requirements of the collaborative decision-making rules.

[0139] If, during the full-time-cycle operational condition adaptation verification process, it is found that path segments or collaborative operation plans at certain timestamps in the intermediate collaborative path still do not meet the requirements of the collaborative decision-making rules, then it is necessary to re-examine whether the matching relationship between the operational condition prediction sequence at that timestamp and the collaborative decision-making rules is accurate, and whether there are any errors in rule application or matching deviations in the combination of scenario features. Based on the results of the re-analysis, the pre-adaptation adjustment plan is adjusted, and the non-compliant parts of the intermediate collaborative path are modified again. After the modification is completed, the full-time-cycle operational condition adaptation verification is executed again, and the comparison and adjustment process is repeated until all path segments and collaborative operation plans at all timestamps in the intermediate collaborative path plan fully meet the corresponding collaborative decision-making rule requirements and are adapted to the operational condition prediction sequence.

[0140] Step S1510: Optimize the verified intermediate collaborative path to generate the final multi-device collaborative path scheme.

[0141] For example, step S1510 includes: step S1511: extract the complete path node sequence of each device from the verified intermediate collaborative path, calculate the total path length and total path time of each device, and count the differences in path length and time of each device.

[0142] From the validated intermediate collaborative path scheme, a complete path node sequence is extracted for each device, which is an ordered set of coordinates of all path nodes from the starting point to the ending point. Based on the path node coordinates, the distance between adjacent nodes in the path node sequence of each device is calculated using a distance calculation formula between two points (such as Euclidean distance), and then summed to obtain the total path length of that device. Simultaneously, based on the length of each path segment and the expected travel speed of that segment, the time to traverse each path segment is calculated, and then summed to obtain the total path time for the device to travel the entire path. The total path length of all devices is calculated, and the difference between the longest and shortest paths is taken as the path length difference for each device; similarly, the total path time of all devices is calculated, and the difference between the longest and shortest times is taken as the time difference for each device.

[0143] Step S1512: Identify the detour path segments in each device path. The detour path segments refer to path segments that deviate from the straight line and do not require necessary avoidance. By comparing the device's path node sequence with the device's straight path node sequence, determine the start and end nodes of the detour path segments.

[0144] For each device, a theoretical straight-line path node sequence is generated based on its start and end coordinates (i.e., a sequence containing only the start and end nodes, or an approximate straight-line path formed by uniformly inserting a small number of intermediate nodes between the start and end for simplified calculation). The actual path node sequence of the device is compared with the straight-line path node sequence, and the deviation of each path segment (path segment between adjacent nodes) in the actual path from the corresponding line segment of the straight path is calculated. If a path segment deviates significantly from the straight path, and according to the current working condition prediction sequence and collaborative decision-making rules, the deviation of the path segment is not due to necessary avoidance requirements (such as avoiding temporary obstacles or adverse geological areas), then the path segment is identified as a detour path segment. The start and end node numbers of the detour path segment are recorded to clarify its position in the path node sequence.

[0145] Step S1513: Optimize the identified detour path segments, replan the path between the starting and ending nodes of the detour path segments, and the replanned path must meet the spatial collaborative decision-making rules of the multi-device collaborative decision-making matrix, and the length of the replanned path must be shorter than the original detour path segment.

[0146] For each identified detour path segment, the starting and ending nodes of that segment are used as new start and end points. A path optimization algorithm (such as a more efficient A* variant or simulated annealing algorithm) is then used to replan the path. During the replanning process, the spatial collaborative decision-making rules in the multi-device collaborative decision matrix must be strictly followed, such as requirements for path avoidance distance, area division, and obstacle avoidance. The goal is to find a new path that, while satisfying all rule constraints, is shorter than the original detour path segment, thereby reducing unnecessary travel distance and time consumption.

[0147] Step S1514: Extract all collaborative interaction nodes in the intermediate collaborative path and analyze the operation process of each collaborative interaction node. The operation process analysis includes the number of devices participating in the collaboration, the operation order of each device, and the allocation of operation time.

[0148] Extract all defined inter-device collaborative interaction nodes from the intermediate collaborative path scheme, such as loading points, unloading points, and meeting points. For each collaborative interaction node, analyze its operation process in detail: determine the type and number of all devices participating in the collaborative interaction, i.e., the number of devices participating in the collaboration; clarify the operation order of each device at this node, such as which device enters first and performs what operation, which device enters last and performs what operation; calculate the operation time allocation of each device at this node, including waiting time, operation execution time, and departure preparation time, and analyze the rationality and efficiency of the operation process.

[0149] Step S1515: Identify redundant steps in the operation process, delete redundant steps in the operation process, adjust the operation sequence and time allocation of each device, so that the total time of collaborative operation is shortened to the preset range.

[0150] Based on the analysis of the collaborative interaction node operation process, a thorough examination is conducted to identify any redundant operational steps, such as unnecessary waiting periods, repetitive signal confirmations, and unreasonable equipment entry / exit sequences. These redundant steps are removed to simplify the operation process. Simultaneously, based on equipment operating efficiency, task priority, and timing-based collaborative decision rules, the operational sequence of each device at the collaborative interaction node is readjusted, and the allocation of operation time is optimized, such as reducing the waiting time of high-priority devices and evenly distributing the operation time across all devices. The goal is to reduce the total collaborative operation time at the collaborative interaction node (the total time from the arrival of the first device to the departure of the last device) to a preset time range, thereby improving overall operational efficiency.

