Unmanned coal mine production method

Abstract

The invention provides an unmanned coal mine production method, which comprises the following steps that: an unmanned coal mine production system receives a coal mine production order, and calculates the numbers of coal detection robots, coal breaking robots, coal loading robots, coal conveying robots, supporting robots, goaf processing robots and inspection robots according to the order quantity, the order delivery time and the size of a mining area; a multi-objective function with the minimum order completion time, the minimum total adjustment time and the minimum robot total energy consumption as the core is established, constraint conditions of a coal exploration robot, a coal breaking robot, a coal charging robot, a coal conveying robot, a supporting robot, a goaf processing robot and an inspection robot are established, and the optimal operation site and operation time of each robot are calculated. According to the method, an intelligent dispatching tool and an unmanned production method are provided for the coal mine production robot, and the coal mine production efficiency and safety are further improved.

Description

Technical Field

[0001] This invention relates to the field of coal mine production robot scheduling technology, specifically to an unmanned coal mine production method. Background Technology

[0002] Coal mines are an important energy source for human society and a vital industry in the national economy. They are generally divided into underground coal mines, which are far from the surface, and open-pit coal mines, which are very close to the surface. The vast majority of coal mines in my country are underground mines. The five major disasters commonly found in coal mines are coal dust, water, fire, gas, and roof collapse. Roof collapse accidents rank first in both frequency and fatality rate among all types of coal mine accidents. This occurs when the roof strata above the coal seam lose their support, causing the ventilation ducts to collapse; it is also known as a roof fall accident. In recent years, robots have also been used in coal mine operations, greatly improving operational efficiency and safety. CN121291832A discloses a disaster detection system and method for an intrinsically safe variable-structure aerial-ground robot in coal mines; CN118378872A discloses a collaborative management platform for coal mine robot clusters; CN120853282A discloses an inspection robot and its usage method for special underground coal mine environments; and CN121024593A discloses a remotely controlled intelligent tunneling robot for coal mines.

[0003] Existing coal mine production robot scheduling technologies have significant shortcomings in areas such as architecture design, objective optimization, dynamic response, algorithm performance, and reliability assurance. Centralized scheduling architectures suffer from inherent limitations, incomplete multi-objective optimization design, weak dynamic disturbance response capabilities, insufficient algorithm efficiency and robustness, and a lack of reliability and traceability in the scheduling process. Therefore, there is an urgent need to develop an unmanned coal mine production method based on a distributed architecture, considering multi-objective optimization, possessing rapid dynamic response capabilities, and high reliability. This is a pressing requirement for overcoming current technological bottlenecks and promoting high-quality development of the coal mining industry. Summary of the Invention

[0004] The purpose of this invention is to solve the technical problems mentioned above and to propose an unmanned coal mine production method, comprising the following steps: S1. Accept coal mine production orders, calculate the number of work sites and the number of robots based on order quantity, order delivery time, and mine size; establish a multi-objective function with minimizing order completion time, total robot adjustment time, and total robot energy consumption as the core, and establish constraints for the robots; S2. The unmanned coal mine production system encodes the work location, operation, and time of each robot into a feasible solution; the robots operate in a distributed manner using artificial intelligence algorithms to jointly solve the optimal solution of the multi-objective function that meets the constraints in step S1, that is, the optimal work location, operation, and time of each robot. S3. The robot goes to the optimal work location according to the optimal solution obtained in step S2 and carries out coal mine production according to the work operation and work time. When changes and disturbances occur in the coal mine production process, it returns to steps S1 and S2 to adjust the work location, work operation and work time of each robot in real time until all coal mine production orders are completed. The robots include coal exploration robots, coal breaking robots, coal loading robots, coal transportation robots, support robots, goaf treatment robots, and inspection robots.

[0005] In a preferred embodiment, step S1 includes: The total number of work sites, the size of the mining area, and the corresponding coal mining volume should meet the order quantity requirements; The number of work locations, the size of the area, the number of robots, and the work time should meet the order delivery time requirements.

[0006] In the multi-objective function, the order completion time refers to the maximum value of the coal mine production order completion time for all robots; The total robot setup time in the multi-objective function includes the tool setup time and position setup time for all robots; the robot tool setup time is the time for the robot to adjust and change tools; the robot position setup time is the time for the robot to adjust and change its work position.

[0007] In the preferred embodiment, the total robot energy consumption in the multi-objective function includes robot operation energy consumption, robot adjustment energy consumption, and robot idle energy consumption, specifically satisfying: Total robot energy consumption = robot operation energy consumption + robot adjustment energy consumption + robot idle energy consumption; Among them, robot operation energy consumption is the sum of the products of each robot's operation power and corresponding operation time; robot adjustment energy consumption is the sum of the products of each robot's adjustment power and corresponding adjustment time; and robot idle energy consumption is the sum of the products of each robot's idle power and corresponding idle time.

