Intelligent dispatching method and system for underground coal mine trackless rubber-tyred vehicle

Abstract

The invention provides a coal mine underground trackless rubber-tyred vehicle intelligent scheduling method and system, and relates to the field of coal mine underground vehicle scheduling, and the method comprises the steps: obtaining vehicle application tasks of underground group personnel and the position state information of vehicles, the vehicle application tasks comprising the number of required vehicle personnel, the getting-on and getting-off positions, the vehicle types and the use time; and based on a preset intelligent scheduling algorithm, carrying out joint optimization on the vehicle application task and the position state information by taking the path cost and the vehicle calling cost as an optimization target, and carrying out trackless rubber-tyred vehicle scheduling according to a scheme obtained through optimization. According to the embodiment of the invention, the underground trackless rubber-tyred vehicle of the coal mine can be intelligently dispatched through an information technology means, so that safe, efficient and ordered operation of the trackless rubber-tyred vehicle is realized.

Description

Technical Field

[0001] The present invention relates to the technical field of vehicle dispatching in coal mines, and in particular to an intelligent dispatching method and system for trackless rubber-tyred vehicles in coal mines. Background Art

[0002] As a vital energy source and industrial raw material, coal has always played a vital role in energy supply and supporting industrial operations. With increasing mining depths, the increasing use of trackless auxiliary transportation equipment, such as man-carriages, material carts, and multi-purpose vehicles, has led to underground traffic congestion, frequent traffic accidents, air pollution caused by excessive exhaust emissions, and disorderly waiting areas, all of which have seriously impacted mine safety and production. However, traditional manual transportation scheduling relies on empirical decision-making, resulting in low efficiency and prone to errors. Summary of the Invention

[0003] To solve the above problems, an embodiment of the present invention provides an intelligent scheduling method for rubber-tyred trackless vehicles in coal mines, the method comprising: obtaining vehicle application tasks of underground team members and vehicle location status information, the vehicle application tasks comprising: the number of personnel required to use the vehicle, boarding and alighting locations, vehicle type and usage time; based on a preset intelligent scheduling algorithm, the vehicle application tasks and the location status information are jointly optimized with the combined minimization of path cost and vehicle call cost as the optimization goal, and the trackless rubber-tyred vehicles are scheduled according to the optimized solution; the intelligent scheduling algorithm comprises: global dynamic path planning based on the optimal path algorithm, priority evaluation and task allocation mechanism, and multi-vehicle collaboration and mixed task scheduling based on a genetic algorithm.

[0004] The intelligent dispatching method for rubber-tyred trackless vehicles in underground coal mines provided by the embodiment of the present invention can intelligently dispatch rubber-tyred trackless vehicles in underground coal mines through information technology means, thereby realizing safe, efficient and orderly operation of the rubber-tyred trackless vehicles.

[0005] Optionally, the global dynamic path planning based on the optimal path algorithm includes: using a deep Q network to dynamically learn the characteristics of the coal mine underground traffic network, and incorporating real-time congestion conditions and underground harmful gas concentrations into the calculation of the path cost, and dynamically calculating the path cost; the multi-vehicle collaboration and mixed task scheduling based on the genetic algorithm includes: incorporating the path cost into the genetic algorithm, and the genetic algorithm performs path optimization with the comprehensive minimization of path cost and vehicle call cost as the optimization goal.

[0006] Optionally, the multi-vehicle collaboration and mixed task scheduling based on genetic algorithms also includes: splitting the tasks in which the number of personnel required for the vehicle does not exceed or exceeds the capacity of a single vehicle, and incorporating different splitting results into the genetic algorithm respectively to obtain a path plan with the lowest combined path cost and vehicle call cost.

[0007] Optionally, the genetic algorithm includes: if the number of people required for a certain task demand point is greater than the maximum capacity limit of a single vehicle, first use the prior splitting method to formulate a strategy in advance to decompose the demand point task, and then use the genetic algorithm to allocate the split demand to different vehicles.

[0008] Optionally, the genetic algorithm further includes: performing chromosome encoding on the vehicle orders obtained by splitting the tasks to obtain an ordered sequence, cutting the sequence according to the capacity constraints of the vehicles to obtain driving routes for several vehicles; splitting the chromosomes and allocating them to different vehicles to obtain driving routes for several vehicles.

[0009] Optionally, the genetic algorithm further comprises: using a labeling method to record all vehicles and paths that may enter the current node, then deleting the dominated labels through Pareto optimization, and finally selecting a combined path with the lowest cost from the remaining non-dominated labels.

[0010] Optionally, the priority evaluation and task allocation mechanism includes: establishing a task priority model, sorting the tasks based on the urgency of the tasks and a model trained with historical data; predicting task completion time and vehicle utilization through a decision tree algorithm, and adjusting the allocation strategy in real time to dynamically allocate tasks.