[0151] Step S1516: Calculate the workload of each device in the intermediate collaborative path. The workload is determined based on the load distribution ratio of the device, the length of the operation time and the operation intensity parameters, and statistically analyze the differences in workload of each device.

[0152] Based on the load allocation ratio of each device determined by the load coordination decision rules in the multi-device collaborative decision matrix, combined with the operating time length of each device in the intermediate collaborative path (i.e., the total path time or effective operating time), and the operation intensity parameters reflecting the difficulty of the operation (such as the difficulty of mineral occurrence in geological features, the resistance that the equipment needs to overcome, etc.), the operating load of each device is calculated using a preset operating load calculation formula (e.g., load = load allocation ratio × operating time length × operation intensity parameter). Operating load is a comprehensive indicator reflecting the degree of workload of the equipment. The operating load values ​​of all devices are statistically analyzed, and the difference between the device with the highest load and the device with the lowest load is calculated as the operating load difference among the devices.

[0153] Step S1517: Adjust the load of equipment whose workload difference exceeds the preset difference threshold. Transfer part of the load of the equipment with excessive workload to the equipment whose workload is in the preset normal load range and has not reached the load limit. The adjustment must comply with the load coordination decision rules of the multi-device coordination decision matrix, and the load of the transferred equipment shall not exceed the load limit of the equipment.

[0154] Set a preset threshold for workload difference. If the statistically obtained workload difference among devices exceeds this threshold, it indicates an uneven load distribution among the devices, requiring adjustment. Identify devices with excessively high workloads and devices with low workloads within a preset normal load range (e.g., load rate between 60% and 85%) whose current load has not reached their own load limit. According to the load coordination decision rules in the multi-device collaborative decision matrix, reasonably transfer part of the load borne by the overloaded device to the selected device with a lower load. During the transfer process, ensure that the transferred load does not cause the receiving device's load to exceed its load limit, while maintaining the overall task load balance and schedule requirements.

[0155] Step S1518: Integrate the optimized path node sequence, collaborative interaction node operation process and equipment load allocation to form an optimized collaborative path scheme. Evaluate the performance of the optimized collaborative path scheme and calculate the total operation efficiency, equipment utilization rate and collaborative conflict risk of the optimized collaborative path scheme.

[0156] The optimized sequence of equipment path nodes after detour path optimization, the optimized operation process at collaborative interaction nodes, and the adjusted equipment load distribution are comprehensively integrated to form a complete optimized collaborative path scheme. Then, the performance of this optimized scheme is evaluated: overall operational efficiency is measured by calculating the expected total output or total workload completed per unit time; equipment utilization is assessed by calculating the ratio of effective equipment operating time to total available time; collaborative conflict risk is predicted by analyzing factors such as the number of path intersections, equipment waiting time, and load balancing, combined with historical collaborative conflict data, to predict the likelihood and severity of collaborative conflicts during the scheme's implementation.

[0157] Step S1519: If all performance evaluation indicators meet the preset standards, the optimized collaborative path scheme is determined as the final multi-device collaborative path scheme; if the preset standards are not met, the detour path optimization, operation process optimization or load adjustment are carried out again, and the evaluation is carried out again until all performance evaluation indicators meet the preset standards.

[0158] Set preset standard values ​​or acceptable ranges for overall operation efficiency, equipment utilization, and collaborative conflict risk. If all performance evaluation indicators of the optimized collaborative path scheme meet or exceed the preset standards, the scheme meets the requirements and is determined as the final multi-device collaborative path scheme. If any indicator fails to meet the preset standard, it is necessary to return to steps S1512 to S1517 to re-optimize the detour path, optimize the collaborative interaction node operation process, or adjust the equipment load. After the adjustment is completed, an optimized scheme is formed again and its performance is evaluated. Repeat the above optimization and evaluation process until all performance evaluation indicators meet the preset standards, and finally determine the final multi-device collaborative path scheme.

[0159] Step S160: Generate a multi-device collaborative execution network based on the final scheme of the multi-device collaborative path. The multi-device collaborative execution network marks the path node timing of each device, the operation instructions of the collaborative interaction nodes, and the contingency plan for changes in working conditions. Send the multi-device collaborative execution network to the control terminal of each mining device to drive the device to perform mining operations according to the collaborative execution network.

[0160] Step S161: Extract the path node sequence and corresponding timestamp of each device from the final multi-device collaborative path scheme, classify and organize them according to device identifier, and form a path node time sequence table for each device. The path node time sequence table includes node number, node coordinates, arrival time, departure time and operation at the node.

[0161] From the final multi-device collaborative path scheme, the optimized complete path node sequence is extracted for each device, along with the estimated arrival and departure timestamps for each path node. This information is then categorized and organized according to the device's unique identifier (e.g., device number), creating a path node time sequence table for each device. This table details: node number (numbered sequentially from start to finish); the three-dimensional coordinates (X, Y, Z) of each node in the mining area coordinate system; the estimated arrival time of the device at that node (accurate to minutes or seconds); the estimated departure time of the device from that node; and the operational tasks the device needs to perform at that node, such as "stop and wait," "start loading," "complete unloading," "turn," "accelerate," and "decelerate."