[0008] In the preferred embodiment, the coal exploration robot is equipped with coal mine detection tools and is able to perform coal exploration operations, explore the location of coal in the coal mining area, plan suitable mining areas and mining routes, and plan work locations for other robots. The coal breaking robot is equipped with coal mining tools and can perform coal breaking operations, removing rocks and coverings above the coal seam to expose the coal and strip the coal from the rock mass to complete the mining; the coal loading robot is equipped with coal loading tools and can perform coal loading operations, loading the coal stripped from the rock mass by the coal breaking robot into the coal transport robot. Coal transport robots, equipped with coal containers, are capable of performing coal transport operations, holding coal stripped from the rock mass, and transporting it to a designated location. Support robots, equipped with support tools, can perform support operations, install supports in coal mining areas, and prevent rock mass collapse during coal stripping. Goaf treatment robots are equipped with goaf treatment tools and can perform goaf treatment operations. After coal mining, they can promptly fill the goaf with gangue or concrete to maintain the balance of ground pressure, or blast the rocks above the goaf. The inspection robot is equipped with an infrared camera, microphone, temperature sensor, humidity sensor, and gas sensor. It can perform inspection operations, plan inspection routes in coal mining areas, and monitor and provide early warnings for five major disasters: coal dust, water, fire, gas, and roof collapse. In the preferred embodiment, the coal exploration robot starts its work at the same work site earlier than all other robots until all orders are completed, and the same coal exploration robot does not support simultaneous operation at multiple work sites; the coal exploration robot is limited to operating in locations within the coal mine where no other robots are operating. The coal breaking robot starts its work at the same work site later than the coal exploration robot, but earlier than the support robot, coal loading robot, coal transportation robot, and goaf treatment robot. It does not support the same coal loading robot to work at multiple work sites at the same time. Support robots start their operations at the same work site later than coal breaking robots, but earlier than coal loading robots, coal conveying robots, and goaf treatment robots. They also do not support the same support robot operating simultaneously at multiple work sites. The coal loading robot starts its work at the same work site later than the support robot but earlier than the coal transportation robot and the goaf treatment robot, and does not support the same coal loading robot to work at multiple work sites simultaneously. The coal transport robot starts its work at the same work site later than the support robot but earlier than the goaf treatment robot. It does not support the same coal transport robot to work at multiple work sites at the same time. The coal transport robot will only start transporting coal when its coal container is full or when the coal stripped from the rock mass at the same work site is empty. The goaf treatment robot starts its work later than the coal transportation robot at the same work site, and does not support the same goaf treatment robot to work simultaneously at multiple work sites. The inspection robot starts its work later than the coal exploration robot at the same work site and ends its work later than all other robots. It does not support the same inspection robot to work simultaneously at multiple work sites.

[0009] In the preferred embodiment, the value is set to 1 if the robot selects a work location, and 0 if it does not select a work location; the value is set to 1 if it selects a work operation, and 0 if it does not select a work operation.

[0010] In the preferred scheme, the operation includes seven categories of operations and corresponding operation procedures, including the use of specific robot tools, coal exploration, coal breaking, coal loading, coal transportation, support, goaf treatment, and inspection, as well as the specific operation time of each operation procedure. The robots use their own embedded computing platforms and communication networks to perform mobile edge computing; all robots run the same artificial intelligence algorithm, using the multi-objective function of the constraints in step S1 as the fitness function, and the feasible solution with the highest fitness is the optimal solution of the multi-objective function.

[0011] In the preferred embodiment, the robots use their own embedded computing platforms and communication networks to compare their calculated optimal solutions. When the predetermined error conditions and number of iterations are met, they jointly select the optimal solution with the highest fitness, which is the optimal working location and operation for each robot. Based on the work locations and operations of all robots, as well as the order of the operations and the corresponding duration of the operations, calculate the work time for each robot at its work location and during its operations.

[0012] In the preferred embodiment, the coal exploration robot uses coal mine detection tools to explore the location of coal in the coal mine area according to the order of coal exploration operations and the corresponding operation time, and plans suitable mining areas and mining routes, and plans operation locations for other robots. The coal breaking robot uses coal mining tools to remove the rocks and coverings above the coal seam at the work site according to the order of coal breaking operations and the corresponding operation time, thereby exposing the coal and stripping the coal from the rock mass to complete the mining. The coal loading robot uses coal loading tools in the mine to load the coal stripped from the rock by the coal breaking robot into the coal transport robot at the work site according to the order of coal loading operations and the corresponding operation time. Coal transport robots use coal mine containers to hold coal stripped from the rock mass. At the work site, they follow the order of coal transport operations and the corresponding work time, and begin transporting the coal to the designated location when the coal mine container is full or when the coal stripped from the rock mass at the same work site is empty. Support robots use support tools to install supports at the work site in the coal mine area according to the sequence of support operations and the corresponding operation time, so as to prevent rock mass collapse during coal stripping. Goaf treatment robots use goaf treatment tools to fill gangue or concrete in a timely manner after coal mining, maintain ground pressure balance, or blast the rocks above the goaf, according to the order of goaf treatment operations and the corresponding operation time at the work site. The inspection robot uses infrared cameras, microphones, temperature sensors, humidity sensors, and gas sensors to patrol the coal mine area according to the order of inspection operations and the corresponding operation time, and to monitor and warn of five major disasters: coal dust, water, fire, gas, and roof collapse. In the preferred solution, if the coal exploration robot detects new or changed coal mine production orders, the number of work locations, the size of the mining area, and the total amount of coal mined, and these quantities cannot meet the order volume requirements, then the coal exploration operation and the corresponding operation time need to be increased. When the number of work locations, the size of the mining area, the number of robots, and the working time are detected to be insufficient to meet the order delivery time requirements, it is necessary to increase the number of coal breaking robots, add new coal breaking operations, and adjust the corresponding working time. When the support robot detects a change in the working location of the coal breaking robot, it follows the working location of the coal breaking robot to perform support operations and re-estimates the working time corresponding to the new support operations. When the coal loading robot detects a change in the working location of the coal breaking robot, it follows the working location of the coal breaking robot to perform coal loading operations and re-estimates the working time corresponding to the new coal loading operations. When the coal-carrying robot detects a change in the working location of the coal-loading robot, it follows the working location of the coal-loading robot to carry out coal-carrying operations and re-estimates the working time corresponding to the new coal-carrying operations. When the goaf treatment robot detects changes in the work location and completion of corresponding tasks by the coal breaking robot, coal conveying robot, and support robot, it follows the completed work location to perform goaf treatment operations and re-estimates the corresponding operation time for the newly added goaf treatment operation. When the inspection robot detects five major hazards—coal dust, water, fire, gas, and roof collapse—it issues an early warning. All other robots, including coal exploration robots, coal breaking robots, coal loading robots, coal transportation robots, support robots, and goaf treatment robots, suspend operations and calculate evacuation routes according to the artificial intelligence algorithm in step S2. They then evacuate in an orderly manner through the escape passages. The inspection robot evacuates last after monitoring that all other robots have evacuated.