[0011] An embodiment of the present invention provides an intelligent dispatching system for trackless rubber-tyred vehicles in underground coal mines. The intelligent dispatching system for people and vehicles includes: a user-end application for submitting vehicle applications and checking vehicle status, a sensor module for real-time feedback of vehicle and personnel information, a storage module for storing historical vehicle operation data, a dispatching center server for receiving user applications and executing an intelligent dispatching algorithm to dynamically manage vehicle resources, and a cloud computing platform for providing computing power support; the dispatching center server is used to execute any of the above-mentioned intelligent dispatching methods for trackless rubber-tyred vehicles in underground coal mines.

[0012] Optionally, the sensor module includes a vehicle position sensor, a vehicle load sensor, and a vehicle status sensor; the vehicle position sensor is used to obtain the vehicle's position information in real time and feed it back to the dispatch center server in real time; the vehicle load sensor is used to detect the number of passengers and load conditions in the vehicle, and feed it back to the dispatch center server in real time; the vehicle status sensor is used to monitor the vehicle's operating status and safety status, and feed it back to the dispatch center server in real time.

[0013] Optionally, the storage module includes a data storage unit and a history record management unit; the data storage unit is used to store the position, load and status data fed back by the sensor module; the history record management unit is used to store the vehicle's historical operation trajectory, scheduling records and operation status data.

[0014] The intelligent dispatching system for trackless rubber-tyred vehicles in underground coal mines provided by the embodiment of the present invention can achieve the same technical effect as the above-mentioned intelligent dispatching method for trackless rubber-tyred vehicles in underground coal mines. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are merely embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on the provided drawings without paying any creative work.

[0016] Figure 1 The present invention is a schematic flow chart of an intelligent dispatching method for trackless rubber-tyred vehicles in underground coal mines according to an embodiment of the present invention. DETAILED DESCRIPTION

[0017] In order to make the above-mentioned objects, features and advantages of the present invention more clearly understood, the following detailed description of the specific embodiments of the present invention is given in conjunction with the accompanying drawings. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0018] The embodiment of the present invention provides an intelligent dispatching method and system for trackless rubber-tyred vehicles in coal mines, which replaces manual transportation dispatching with intelligent dispatching by information technology means, thereby realizing safe, efficient and orderly operation of trackless rubber-tyred vehicles.

[0019] The intelligent dispatching system for rubber-tyred trackless vehicles in underground coal mines includes: a user-side application for submitting vehicle applications and checking vehicle status, a sensor module for real-time feedback of vehicle and personnel information, a storage module for storing historical vehicle operation data, a dispatching center server for receiving user applications and executing intelligent dispatching algorithms to dynamically manage vehicle resources, and a cloud computing platform for providing computing power support.

[0020] Among them, the user-side application includes vehicle application, vehicle information query and feedback functions. Specifically: team members fill out and submit vehicle applications based on the number of people required, vehicle type and usage time. They can also view the vehicle's current location and estimated arrival time in real time through the mobile application, and allow feedback or application for task adjustments, thereby improving the flexibility and adaptability of scheduling.

[0021] The sensor module mainly includes a vehicle position sensor, a vehicle load sensor, and a vehicle status sensor. Specifically:

[0022] Vehicle position sensors are primarily used to obtain real-time vehicle location information and provide feedback to the dispatch center server. Vehicle load sensors are primarily used to detect the number of passengers and load in a vehicle for reference in dispatch decisions. Vehicle status sensors are primarily used to monitor the vehicle's operating status (such as speed and remaining battery power) and safety status (such as fault detection and shutdown status), and provide real-time feedback to the dispatch center server.

[0023] The information storage module mainly includes a data storage unit and a history record management unit, specifically:

[0024] The data storage unit is primarily used to store real-time position, load, and status data fed back by the sensor module. The history management unit is primarily used to store the vehicle's historical operation trajectory, dispatch records, and operating status to facilitate subsequent analysis and optimization.

[0025] The dispatch center server is primarily responsible for receiving user applications and dynamically managing vehicle resources through a built-in intelligent scheduling algorithm. This intelligent scheduling algorithm primarily includes the following functions: global dynamic path planning based on an optimal path algorithm, priority assessment and task allocation mechanisms, and multi-vehicle coordination and mixed task scheduling based on a genetic algorithm.

[0026] For example, the dispatch center server uses the D* algorithm for global dynamic path planning. The algorithm comprehensively considers factors such as transportation distance, time, road conditions and tire wear, and automatically adjusts the driving route when real-time road conditions change, such as traffic jams, to ensure optimal transportation efficiency and safety.

[0027] The intelligent scheduling system establishes a task priority model, sorts tasks based on the urgency of the task (such as shift handover, emergency tasks, etc.) and the model trained with historical data, and uses the decision tree algorithm to predict task completion time and vehicle utilization, adjust the allocation strategy in real time, and dynamically allocate transportation tasks.