[0162] Step S162: Extract the collaborative interaction nodes in the final solution of the multi-device collaborative path, and determine the device identifier, collaborative operation content, collaborative start time, collaborative end time and collaborative operation standard of each collaborative interaction node.

[0163] Extract detailed information of all collaborative interaction nodes from the final multi-device collaborative path scheme. For each collaborative interaction node, clearly define the device identifiers (numbers) of all devices participating in the collaborative interaction operation; define the collaborative operation content, such as "electric shovel A loads ore for dump trucks B and C" and "dump trucks D and E alternate passing at the meeting point"; specify the collaborative start time (the time when the first device should arrive) and collaborative end time (the time when the last device completes the operation and leaves); and formulate collaborative operation standards, including operational accuracy requirements (such as loading error range), safety distance requirements, signal interaction protocols, quality acceptance standards, etc., to ensure that the collaborative operation is carried out in a standardized and orderly manner.

[0164] Step S163: Generate a corresponding operation instruction for each collaborative interaction node. The operation instruction includes an instruction number, device identifier, operation content, operation parameters, execution time window, and operation result feedback requirements.

[0165] For each collaborative interaction node, specific operation instructions are generated for each participating device based on the collaborative operation content and standards. Each operation instruction has a unique instruction number, including: device identifier (indicating which device the instruction is sent to); detailed operation content (e.g., "Proceed to loading point P3 and accept loading from electric shovel A", "Travel to meeting point Q2 at speed V1"); operation parameters (e.g., specific values ​​or ranges such as target loading amount, travel speed limit, operating pressure, angle, etc.); execution time window (the time interval between when the instruction must start and complete execution); and operation result feedback requirements, such as the information to be fed back after the operation (e.g., actual loading amount, completion time, whether any abnormalities occurred), feedback method (e.g., automatic upload via vehicle terminal, manual confirmation), and feedback time limit.

[0166] Step S164: Extract the predicted values ​​and confidence levels of the operating condition parameters corresponding to each timestamp from the operating condition evolution trend prediction results. For timestamps where the predicted values ​​of the operating condition parameters may exceed the safe range, formulate a contingency plan for changes in operating conditions. The contingency plan for changes in operating conditions includes early warning triggering conditions, equipment path nodes that need to be adjusted, adjusted operation instructions, and contingency plan execution priority.

[0167] Review the predicted operating condition evolution trends and extract the predicted values ​​and confidence levels of the operating condition parameters for each core influencing factor corresponding to each time stamp. Time stamps where the predicted operating condition parameters are close to or may exceed preset safety ranges (e.g., wind speed exceeding the equipment's safe operating limit, dust concentration exceeding the health threshold, equipment temperature approaching the fault threshold) are identified as potential risk time points. For these potential risk time points, develop contingency plans to address changes in operating conditions. The plans specify: the warning trigger conditions, i.e., the threshold at which the real-time monitored operating condition parameters reach to activate the plan; the equipment path nodes to be adjusted, indicating which equipment and which path nodes need to be changed; the adjusted operating instructions, such as modified path node coordinates, travel speed, operating mode, and collaborative interaction node operations; and the plan execution priority, determining the execution order when multiple plans exist simultaneously.

[0168] Step S165: Select a network generation tool and set the structural framework of the multi-device collaborative execution network. The structural framework of the multi-device collaborative execution network includes a device path layer, a collaborative operation layer, and a contingency plan layer. The device path layer displays the timing sequence of the path nodes of each device, the collaborative operation layer displays the operation instructions of the collaborative interaction nodes, and the contingency plan layer displays contingency plans for changes in operating conditions.

[0169] Choose a suitable network modeling or flowchart generation tool, such as Visio, Lucidchart, or professional industrial control process design software. Use this tool to set up a hierarchical framework for the multi-device collaborative execution network. The device path layer, as the bottom layer, uses a timeline as its basis and displays the time sequence table of each participating device's path nodes in parallel, including path nodes, arrival / departure times, and node operations, intuitively presenting the device's trajectory and schedule. The collaborative operation layer, located in the middle layer, corresponds to the position of collaborative interaction nodes on the timeline, displaying the device identifiers, collaborative operation content, operation instructions, and execution time windows involved in the collaboration, and representing the collaborative relationships between devices through directed lines. The contingency plan layer, as the top layer, marks potential risk timestamps on the timeline, displays the warning trigger conditions, adjustment plans, and execution priorities of the corresponding contingency plan for responding to changes in operating conditions, and connects to the affected parts in the device path layer and collaborative operation layer through connecting lines.

[0170] Step S166: Import the path node timing table of each device into the device path layer of the network generation tool, arrange the nodes in timeline order, and label the node coordinates, arrival time and operation at the node.

[0171] The path node timing table created for each device is imported into the device path layer of the multi-device collaborative execution network according to the format required by the network generation tool. Within the tool, a parallel timeline is assigned to each device. Path nodes are arranged on the timeline in chronological order of arrival time, with each node represented by a graphic symbol (such as a dot or square). Next to the node symbol, the node's coordinates, estimated arrival time, and the operation performed at the node are labeled. Adjacent nodes are connected with lines to form the complete path trajectory line for the device, and the estimated travel speed or time for that path segment can be labeled next to the line segment.

[0172] Step S167: Import the operation instructions of the collaborative interaction node into the collaborative operation layer of the network generation tool, mark the device identifier, collaborative operation content and execution requirements of the participating devices at the corresponding timestamp position, and connect the path nodes of the participating devices through the connection.