[0013] Compared with the prior art, the beneficial effects of the present invention are as follows: (1) Robot cluster operation enables unmanned coal mine production. This invention requires no human intervention. The multi-objective function simultaneously considers three core indicators: order completion time, total robot adjustment time, and total robot energy consumption. By combining various robot operation operations with dynamic adaptation of virtualizable resources, the optimal combination of production resources is achieved, which significantly improves resource utilization and order delivery efficiency.

[0014] (2) Distributed edge computing enables high efficiency in coal mine production scheduling. The hybrid optimization algorithm improves population diversity through three initialization rules and combines dynamic crossover mutation and tabu search optimization to balance global search and local convergence efficiency. In response to interference such as resource failure and emergency order insertion, a short time window rescheduling mechanism is adopted to achieve rapid response after interference, solving the pain point of weak anti-interference capability of traditional static scheduling.

[0015] (3) Coal exploration, coal breaking, coal loading, coal transportation, support, goaf treatment, and inspection are integrated, making coal mine production safer. When the inspection robot detects the five major disasters of coal dust, water, fire, gas, and roof collapse, it will issue an early warning. The coal exploration robot, coal breaking robot, coal loading robot, coal transportation robot, support robot, and goaf treatment robot will all stop working and run the artificial intelligence algorithm to calculate the evacuation route according to step S2. They will evacuate in an orderly manner through the escape passage. The inspection robot will evacuate last after monitoring that all robots have evacuated.

[0016] (4) Various types of robots use their own embedded computing platforms and communication networks for mobile edge computing, resulting in low computing costs. Virtualizable resources cover all types of manufacturing equipment, personnel, and software, making them suitable for multi-variety, small-batch coal mine production scenarios, with stronger adaptability and scalability. Attached Figure Description

[0017] Figure 1 This is a system structure diagram of an unmanned coal mine production method.

[0018] Figure 2 This is a flowchart of an unmanned coal mine production method. Detailed Implementation

[0019] This embodiment provides a system structure diagram of an unmanned coal mine production method, as shown below. Figure 1 As shown, it includes the following components: rock, coal, work site, support, goaf, coal exploration robot, coal breaking robot, coal loading robot, coal transportation robot, support robot, goaf treatment robot, and inspection robot; furthermore, the coal exploration robot, coal breaking robot, coal loading robot, coal transportation robot, support robot, goaf treatment robot, and inspection robot constitute a coal mine production robot cluster. This embodiment provides an unmanned coal mine production method, such as... Figure 2 As shown, it includes the following steps: S1. The unmanned coal mine production system accepts coal mine production orders and calculates the number of work sites, coal exploration robots, coal breaking robots, coal loading robots, coal transportation robots, support robots, goaf treatment robots, and inspection robots based on the order quantity, order delivery time, and mine size. It establishes a multi-objective function with the core objective of minimizing order completion time, total robot adjustment time, and total robot energy consumption, and establishes constraints for the coal exploration robots, coal breaking robots, coal loading robots, coal transportation robots, support robots, goaf treatment robots, and inspection robots. Step S1 includes the following steps: S1-1. The unmanned coal mine production system accepts coal mine production orders and calculates the number of work sites, coal exploration robots, coal breaking robots, coal loading robots, coal transportation robots, support robots, goaf treatment robots, and inspection robots based on the order quantity, order delivery time, and mine area size. In the preferred scheme, the number of work sites, the size of the mining area, and the total amount of coal mined should meet the order volume requirements; In the preferred scheme, the number of work sites, the size of the mining area, the number of robots, and the work time should meet the order delivery time requirements; The coal exploration robot is equipped with coal mine detection tools and can perform coal exploration operations, exploring the location of coal in the coal mining area and planning suitable mining areas and routes, thus planning work locations for other robots. The coal breaking robot is equipped with coal mining tools and can perform coal breaking operations, removing rocks and coverings above the coal seam to expose the coal and stripping it from the rock mass to complete the mining. The coal loading robot is equipped with coal loading tools and can perform coal loading operations, loading the coal stripped from the rock mass by the coal breaking robot into the coal transport robot. The coal transport robot is equipped with a coal container and can perform coal transport operations to hold the coal stripped from the rock mass. The coal is transported to a designated location; the support robot, equipped with support tools, can perform support operations, installing supports in the coal mining area to prevent rock collapse during coal stripping; the goaf treatment robot, equipped with goaf treatment tools, can perform goaf treatment operations, promptly filling the goaf with gangue or concrete after coal mining to maintain ground pressure balance, or blasting the rock above the goaf; the inspection robot, equipped with an infrared camera, microphone, temperature sensor, humidity sensor, and gas sensor, can perform inspection operations, planning inspection routes in the coal mining area, and monitoring and issuing early warnings for five major hazards: coal dust, water, fire, gas, and roof collapse. S1-2. Establish a multi-objective function with the core objective of minimizing order completion time, total robot setup time, and total robot energy consumption; 1. Minimize the maximum completion time:

[0020] The completion time of work location p is determined by the completion time of the last operation at that work location; The start time of the qth operation at work location p on robot r must be such that the preceding operation is completed and the robot is available. Let be the adjustment time required for robot r before the q-th operation at work location p. If the preceding operation performs the same operation at the same work location, then... =0; The idle time of robot r before the qth operation at work site p is caused by the robot waiting or not being assigned a task. Let q be the actual operation time of the qth operation at work location p on robot r.