[0028] The intelligent dispatch system supports the coordinated dispatch of multiple vehicle types. Its algorithms automatically match vehicle types based on factors such as task type, distance, and timeliness, effectively allocating tasks. When task requirements exceed the capacity of a single vehicle, the system splits the task and distributes it among multiple vehicles. Using a variety of dispatch algorithms, the system achieves the decomposition and optimal allocation of complex tasks.

[0029] Specifically, if the number of passengers required at a task demand point is greater than the maximum capacity limit of a single vehicle, a strategy is first formulated in advance using the prior splitting method to decompose the task at the demand point, and then a genetic algorithm is used to allocate the split demand to different vehicles.

[0030] Specifically, when the shift is changing and the mine is being lifted up, the locations of underground vehicles and task demand points will be particularly scattered. At this time, the labeling method is used to generate alternative labels to assist in task allocation and path planning. The Pareto idea is introduced when designing the genetic algorithm to inherit high-quality labels and intelligently match available vehicles.

[0031] The cloud computing platform can provide powerful computing capabilities, capable of processing and analyzing sensor data from various vehicles and personnel underground in real time, and executing the built-in algorithms of the dispatch center server to provide support for the dispatch center server.

[0032] Figure 1 The following is a schematic flow chart of an intelligent dispatching method for trackless rubber-tyred vehicles in a coal mine according to an embodiment of the present invention, wherein the method comprises the following steps:

[0033] S102: Obtain vehicle application tasks and vehicle location status information from underground team members. The vehicle application tasks include: the number of personnel required, boarding and alighting locations, vehicle type, and usage time; the vehicle location status information includes vehicle location information, vehicle passenger count and load, vehicle operating status (such as speed and remaining battery power), and safety status (such as fault detection and shutdown status).

[0034] S104: Using a pre-set intelligent scheduling algorithm, the vehicle application task and location status information are jointly optimized to minimize the combined path cost and vehicle call cost. Rubber-tyred trackless vehicles are then dispatched based on the resulting solution. This intelligent scheduling algorithm includes global dynamic path planning based on an optimal path algorithm, a priority assessment and task allocation mechanism, and multi-vehicle coordination and mixed task scheduling based on a genetic algorithm.

[0035] Among them, the global dynamic path planning based on the optimal path algorithm includes: using the deep Q network to dynamically learn the characteristics of the underground traffic network in coal mines, and incorporating real-time congestion conditions and underground harmful gas concentrations into the calculation of path costs, and dynamically calculating the path costs.

[0036] The Deep Q-Network (DQN) is used to dynamically learn the characteristics of the underground traffic network. In addition to the underground road conditions and the distances between underground road network nodes, data such as the real-time underground congestion situation and the concentration of harmful gases underground are incorporated into the calculation of path costs. The cost of path calculation is dynamically adjusted and incorporated into the genetic algorithm intelligent scheduling algorithm to minimize the combined path cost and vehicle call cost.

[0037] In the coal mine underground dispatching system, the Deep Q Network (DQN) is used to dynamically learn the characteristics of the underground traffic network, construct a path cost optimization objective function, and dynamically adjust it in combination with the reinforcement learning model. The algorithm is as follows:

[0038] 1. Path cost objective function

[0039] Define the path cost C path As one of the optimization objectives, the specific expression is as follows:

[0040]

[0041] Where: d i is the physical distance of the i-th path (meters); c i is the real-time congestion coefficient of the i-th path, reflecting the underground traffic efficiency (value range: 0≤c i ≤1); g i is the concentration coefficient of harmful gases on the i-th path (ppm); α and β are weight parameters that control the impact of congestion and harmful gases on path cost.

[0042] 2. Constraints

[0043] To ensure driving safety and scheduling efficiency, we set the following constraints:

[0044] 1) Path reachability constraints:

[0045]

[0046] Among them, x i Indicates that path i is selected, otherwise it is not selected.

[0047] 2) Gas concentration safety threshold constraints:

[0048] g i ≤G max

[0049] Among them, G max The safety concentration threshold is set, and the path beyond this value is unavailable.

[0050] 3) Driving time constraints:

[0051]

[0052] Among them, v i is the average speed of the i-th path, T max The maximum allowed travel time for the task.

[0053] 3. Deep Reinforcement Learning (DQN) Optimized Path Planning

[0054] In path optimization, DQN dynamically adjusts path selection through the state-action-reward mechanism:

[0055] Status S t: Underground road network status, including path length, real-time congestion, harmful gas concentration, etc.;

[0056] Action A t : Select the next path;

[0057] Reward function R t :

[0058] R t =-C path (S t ,A t )

[0059] The reward is the negative of the path cost, which makes the model tend to choose the low-cost path.