[0173] The operation instructions generated for the collaborative interaction nodes are imported into the collaborative operation layer of the multi-device collaborative execution network. On the timeline, locate the collaboration start time and collaboration end time interval for each collaborative interaction node. Within this interval, label the list of participating device identifiers, the textual description of the collaborative operation content, and the execution requirements (such as operation parameters and time windows). Using lines of different colors or styles, connect the path nodes (usually the nodes leading to and leaving the collaborative interaction node) of each participating device in the device path layer to the operation instruction box of that collaborative interaction node in the collaborative operation layer, clearly demonstrating how the device participates in the collaborative operation.

[0174] Step S168: Import the contingency plan for changes in operating conditions into the contingency plan layer of the network generation tool, mark the warning triggering conditions, adjustment plan and contingency plan execution priority at the corresponding warning timestamp position, and associate it with the equipment path layer and collaborative operation layer through special symbols.

[0175] Import the developed contingency plans for changes in operating conditions into the contingency plan layer of the multi-device collaborative execution network. Precisely mark the warning timestamp position corresponding to each plan on the timeline. At this position, use a prominent special symbol (such as a triangle warning sign or exclamation mark icon) to indicate the warning trigger condition (e.g., "wind speed > 15 m / s"), the specific adjustment plan (e.g., "device X1 adjusts its path to the backup route R2", "the operation time window of collaborative interaction node Y is extended by T minutes"), and the plan's execution priority (e.g., represented by numbers 1, 2, 3..., with 1 being the highest). Use dashed lines or special lines with arrows to associate the plan symbols with the path nodes that need adjustment in the device path layer and the operation instructions that need to be changed in the collaborative operation layer, indicating the scope and objects affected by the plan.

[0176] Step S169: The multi-device collaborative execution network is sent to the control terminal of each mining equipment. After receiving the multi-device collaborative execution network, the control terminal of the mining equipment parses the path node timing, operation instructions and working condition change response plan in the multi-device collaborative execution network, drives the equipment to move according to the path node timing in the multi-device collaborative execution network, executes the corresponding operation instructions at the collaborative interaction node, and automatically executes the corresponding working condition change response plan when the working condition change triggers the early warning trigger condition.

[0177] The completed multi-device collaborative execution network data file (such as XML, JSON, or a tool-supported proprietary format) is sent to the vehicle-mounted control terminal or local control system of each participating mining equipment via industrial Ethernet, wireless networks (such as 4G / 5G, Wi-Fi), or dedicated data transmission links within the mining area. Upon receiving the multi-device collaborative execution network data, the equipment control terminal initiates a parsing process to extract path node timing information relevant to its own equipment (determining the travel route and time), operational instructions from collaborative interaction nodes (clarifying when, where, and what operations to perform), and contingency plans for changes in operating conditions (understanding potential risks and countermeasures). Based on the parsed path node timing, the control terminal, through the equipment's drive system (such as the engine, hydraulic system, and steering system), drives the equipment to move according to the planned path nodes and time sequence. When the equipment reaches a collaborative interaction node, the control terminal controls the equipment to execute the corresponding collaborative operation according to the operational instructions. Meanwhile, the equipment control terminal monitors the operating parameters collected by its own sensors in real time. When a certain operating parameter is detected to have reached the warning trigger condition in the operating condition change response plan, the corresponding operating condition change response plan is automatically invoked and executed, and the equipment's travel path or operation instructions are adjusted to adapt to the operating condition change, so as to ensure that the mining operation is carried out safely, efficiently and collaboratively.

[0178] In one exemplary embodiment, a multi-device collaborative mining operation path planning system is provided. This system can be a terminal, a server, etc., and its internal structure diagram can be as follows: Figure 2As shown, this multi-device collaborative mining operation path planning system includes a processor, memory, input / output interface, communication interface, display unit, and input device. The processor, memory, and input / output interface are connected via a system bus, and the communication interface, display unit, and input device are also connected to the system bus via the input / output interface. The processor provides computing and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system and computer programs. The internal memory provides the environment for the operation of the operating system and computer programs in the non-volatile storage media. The input / output interface is used for exchanging information between the processor and external devices. The communication interface is used for wired or wireless communication with external terminals; wireless communication can be achieved through Wi-Fi, mobile cellular networks, near-field communication, or other technologies. When the computer program is executed by the processor, it implements a multi-device collaborative mining operation path planning method. The display unit is used to form a visually visible image and can be a display screen, projection device, or virtual reality imaging device. The display screen can be an LCD screen or an e-ink screen. The input device can be a touch layer covering the display screen, or a button, trackball, or touchpad set on the casing of a multi-device collaborative mining operation path planning system, or an external keyboard, touchpad, or mouse, etc.

[0179] It should be noted that, in order to simplify the description of the present invention and thus help to understand one or more embodiments of the invention, multiple features may sometimes be grouped into one embodiment, drawing or description thereof in the foregoing description of the embodiments of the present invention.