[0021] In the preferred scheme, the order completion time in the multi-objective function refers to the maximum value of the coal mine production order completion time for all robots; 2. Minimize total adjustment time:

[0022] In the preferred embodiment, the total robot adjustment time in the multi-objective function includes the tool adjustment time and position adjustment time of all robots; the robot tool adjustment time is the time for the robot to adjust and change tools; the robot position adjustment time is the time for the robot to adjust and change its work position. 3. Minimize total energy consumption:

[0023] in:

[0024] Where E represents energy consumption and P represents power.

[0025] In the preferred embodiment, the total robot energy consumption in the multi-objective function includes robot operation energy consumption, robot adjustment energy consumption, and robot idle energy consumption, specifically satisfying: Total robot energy consumption = robot operation energy consumption + robot adjustment energy consumption + robot idle energy consumption; Among them, robot operation energy consumption is the sum of the products of each robot's operation power and corresponding operation time; robot adjustment energy consumption is the sum of the products of each robot's adjustment power and corresponding adjustment time; and robot idle energy consumption is the sum of the products of each robot's idle power and corresponding idle time.

[0026] S1-3. Establish constraints for coal exploration robots, coal breaking robots, coal loading robots, coal transportation robots, support robots, goaf treatment robots, and inspection robots. The constraints include the following three main constraints: Resource exclusivity: At any given time, a robot can only handle one task.

[0027] Operation constraints: Subsequent operations can only begin after the preceding operations are completed.

[0028] Dynamic availability of resources: .

[0029] The coal exploration robot starts its work earlier than all other robots at the same work site and continues until all orders are completed. It does not support the same coal exploration robot working simultaneously at multiple work sites. The coal exploration robot is limited to working in locations within the coal mine where no other robots are operating. The coal breaking robot starts its work at the same work site later than the coal exploration robot, but earlier than the support robot, coal loading robot, coal transportation robot, and goaf treatment robot, and does not support the same coal loading robot to work at multiple work sites simultaneously. The support robot starts its work at the same work site later than the coal breaking robot, but earlier than the coal loading robot, coal transport robot, and goaf treatment robot, and does not support the same support robot to work at multiple work sites simultaneously. The coal loading robot starts its work at the same work site later than the support robot but earlier than the coal conveying robot and the goaf treatment robot, and does not support the same coal loading robot to work at multiple work sites at the same time. The coal transport robot starts its work at the same work site later than the support robot but earlier than the goaf treatment robot. It does not support the same coal transport robot to work at multiple work sites at the same time. The coal transport robot will only start transporting coal when its coal container is full or when the coal stripped from the rock mass at the same work site is empty. The goaf treatment robot starts its work later than the coal transportation robot at the same work site, and does not support the same goaf treatment robot to work simultaneously at multiple work sites. The inspection robot starts its work later than the coal exploration robot at the same work site and ends its work later than all other robots. It does not support the same inspection robot to work simultaneously at multiple work sites. S2. The unmanned coal mine production system encodes the working location, operation, and time of each robot into a feasible solution. The coal exploration robot, coal breaking robot, coal loading robot, coal transportation robot, support robot, goaf treatment robot, and inspection robot operate in a distributed manner using artificial intelligence algorithms to jointly solve the optimal solution of the multi-objective function that meets the constraints in step S1, that is, the optimal working location, operation, and time of each robot. Step S2 includes the following steps: S2-1, The unmanned coal mine production system encodes the working locations and operations of each robot into feasible solutions; In the preferred embodiment, the coal exploration robot, coal breaking robot, coal loading robot, coal transportation robot, support robot, goaf treatment robot, and inspection robot are configured such that selecting a specific work location sets the value to 1, not selecting a specific work location sets the value to 0, selecting a specific operation sets the value to 1, and not selecting a specific operation sets the value to 0. The operation includes the specific tool usage of the coal exploration robot, coal breaking robot, coal loading robot, coal transportation robot, support robot, goaf treatment robot, and inspection robot, and seven types of operation and corresponding operation procedures, as well as the specific operation time of each operation procedure. Decision variables:

[0030] Indicates whether the q-th operation at work location p is processed on robot r.

[0031] S2-2, Coal exploration robot, coal breaking robot, coal loading robot, coal transportation robot, support robot, goaf treatment robot, inspection robot, distributed operation artificial intelligence algorithm, jointly solve the optimal solution of the multi-objective function that meets the constraints in step S1; In the preferred embodiment, the coal exploration robot, coal breaking robot, coal loading robot, coal transportation robot, support robot, goaf treatment robot, and inspection robot use their respective embedded computing platforms and communication networks for mobile edge computing; all robots run the same artificial intelligence algorithm, using the multi-objective function of the constraints in step S1 as the fitness function, and the feasible solution with the highest fitness is the optimal solution of the multi-objective function; The specific process is as follows: Sub-step S2-2-1: Input of each coal mine production robot cluster to the artificial intelligence algorithm: Coal mine production order operation set Robot Collection The operation time of each task on the robot Robot adjustment time Robot power parameters .

[0032] Sub-step S2-2-2: Each coal mine production robot cluster runs an artificial intelligence algorithm to calculate: Furthermore, coding design: Robot selection layer: an integer sequence representing the robot number selected for each job operation.

[0033] Job operation sorting layer: The integer sequence represents the order of job operations, and the repetition count represents the number of job operations at the job location.

[0034] Furthermore, initialize the rules: Completely random initialization: Robot selection layer: Randomly assigns an available robot to each job operation.

[0035] Job operation sorting layer: Randomly generates the processing order of job operations.

[0036] Shortest task time priority: Robot selection layer: Select the operation time for each task. The shortest robot.

[0037] Job operation sorting layer: sorted in ascending order by the remaining time of the job location (remaining time = the sum of the time of subsequent job operations).

[0038] Lowest energy consumption priority: Robot selection layer: For each operation, the robot with the lowest energy consumption is selected (energy consumption = ).

[0039] Job operation sorting layer: sorted in ascending order of total energy consumption.

[0040] Furthermore, decoding chromosomes: Robot selection decoding: Based on the robot selection layer sequence, determine the processing robot for each operation. .

[0041] Job operation sorting and decoding: Based on the sorting layer sequence, generate the actual processing order of the job operations (which must meet the job operation timing constraints).

[0042] Furthermore, the objective function is calculated using the artificial intelligence algorithm: 1. Maximum completion time:

[0043] 2. Total adjustment time:

[0044] 3. Total energy consumption:

[0045] Non-dominated sorting: Based on the target value ( Pareto rank is used to classify individuals in the population.

[0046] Level 1: Solutions that are not dominated by any other individual (i.e., the optimal solution set).

[0047] Level 2: Solutions dominated only by Level 1 individuals, and so on.

[0048] Furthermore, crossover and mutation operations: For the obtained feasible solutions, perform cross-cutting based on Pareto levels: 1. Crossover rate:

[0049] , Pareto rank of the parent individual (the lower the rank, the better).

[0050] 2. Crossover method: Robot selection layer: Two points intersect, and the robot selection sequence of the parent generation is swapped.

[0051] Job operation sorting layer: Sequential cross (OX), retains fragments of the parent job operation sequence.

[0052] Sub-step S2-2-3: Output of artificial intelligence algorithm by each coal mine production robot cluster: Pareto Frontier Solution Set: The AI ​​algorithm output of coal mine production robot clusters in various locations contains multiple non-dominated solutions, each corresponding to a scheduling scheme.

[0053] Optimal scheduling scheme: Coal mine production robot clusters in various locations select the solution with the minimum proximity from the Pareto front using the TOPSIS method.

[0054] S2-3, All robots determine the optimal solution; Furthermore, the termination conditions and optimization strategies for artificial intelligence algorithms are as follows: Sub-step S2-3-1. Termination condition: The improvement rate of the optimal solution over 10 consecutive generations is less than 0.5%.

[0055] The total number of iterations reaches the preset value (e.g., 500 times).

[0056] Sub-step S2-3-2. Elite retention strategy: Each generation of coal mine production robots retains the top 10% of non-dominated solutions to avoid losing high-quality solutions.

[0057] Sub-step S2-3-3. Local search enhancement: Each coal mine production robot performs tabu search on the Pareto front solution, and the neighborhood action is to randomly exchange the operation order or adjust the resource selection.

[0058] Sub-steps 2-3-4: Different coal mine production robots send their optimal solutions for completing their respective coal mine production orders to the coal mine production robot cluster for comparison, and select the optimal solution from them; Furthermore, the optimal solution for achieving the coal mine production order completion target is compared using a multi-objective decision-making method, as follows: Sub-step S2-3-5. TOPSIS evaluation: (1) Standardize the three objective values ​​and calculate the closeness of the optimal solution for each robot:

[0059] in, The weights are determined using the analytic hierarchy process (AHP).

[0060] (2) Select the solution with the smallest proximity as the optimal solution, and broadcast the robot with the highest optimal solution to the entire coal mine production robot cluster.

[0061] Sub-step 2-3-6. Optimal solution consensus: The optimal solution is distributed and stored in the coal mine production robot cluster. All coal mine production robot clusters reach a consensus and confirm the solution, ensuring that the coal mine production order allocation results are consistent and tamper-proof.

[0062] In the preferred scheme, the coal exploration robot, coal breaking robot, coal loading robot, coal transportation robot, support robot, goaf treatment robot, and inspection robot use their respective embedded computing platforms and communication networks to calculate their optimal solutions and compare them. If they meet the predetermined error conditions and iteration number, they jointly select the optimal solution with the highest fitness, that is, the optimal working location and operation of each robot. Furthermore, based on the working locations and operations of all robots, as well as the order of the operations and the corresponding working time, the working time of each robot in terms of working location and operation is calculated. S3. Coal exploration robots, coal breaking robots, coal loading robots, coal transportation robots, support robots, goaf treatment robots, and inspection robots, according to the optimal solution obtained in step S2, go to the optimal work location and carry out coal mine production according to the work operation and work time; and when changes and interference occur in the coal mine production process, return to steps S1 and S2 to adjust the work location, work operation, and work time of each robot in real time until all coal mine production orders are completed.