Claims

1. A method for multi-equipment collaborative mining operation path planning, characterized in that, The method includes: A mining operation scenario profile is constructed, which includes the geological features of the operation area, the historical features of equipment collaboration, the real-time environmental interference features, and the features of the task objectives. The geological features of the operation area, the historical features of equipment collaboration, the real-time environmental interference features, and the features of the task objectives are used to form a scenario feature association network through multi-source data association modeling. A multi-device collaborative decision matrix is ​​generated based on the mining operation scenario profile. The multi-device collaborative decision matrix includes spatial collaborative decision rules, temporal collaborative decision rules, and load collaborative decision rules. Each rule is used to define the basis for the collaborative path selection of equipment under different combinations of scenario features. Based on the multi-device collaborative decision matrix and task objective characteristics, an initial multi-device collaborative path scheme is generated. The initial multi-device collaborative path scheme includes the path node sequence of each device, collaborative interaction nodes between devices, and collaborative operation specifications. A mining operation condition evolution model is established. Real-time collected operation condition data is input into the mining operation condition evolution model to obtain the operation condition evolution trend prediction results. The operation condition evolution trend prediction results include the geological state change trend, equipment operation status change trend, and environmental disturbance change trend within a future preset time period. By combining the prediction results of the operating condition evolution trend with the multi-device collaborative decision matrix, the initial collaborative path scheme of the multi-device is pre-adapted and adjusted to obtain the final collaborative path scheme of the multi-device. A multi-device collaborative execution network is generated based on the final scheme of the multi-device collaborative path. The multi-device collaborative execution network marks the path node timing of each device, the operation instructions of the collaborative interaction nodes, and the contingency plan for changes in working conditions. The multi-device collaborative execution network is sent to the control terminal of each mining device to drive the device to perform mining operations according to the collaborative execution network.

2. The multi-equipment collaborative mining operation path planning method according to claim 1, characterized in that, The construction of the mining operation scenario profile includes: Collect multi-source basic data of the mining operation area, including geological exploration data of the operation area, historical operation data of each mining equipment, historical coordination conflict data between equipment, real-time environmental monitoring data, and current mining task data. Geological exploration data is subjected to feature extraction to obtain a set of geological features, which includes rock strata distribution features, soil density features, mineral occurrence features and geological fault features. Each geological feature is labeled with its corresponding spatial distribution range. Feature extraction is performed on historical equipment operation data to obtain a set of equipment operation features. The set of equipment operation features includes historical equipment movement speed features, historical equipment load change features, historical equipment failure frequency features, and historical equipment operating efficiency features. Each equipment operation feature is labeled with its corresponding time series distribution. Feature extraction is performed on historical collaboration conflict data between devices to obtain a collaboration conflict feature set. The collaboration conflict feature set includes the spatial location features of historical conflicts, the time node features of historical conflicts, the equipment combination features involved in historical conflicts, and the features of historical conflict resolution methods. Each collaboration conflict feature is labeled with the corresponding conflict severity. Feature extraction is performed on real-time environmental monitoring data to obtain an environmental disturbance feature set. The environmental disturbance feature set includes wind speed change features, dust concentration change features, light intensity change features, and temporary obstacle distribution features in the work area. Each environmental disturbance feature is labeled with its corresponding real-time change rate. Feature extraction is performed on the current mining task data to obtain a task target feature set, which includes the task operation area, task mineral extraction target, task operation time limit requirement, and a list of equipment involved in the task. A multi-source feature association model is established. Geological feature set, equipment operation feature set, collaborative conflict feature set, environmental interference feature set and task target feature set are input into the multi-source feature association model. The association strength between different feature sets is calculated and a feature association weight matrix is ​​generated. Based on the feature association weight matrix, each feature set is fused, and features with association strength higher than a preset threshold are combined into scene feature units. Each scene feature unit contains associated feature content, association strength and applicable scene of the feature. All scene feature units are structurally integrated and classified according to geological, equipment, collaboration, environmental and task dimensions to form an initial mining operation scene profile; Real-time scene feedback data from the mining operation site is collected, and the real-time scene feedback data is compared with the scene feature units in the initial mining operation scene profile. Scene feature units whose correlation strength deviates from the actual scene are corrected, and the content and applicable scene of the scene feature units are updated to obtain the final mining operation scene profile.

3. The multi-equipment collaborative mining operation path planning method according to claim 1, characterized in that, The generation of a multi-device collaborative decision matrix based on mining operation scenario profiles includes: Spatial dimension-related features are extracted from the mining operation scene profile. These features include the spatial distribution range of the geological feature set, the distribution features of temporary obstacles in the environmental disturbance feature set, and the spatial location features of historical conflicts in the collaborative conflict feature set. Based on spatial dimension-related features, spatial collaborative decision-making rules are established. These rules define the path avoidance distance, path detour priority, and spatial interaction area division criteria for devices under different combinations of spatial features. Extract time-related features from the mining operation scenario profile. These time-related features include the time series distribution in the equipment operation feature set, the historical conflict time node features in the collaborative conflict feature set, and the task operation time limit requirements in the task objective feature set. A time-series collaborative decision-making rule is established based on time-related features. The time-series collaborative decision-making rule defines the work start sequence, work time interval, and time window of collaborative interaction nodes of the equipment under different combinations of time features. The load dimension-related features are extracted from the mining operation scenario profile. These load dimension-related features include historical load change features in the equipment operation feature set, task mineral mining volume targets in the task target feature set, and mineral occurrence features in the geological feature set. Based on load dimension-related features, load coordination decision rules are established. These rules define the load allocation ratio, load adjustment frequency, and coordinated transfer scheme when the load is overloaded for equipment under different combinations of load characteristics. Collect spatial collaborative decision-making rules, temporal collaborative decision-making rules, and load collaborative decision-making rules, and assign rule priorities to each decision-making rule. The rule priorities are determined based on the importance of the task in the task objective feature set and the severity of the conflict in the collaborative conflict feature set. Establish a decision rule conflict resolution mechanism. When decision rules of different dimensions conflict under the same combination of scenario features, determine the priority of the decision rule to be executed based on the rule priority. If the priorities are the same, select the decision rule with higher correlation to the scenario by combining the feature association weight matrix in the mining operation scenario profile. The spatial collaborative decision-making rules, temporal collaborative decision-making rules, load collaborative decision-making rules, and decision rule conflict resolution mechanisms are imported into the matrix generation tool to construct an initial multi-device collaborative decision-making matrix. The initial multi-device collaborative decision matrix is ​​verified by using historical collaborative conflict data in the mining operation scenario profile. The matching rate between the decision results of the multi-device collaborative decision matrix and the optimal solution for resolving historical conflicts is calculated. If the matching rate is lower than a preset threshold, the content or priority of the decision rules is adjusted, and the verification is repeated until the matching rate reaches the preset threshold, thus obtaining the final multi-device collaborative decision matrix.

4. The multi-equipment collaborative mining operation path planning method according to claim 1, characterized in that, The establishment of the mining operation condition evolution model involves inputting real-time collected operation condition data into the model to obtain prediction results of the operation condition evolution trend, including: The core influencing factors of mining operation conditions are determined. These core influencing factors include geological conditions, equipment operating conditions, environmental disturbances, and operation progress. Each core influencing factor corresponds to multiple specific operating condition parameters. Collect historical working condition data sequences from mining operation sites. These historical working condition data sequences contain records of changes in working condition parameters corresponding to each core influencing factor over time within a preset time period. Each record is marked with a corresponding timestamp and scene feature label. The historical operating condition data sequence is preprocessed, and the preprocessed historical operating condition data sequence is classified according to the scene feature label to form a historical operating condition data set under different scenes. For each core influencing factor, a corresponding time series forecasting algorithm is selected. The classified historical operating condition data set is input into the corresponding time series forecasting algorithm to train the single-factor operating condition forecasting model for each core influencing factor. By adjusting the algorithm parameters, the error between the prediction results of the single-factor operating condition forecasting model and the historical actual data is made lower than the preset error threshold. A multi-factor coupling model is established, and the prediction results output by the single-factor working condition prediction model of each core influencing factor are used as input to analyze the coupling relationship between different core influencing factors and generate a coupling relationship coefficient matrix. Based on the multi-factor coupling model and coupling relationship coefficient matrix, a single-factor working condition prediction model is integrated with each core influencing factor to construct a mining operation working condition evolution model. The mining operation working condition evolution model includes an input layer, a coupling analysis layer and an output layer. The input layer receives real-time working condition data, the coupling analysis layer calculates the coupling influence of multiple factors, and the output layer outputs the working condition evolution trend. Real-time operational data of the mining operation site is collected. The real-time operational data includes operational parameters corresponding to each core influencing factor at the current time point. The real-time operational data is then input into the input layer of the mining operation condition evolution model. The real-time operation data is processed by the coupling analysis layer of the mining operation condition evolution model. Combined with the change pattern of historical operation data set and the coupling relationship coefficient matrix, the parameter change trend of each core influencing factor in the future preset time period is calculated. The parameter change trends of each core influencing factor are integrated by timestamp, and the predicted value and prediction confidence of the working condition parameter corresponding to each timestamp are marked to generate the prediction result of the working condition evolution trend. If the prediction confidence of a core influencing factor is lower than the preset confidence threshold, the single-factor working condition prediction model parameters of the core influencing factor are readjusted, and real-time data is input again for prediction until the prediction confidence of all core influencing factors reaches the preset confidence threshold.

5. The multi-equipment collaborative mining operation path planning method according to claim 1, characterized in that, The process of combining the predicted operating condition evolution trend results with the multi-device collaborative decision matrix to pre-adapt and adjust the initial multi-device collaborative path scheme, resulting in the final multi-device collaborative path scheme, includes: The predicted values ​​of operating condition parameters corresponding to each time stamp within a preset time period are extracted from the operating condition evolution trend prediction results. The operating condition prediction sequences are formed by classifying them according to the core influencing factors. Each operating condition prediction sequence contains the parameter changes of any core influencing factor at different time stamps. The predicted operating conditions sequence is compared with the scene feature combination in the multi-device collaborative decision matrix to identify the scene feature combination that matches the predicted operating condition parameter value at each time stamp, and to determine the collaborative decision rule and rule priority corresponding to each time stamp. Extract the path node sequence and collaborative interaction node of each device from the initial collaborative path scheme of multiple devices, and split them according to timestamp to obtain the path segment and collaborative operation plan corresponding to each timestamp. For each timestamp, the path segment and collaborative operation plan are combined with the collaborative decision-making rule matched by that timestamp to determine whether the path segment meets the requirements of the collaborative decision-making rule. If a path segment or collaborative operation plan does not meet the requirements of the collaborative decision-making rules, then the content that needs to be adjusted is determined. Based on the rule priority and the changing trend of the work condition prediction sequence, a pre-adaptation adjustment plan is formulated. The pre-adaptation adjustment plan should ensure that the adjusted path segments and collaborative operation plans conform to the collaborative decision-making rules of the current timestamp and adapt to the work condition evolution trend of subsequent timestamps. According to the pre-adaptation adjustment scheme, the path segments and collaborative operation plans with corresponding timestamps in the initial collaborative path scheme of multiple devices are modified to generate intermediate collaborative paths; The intermediate collaborative path is verified for full-time working condition adaptation. The path segments with all timestamps and collaborative operation plans in the intermediate collaborative path are compared with the corresponding working condition prediction sequences and collaborative decision rules. If there are situations that do not meet the requirements of the collaborative decision-making rules, the matching relationship between the working condition prediction sequence and the collaborative decision-making rules will be re-analyzed, the pre-adaptation adjustment scheme will be adjusted, the intermediate collaborative path will be modified again and verified, until the content of all timestamps meets the requirements of the collaborative decision-making rules. The verified intermediate collaboration paths are optimized to generate the final multi-device collaboration path scheme.