[0063] Step S3 includes the following steps: S3-1. Coal exploration robots, coal breaking robots, coal loading robots, coal transportation robots, support robots, goaf treatment robots, and inspection robots, according to the optimal solution obtained in step S2, go to the optimal work location and work time to carry out coal mine production. The coal mine production robot cluster sends the processed coal to the user, and the user settles the payment with the unmanned coal mine production system, thus ending the coal mine production order. In the preferred embodiment, the coal exploration robot uses coal mine detection tools to explore the location of coal in the coal mine area according to the order of coal exploration operations and the corresponding operation time, and plans suitable mining areas and mining routes, and plans operation locations for other robots. Furthermore, the coal breaking robot uses coal mining tools to remove the rocks and coverings above the coal seam at the work site according to the order of coal breaking operations and the corresponding operation time, exposing the coal and stripping the coal from the rock mass to complete the mining. Furthermore, the coal loading robot uses coal mine loading tools to load the coal stripped from the rock mass by the coal breaking robot into the coal transport robot at the work site according to the order of coal loading operations and the corresponding operation time. Furthermore, the coal transport robot uses a coal mine container to hold the coal stripped from the rock mass. At the work site, it follows the order of coal transport operations and the corresponding work time, and begins transporting the coal to the designated location when the coal mine container is full or when the coal stripped from the rock mass at the same work site is empty. Furthermore, the support robot, using support tools, installs supports at the work site in the coal mine area according to the sequence of support operations and the corresponding operation time, to prevent rock mass collapse during coal stripping. Furthermore, the goaf treatment robot uses goaf treatment tools to fill the goaf with gangue or concrete in a timely manner after coal mining, according to the order of goaf treatment operations and the corresponding operation time, to maintain the balance of ground pressure, or to blast the rocks above the goaf. Furthermore, the inspection robot uses infrared cameras, microphones, temperature sensors, humidity sensors, and gas sensors to inspect the coal mine area according to the order of inspection operations and the corresponding operation time, and to monitor and warn of the five major disasters of coal dust, water, fire, gas, and roof collapse. S3-2. When changes or disturbances occur in the coal mine production process, return to steps S1 and S2 to adjust the working location and working time of each robot in real time until all coal mine production orders are completed. The dynamic robot resource state modeling is as follows: Sub-step 3-2-1. Define the robot availability matrix for the coal mine production robot cluster assigned to coal mine production orders. , indicating whether robot r is available at time t.

[0064] Sub-step 3-2-2. The coal mine production robot cluster assigned to coal mine production orders introduces fuzzy time windows to handle sudden maintenance events:

[0065] The decay function characterizes the effect of uncertainty.

[0066] Sub-step 3-2-3: Inputting the coal mine production robot cluster to the artificial intelligence algorithm: (1) Real-time robot resource status: Robot load rate , representing the utilization rate of robot r at time t.

[0067] Equipment Health Index It is dynamically calculated using sensor data (such as vibration and temperature).

[0068] (2) Set of remaining operations Unfinished tasks and their dynamic priorities (such as prioritizing urgent orders).

[0069] (3) Dynamic power parameters: Real-time processing power adjusted based on equipment health status.

[0070] Adjusting power over time Cumulative loss coefficient ( (A factor of aging).

[0071] Sub-step 3-2-4: The coal mine production robot cluster performs calculations using the artificial intelligence algorithm. Sub-step 3-2-4-1, Dynamic coding design of coal mine production robot cluster: Robot selection layer: Integer sequences represent robot assignments for job operations, and priority labels are introduced.

[0072] Job operation sorting layer: Introducing time window encoding: Each job operation in the sequence is subject to a time interval constraint to ensure that the job operation is completed within the specified window.

[0073] Sub-step 3-2-4-2, Calculation of dynamic fitness of coal mine production robot cluster: (1) Weighted maximum completion time:

[0074] : Priority weight of the work location p.

[0075] (2) Dynamic total energy consumption adjustment:

[0076] (3) Real-time total energy consumption:

[0077] Remaining time for the task (dynamically updated as progress progresses).

[0078] : Cumulative idle time of robot r.

[0079] Sub-step 3-2-4-3, Crossover and Mutation Driven by Reinforcement Learning for Coal Mine Production Robot Clusters: (1) State-action mapping: State space S: contains Real-time indicators such as the percentage of work progress.

[0080] Action Space A: Robot reassignment, adjustment of job operation time windows, such as compressing or expanding the window.

[0081] (2) Optimization of reward function:

[0082] in, Average equipment health index; encourages the selection of high-reliability robot resources. Health weighting coefficient Sub-step 3-2-4-4, Online taboo search for coal mine production robot clusters: (1) Dynamic neighborhood generation: Robot switching neighborhood: Randomly select a task and assign it to a robot with a lower load or a healthier robot.

[0083] Sliding neighborhood of time window: Slides the time window of a certain operation forward or backward. .

[0084] (2) Taboo list update: Record recent adjustments to avoid repeating the same actions within a short period.

[0085] Sub-step 3-2-5: Output of the artificial intelligence algorithm to the coal mine production robot cluster: Sub-step 3-2-5-1, Dynamic scheduling scheme for coal mine production robot cluster: real-time updated robot allocation and time plan for operation.

[0086] Sub-step 3-2-5-2, Suggestions for handling anomalies in coal mine production robot clusters: For robots with low health, a maintenance time window or replacement solution is recommended.

[0087] Sub-step 3-2-6: Adjust the operation and scheduling of the coal mine production robot according to the interference and dynamic changes in the manufacturing process. Repeat step 3-2 until all the coal ore specified in the coal mine production order has been processed. Among them, the triggering conditions are: Robot malfunction (vibration sensor data exceeds threshold) Urgent order insertion request (user-added orders have higher priority than current tasks). Rescheduling methods: Based on the current state, the remaining processes are extracted, and the MOGATS algorithm is rerun in 15-minute windows to generate a local scheduling scheme.