6. The multi-equipment collaborative mining operation path planning method according to claim 2, characterized in that, The process of fusing each feature set based on the feature association weight matrix, and combining features with association strength higher than a preset threshold into scene feature units, includes: The correlation strength values ​​between each feature set are extracted from the feature correlation weight matrix, and a correlation strength comparison table is established. The correlation strength comparison table includes the feature source set, feature name, associated feature source set, associated feature name, and correlation strength value. Iterate through all the association strength values ​​in the association strength comparison table, and filter out the association combinations whose association strength values ​​are higher than the preset threshold of association strength. Each association combination contains one or more pairs of association features. For each selected association combination, extract the specific content of the association features in the association combination; Calculate the overall association strength of each selected association combination. If an association combination contains multiple pairs of association features, take the average value of the association strength values ​​among the association features as the overall association strength. Analyze the mining operation scenarios corresponding to each selected association combination to determine the applicable scenarios for the features; The associated feature content, comprehensive association strength, and applicable scenarios of the features are integrated into a structured data unit, which is defined as a scenario feature unit. Each scenario feature unit is assigned a unique unit identifier.

7. The multi-equipment collaborative mining operation path planning method according to claim 3, characterized in that, The initial multi-device collaborative decision matrix is ​​verified using historical collaborative conflict data from the mining operation scenario profile. The matching rate between the decision results of the multi-device collaborative decision matrix and the optimal solutions for historical conflict resolution is calculated. If the matching rate is lower than a preset threshold, the decision rule content or rule priority is adjusted, and verification is repeated until the matching rate reaches the preset threshold, resulting in the final multi-device collaborative decision matrix, including: Historical collaborative conflict data is extracted from the mining operation scenario profile. The historical collaborative conflict data includes the combination of scenario features when the conflict occurred, the conflict type, the scope of the conflict's impact, the actual solution adopted, and the evaluation of the solution's effectiveness. Historical collaborative conflict data with excellent resolution effects are selected, and the resolution methods corresponding to the selected historical collaborative conflict data with excellent resolution effects are determined as the optimal solutions for historical conflict resolution, forming a set of optimal solutions for historical conflict resolution. Each entry in the set of optimal solutions for historical conflict resolution includes a combination of conflict scenario features, the optimal solution for historical conflict resolution, and resolution effect parameters. The decision rules corresponding to all scene feature combinations are extracted from the initial multi-device collaborative decision matrix to form a set of decision rules for the multi-device collaborative decision matrix; Iterate through each entry in the dataset of optimal solutions for resolving historical conflicts, combine the conflict scenario features corresponding to that entry and input them into the initial multi-device collaborative decision matrix, and obtain the decision results output by the initial multi-device collaborative decision matrix; Compare the decision results output by the initial multi-device collaborative decision matrix with the corresponding historical conflict resolution optimal solution for that entry, and determine whether the two are consistent. If the decision rule content, rule priority, and conflict resolution solution are the same, it is determined to be a match; otherwise, it is determined to be a mismatch. The ratio of the number of matching entries in the historical conflict resolution optimal solution dataset to the total number of entries is used to obtain the matching rate between the decision results of the multi-device collaborative decision matrix and the historical conflict resolution optimal solution. If the matching rate reaches the preset threshold, the initial multi-device collaborative decision matrix is ​​verified and used as the final multi-device collaborative decision matrix; if the matching rate is lower than the preset threshold, the mismatched entries in the historical conflict resolution optimal solution data set are analyzed to determine the reasons for the mismatch. Adjust the initial multi-device collaborative decision matrix to address the reasons for mismatch. If the decision rule content is inconsistent, modify the decision rule content for the corresponding scenario feature combination. If the rule priority is unreasonable, adjust the rule priority order. If the conflict solution is missing, supplement the corresponding conflict solution. After the adjustment is completed, the modified multi-device collaborative decision matrix is ​​re-verified. The above steps are repeated: traversing the data set of historical conflict resolution best solutions, comparing the decision results with the historical conflict resolution best solutions, and calculating the matching rate. The new matching rate is then calculated. If the new matching rate reaches the preset matching rate threshold, the multi-device collaborative decision matrix verification is confirmed to be successful; if it still does not reach the threshold, the reasons for the mismatch are analyzed and adjusted until the matching rate reaches the preset matching rate threshold, and the final multi-device collaborative decision matrix is ​​obtained.