[0088] In the preferred solution, if the coal exploration robot detects new or changed coal mine production orders, the number of work locations, the size of the mining area, and the total amount of coal mined, and these quantities cannot meet the order volume requirements, then the coal exploration operation and the corresponding operation time need to be increased. If the number of work locations, the size of the mining area, the number of robots, and the working time of the coal breaking robots cannot meet the order delivery time requirements, then the number of coal breaking robots needs to be increased, and new coal breaking operations and corresponding working times need to be added. If the support robot detects a change in the working location of the coal breaking robot, it will follow the working location of the coal breaking robot to perform support operations and re-estimate the working time corresponding to the new support operations. If the coal loading robot detects a change in the working location of the coal breaking robot, it will follow the working location of the coal breaking robot to perform coal loading operations and re-estimate the working time corresponding to the new coal loading operations. If the coal-carrying robot detects a change in the working location of the coal loading robot, it will follow the working location of the coal loading robot to carry out coal-carrying operations and re-estimate the working time corresponding to the new coal-carrying operations. If the goaf treatment robot detects a change in the work location and completion of the corresponding work of the coal breaking robot, coal conveying robot, or support robot, it will follow the completed work location to perform goaf treatment operations and re-estimate the operation time corresponding to the newly added goaf treatment operation. When the inspection robot detects five major hazards—coal dust, water, fire, gas, and roof collapse—it issues an early warning. All other robots, including coal exploration robots, coal breaking robots, coal loading robots, coal transportation robots, support robots, and goaf treatment robots, suspend operations and run an artificial intelligence algorithm to calculate evacuation routes, following step S2. They then evacuate in an orderly manner through the escape passages. The inspection robot evacuates last after all other robots have evacuated.

[0089] The above embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit it. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. An unmanned coal mine production method, characterized in that: Includes the following steps: S1. Accept coal mine production orders, calculate the number of work sites and the number of robots based on order quantity, order delivery time, and mine size; establish a multi-objective function with minimizing order completion time, total robot adjustment time, and total robot energy consumption as the core, and establish constraints for the robots; S2. The unmanned coal mine production system encodes the work location, operation, and time of each robot into a feasible solution; the robots operate in a distributed manner using artificial intelligence algorithms to jointly solve the optimal solution of the multi-objective function that meets the constraints in step S1, that is, the optimal work location, operation, and time of each robot. S3. The robot goes to the optimal work location according to the optimal solution obtained in step S2 and carries out coal mine production according to the work operation and work time. When changes and disturbances occur in the coal mine production process, it returns to steps S1 and S2 to adjust the work location, work operation and work time of each robot in real time until all coal mine production orders are completed. The robots include coal exploration robots, coal breaking robots, coal loading robots, coal transportation robots, support robots, goaf treatment robots, and inspection robots.

2. The unmanned coal mine production method according to claim 1, characterized in that: Step S1 includes the following steps: The total number of work sites, the size of the mining area, and the corresponding coal mining volume should meet the order quantity requirements; The number of work locations, the size of the area, the number of robots, and the work time should meet the order delivery time requirements; In the multi-objective function, the order completion time refers to the maximum value of the coal mine production order completion time for all robots; The total robot setup time in the multi-objective function includes the tool setup time and position setup time for all robots; the robot tool setup time is the time for the robot to adjust and change tools; the robot position setup time is the time for the robot to adjust and change its work position.

3. The unmanned coal mine production method according to claim 1, characterized in that: The total energy consumption of the robot in the multi-objective function includes the robot's operational energy consumption, robot adjustment energy consumption, and robot idle energy consumption, specifically satisfying the following: Total robot energy consumption = Robot operation energy consumption + Robot adjustment energy consumption + Robot idle energy consumption; Among them, robot operation energy consumption is the sum of the products of each robot's operation power and corresponding operation time; robot adjustment energy consumption is the sum of the products of each robot's adjustment power and corresponding adjustment time; and robot idle energy consumption is the sum of the products of each robot's idle power and corresponding idle time.

4. The unmanned coal mine production method according to claim 1, characterized in that: Coal exploration robots are equipped with coal mine detection tools and can perform coal exploration operations. They can explore the location of coal in the coal mining area, plan suitable mining areas and mining routes, and plan work locations for other robots. The coal breaking robot is equipped with coal mining tools and can perform coal breaking operations, removing rocks and coverings above the coal seam to expose the coal and strip the coal from the rock mass to complete the mining; the coal loading robot is equipped with coal loading tools and can perform coal loading operations, loading the coal stripped from the rock mass by the coal breaking robot into the coal transport robot. Coal transport robots, equipped with coal containers, are capable of performing coal transport operations, holding coal stripped from the rock mass, and transporting it to a designated location. Support robots, equipped with support tools, can perform support operations, install supports in coal mining areas, and prevent rock mass collapse during coal stripping. Goaf treatment robots are equipped with goaf treatment tools and can perform goaf treatment operations. After coal mining, they can promptly fill the goaf with gangue or concrete to maintain the balance of ground pressure, or blast the rocks above the goaf. The inspection robot is equipped with an infrared camera, microphone, temperature sensor, humidity sensor, and gas sensor. It can perform inspection operations, plan inspection routes in coal mining areas, and monitor and provide early warnings for five major hazards: coal dust, water, fire, gas, and roof collapse.