8. The multi-equipment collaborative mining operation path planning method according to claim 4, characterized in that, The establishment of a multi-factor coupling model involves using the prediction results from the single-factor operating condition prediction models of each core influencing factor as input, analyzing the coupling relationship between different core influencing factors, and generating a coupling relationship coefficient matrix, including: Identify the types of potential coupling relationships among the core influencing factors; Extract the time series of operating parameters corresponding to each core influencing factor from historical operating condition data sequences; Correlation analysis was performed on the time series of operating parameters for each pair of core influencing factors, and the correlation coefficient between the two time series of operating parameters was calculated using a correlation analysis algorithm. The coupling strength level is determined based on the absolute value of the correlation coefficient, and the coupling relationship type, correlation coefficient and coupling strength level of each pair of core influencing factors are recorded as coupling relationship entries. Collect all coupling relationship entries among the core influencing factors and construct a coupling relationship entry list, which includes coupling factor pairs, coupling type, correlation coefficient, and coupling strength level; A coupling coefficient matrix is ​​constructed based on the list of coupling relationship items. The rows and columns of the coupling coefficient matrix are the core influencing factors, and the elements of the coupling coefficient matrix are the correlation coefficients of the corresponding factor pairs. If two factors are not coupled, the corresponding element of the coupling coefficient matrix is ​​zero. The coupling coefficient matrix is ​​normalized to adjust the numerical range of its elements to a preset range. The coupling relationship coefficient matrix is ​​verified by using historical operating condition data sequences. A portion of historical timestamps are selected, and the actual parameter changes of each core influencing factor are substituted into the coupling relationship coefficient matrix to calculate the predicted value of coupling impact. The predicted value of coupling impact is compared with the actual coupling impact result to determine whether the error is within the preset range. If the error is within the preset range, the coupling coefficient matrix is ​​considered complete; if the error exceeds the preset range, the correlation of the factor pair is re-analyzed, the correlation coefficient value is adjusted, and the verification is repeated until the error meets the requirements, and the final coupling coefficient matrix is ​​generated.

9. The multi-equipment collaborative mining operation path planning method according to claim 1, characterized in that, The process involves generating a multi-device collaborative execution network based on the final solution for multi-device collaborative paths. This network labels the path node sequence, operational instructions for collaborative interaction nodes, and contingency plans for changes in operating conditions for each device. The multi-device collaborative execution network is then sent to the control terminals of each mining device, driving the devices to execute mining operations according to the network. This includes: The path node sequence and corresponding timestamp of each device are extracted from the final solution of the multi-device collaborative path, and sorted and organized according to the device identifier to form the path node time sequence table of each device. The path node time sequence table includes node number, node coordinates, arrival time, departure time and operation at the node. Extract the collaborative interaction nodes from the final multi-device collaborative path scheme, and determine the device identifier, collaborative operation content, collaborative start time, collaborative end time, and collaborative operation standards for each collaborative interaction node. A corresponding operation instruction is generated for each collaborative interaction node. The operation instruction includes an instruction number, device identifier, operation content, operation parameters, execution time window, and operation result feedback requirements. Extract the predicted values ​​and confidence levels of operating condition parameters corresponding to each timestamp from the operating condition evolution trend prediction results. For timestamps where the predicted values ​​of operating condition parameters may exceed the safe range, formulate contingency plans for operating condition changes. The contingency plans for operating condition changes include early warning triggering conditions, equipment path nodes that need to be adjusted, adjusted operation instructions, and contingency plan execution priorities. Select a network generation tool and set the structural framework of the multi-device collaborative execution network. The structural framework of the multi-device collaborative execution network includes a device path layer, a collaborative operation layer, and a contingency plan layer. The device path layer displays the timing sequence of the path nodes of each device, the collaborative operation layer displays the operation instructions of the collaborative interaction nodes, and the contingency plan layer displays contingency plans for changes in working conditions. Import the path node sequence table of each device into the device path layer of the network generation tool, arrange the nodes in timeline order, and label the node coordinates, arrival time and operation at the node. The operation instructions of the collaborative interaction nodes are imported into the collaborative operation layer of the network generation tool. The device identifiers, collaborative operation contents and execution requirements of the participating devices are marked at the corresponding timestamp positions, and the path nodes of the participating devices are connected by lines. Import the contingency plan for changes in operating conditions into the contingency plan layer of the network generation tool, mark the warning triggering conditions, adjustment plan and contingency plan execution priority at the corresponding warning timestamp position, and associate it with the equipment path layer and collaborative operation layer through special symbols; The block sends the multi-device collaborative execution network to the control terminals of each mining device. After receiving the multi-device collaborative execution network, the control terminal of the mining device parses the path node timing, operation instructions and working condition change response plan in the multi-device collaborative execution network, drives the device to move according to the path node timing in the multi-device collaborative execution network, executes the corresponding operation instructions at the collaborative interaction node, and automatically executes the corresponding working condition change response plan when the working condition change triggers the early warning trigger condition.

10. A multi-device collaborative mining operation path planning system, characterized in that, include: processor; A machine-readable storage medium for storing machine-executable instructions of the processor; The processor is configured to execute the multi-device collaborative mining operation path planning method according to any one of claims 1 to 9 by executing the machine-executable instructions.

Images