5. The unmanned coal mine production method according to claim 1, characterized in that: The coal exploration robot starts its work earlier than all other robots at the same work site and continues until all orders are completed. It does not support the same coal exploration robot operating simultaneously at multiple work sites. The coal exploration robot is limited to operating in locations within the coal mine where no other robots are operating. The coal breaking robot starts its work at the same work site later than the coal exploration robot, but earlier than the support robot, coal loading robot, coal transportation robot, and goaf treatment robot. It does not support the same coal loading robot to work at multiple work sites at the same time. Support robots start their operations at the same work site later than coal breaking robots, but earlier than coal loading robots, coal conveying robots, and goaf treatment robots. They also do not support the same support robot operating simultaneously at multiple work sites. The coal loading robot starts its work at the same work site later than the support robot but earlier than the coal transportation robot and the goaf treatment robot, and does not support the same coal loading robot to work at multiple work sites simultaneously. The coal transport robot starts its work at the same work site later than the support robot but earlier than the goaf treatment robot. It does not support the same coal transport robot to work at multiple work sites at the same time. The coal transport robot will only start transporting coal when its coal container is full or when the coal stripped from the rock mass at the same work site is empty. The goaf treatment robot starts its work later than the coal transportation robot at the same work site, and does not support the same goaf treatment robot to work simultaneously at multiple work sites. The inspection robot starts its work later than the coal exploration robot at the same work site and ends its work later than all other robots. It does not support the same inspection robot to work simultaneously at multiple work sites.

6. The unmanned coal mine production method according to claim 1, characterized in that: The value is set to 1 if the robot selects a specific work location, and to 0 if it does not select a specific work location. The value is also set to 1 if the robot selects a specific work operation, and to 0 if it does not select a specific work operation.

7. The unmanned coal mine production method according to claim 1, characterized in that, The operation includes the use of specific tools of the robot, coal exploration, coal breaking, coal loading, coal transportation, support, goaf treatment, and inspection, totaling seven categories of operation and corresponding operation procedures, as well as the specific operation time of each operation procedure. The robots use their own embedded computing platforms and communication networks to perform mobile edge computing; all robots run the same artificial intelligence algorithm, using the multi-objective function of the constraints in step S1 as the fitness function, and the feasible solution with the highest fitness is the optimal solution of the multi-objective function.

8. The unmanned coal mine production method according to claim 1, characterized in that: The robots use their own embedded computing platforms and communication networks to calculate their optimal solutions and compare them. When the predetermined error conditions and number of iterations are met, they jointly select the optimal solution with the highest fitness, which is the optimal working location and operation for each robot. Based on the work locations and operations of all robots, as well as the order of the operations and the corresponding duration of the operations, calculate the work time for each robot at its work location and during its operations.

9. The unmanned coal mine production method according to claim 1, characterized in that: Coal exploration robots use coal mine detection tools to explore coal locations in the mining area according to the order of coal exploration operations and the corresponding operation time, and plan suitable mining areas and mining routes, thus planning operation locations for other robots. The coal breaking robot uses coal mining tools to remove the rocks and coverings above the coal seam at the work site according to the order of coal breaking operations and the corresponding operation time, thereby exposing the coal and stripping the coal from the rock mass to complete the mining. The coal loading robot uses coal loading tools in the mine to load the coal stripped from the rock by the coal breaking robot into the coal transport robot at the work site according to the order of coal loading operations and the corresponding operation time. Coal transport robots use coal mine containers to hold coal stripped from the rock mass. At the work site, they follow the order of coal transport operations and the corresponding work time, and begin transporting the coal to the designated location when the coal mine container is full or when the coal stripped from the rock mass at the same work site is empty. Support robots use support tools to install supports at the work site in the coal mine area according to the sequence of support operations and the corresponding operation time, so as to prevent rock mass collapse during coal stripping. Goaf treatment robots use goaf treatment tools to fill gangue or concrete in a timely manner after coal mining, maintain ground pressure balance, or blast the rocks above the goaf, according to the order of goaf treatment operations and the corresponding operation time at the work site. Inspection robots use infrared cameras, microphones, temperature sensors, humidity sensors, and gas sensors to patrol coal mines according to the order and duration of inspection operations, monitoring and issuing early warnings for five major hazards: coal dust, water, fire, gas, and roof collapse.

10. The unmanned coal mine production method according to claim 1, characterized in that: If the coal exploration robot detects new or changed coal mine production orders, the number of work locations, the size of the mining area, and the total amount of coal mined cannot meet the order quantity requirements, then the coal exploration operation and the corresponding operation time need to be increased. When the number of work locations, the size of the mining area, the number of robots, and the working time are detected to be insufficient to meet the order delivery time requirements, it is necessary to increase the number of coal breaking robots, add new coal breaking operations, and adjust the corresponding working time. When the support robot detects a change in the working location of the coal breaking robot, it follows the working location of the coal breaking robot to perform support operations and re-estimates the working time corresponding to the new support operations. When the coal loading robot detects a change in the working location of the coal breaking robot, it follows the working location of the coal breaking robot to perform coal loading operations and re-estimates the working time corresponding to the new coal loading operations. When the coal-carrying robot detects a change in the working location of the coal-loading robot, it follows the working location of the coal-loading robot to carry out coal-carrying operations and re-estimates the working time corresponding to the new coal-carrying operations. When the goaf treatment robot detects changes in the work location and completion of corresponding tasks by the coal breaking robot, coal conveying robot, and support robot, it follows the completed work location to perform goaf treatment operations and re-estimates the corresponding operation time for the newly added goaf treatment operation. When the inspection robot detects five major hazards—coal dust, water, fire, gas, and roof collapse—it issues an early warning. All other robots, including coal exploration robots, coal breaking robots, coal loading robots, coal transportation robots, support robots, and goaf treatment robots, suspend operations and calculate evacuation routes according to the artificial intelligence algorithm in step S2. They then evacuate in an orderly manner through the escape passages. The inspection robot evacuates last after monitoring that all other robots have evacuated.

Images