[0060] The Q value update formula is as follows:

[0061]

[0062] Where: η is the learning rate; γ is the discount factor, balancing short-term and long-term benefits.

[0063] The mathematical model of the scheduling system is as follows:

[0064] (1) Objective function

[0065] The scheduling objective can be optimized through the following multiple objective functions. Common objectives include minimizing the total path cost, total vehicle call cost, task delay, etc.:

[0066] min C total =α·C path +β·C vehicle +γ·C delay

[0067] Where: C path is the path cost, that is, the total driving cost of the vehicle required for the task. vehicle It is the cost of calling the vehicle, including the cost of using the vehicle, fuel cost, etc. delay is the task delay cost, which is used to measure whether the task is completed on time. α, β, and γ are weight coefficients used to balance the relative importance of each goal.

[0068] (2) Task Splitting and Vehicle Allocation Constraints When the task demand exceeds the capacity of a single vehicle, the task needs to be split and assigned to multiple vehicles. This can be described by the following constraints.

[0069] Task assignment constraint: Each task can be assigned to multiple vehicles (v i For vehicles, T j for tasks): Among them, x ij =1 indicates task Tj Assigned to vehicle v i , otherwise 0.

[0070] Task splitting constraint: Task T j Split into multiple subtasks H k , the demand for each subtask does not exceed the capacity of the vehicle: Among them, d j It is subtask H k The demand, C i Is the vehicle v i capacity.

[0071] (3) Vehicle scheduling constraints

[0072] Vehicle usage constraints: Each vehicle cannot exceed the maximum working time or maximum driving distance: Among them, t ij Indicates vehicle v i Execute Task T j The time required, T max The maximum operating time of the vehicle.

[0073] Task time constraints: Each task must be completed within a specified time: Among them, Completion Time j It is task T j Completion time, D j It is task T j deadline.

[0074] (4) Path planning

[0075] Path planning is the core issue after task assignment, and it needs to consider factors such as the distance between task nodes, vehicle speed, and traffic restrictions. The connection between task nodes and vehicles is described in a graph. The goal of path planning is to find the shortest path and minimize vehicle travel costs.

[0076] Shortest path problem: Among them, d(v i ,v i+1 ) is the node v i To node v i+1 The distance between them.

[0077] Algorithm input and output

[0078] enter:

[0079] (1) Task Demands: task location (starting point and end point), task timeliness requirements (such as start time, completion time, etc.), task demand, and task priority.

[0080] (2) Vehicle Information: The load / capacity of each vehicle, the speed of each vehicle, the driving cost (including fuel costs, maintenance costs, etc.), the vehicle type (7-passenger vehicle, 19-passenger vehicle, electric vehicle, gasoline vehicle), the location of the vehicle, and the availability status (whether it is idle).

[0081] (3) Geographical Information: the distance between task nodes (used for path planning), map information (such as traffic conditions, restricted road sections, etc.), and the distance or time between each task point and the current vehicle.

[0082] (4) Scheduling constraints: maximum driving time or maximum driving distance of a vehicle, restrictions on task splitting, and task priority rules (scheduling priority of high-priority tasks).

[0083] Output:

[0084] Scheduling results: the type of vehicle assigned to each task and the task allocation plan, the task route of each vehicle (the path from the starting point to the end point), the total cost after task allocation, including path cost, vehicle call cost, etc., the vehicle scheduling plan, including the vehicle's departure time and completion time, and the task completion time, to ensure that the timeliness of the task is met.

[0085] The above-mentioned multi-vehicle coordination and mixed task scheduling based on genetic algorithms includes: integrating path costs into the genetic algorithm, and the genetic algorithm optimizes the path with the comprehensive minimization of path costs and vehicle call costs as the optimization goal.

[0086] Path costs are calculated using the aforementioned deep learning algorithm combined with deep reinforcement learning. For example, the vehicle dispatch cost is determined by the type and number of vehicles dispatched, which can generally be categorized as gasoline and electric vehicles, and 7-passenger and 19-passenger vehicles. The overall minimum can be expressed as min(path cost + k * vehicle dispatch cost). By adjusting the coefficient k, different dispatch preferences can be generated. For example, when a temporary task occurs, a large k may favor dispatching vehicles that have already completed their task underground, while a small k may favor dispatching new vehicles from the depot.

[0087] Multi-vehicle coordination and mixed-task scheduling based on genetic algorithms also involves splitting tasks where the number of required vehicles and crews does not exceed or exceeds the capacity of a single vehicle. The resulting splits are then fed into the genetic algorithm to determine the path with the lowest combined path and vehicle dispatch costs. The decision on whether to split tasks and how to split them is based on the algorithm's solution.

[0088] If the number of people required for a task demand point is greater than the maximum capacity limit of a single vehicle, a strategy is first formulated in advance using the prior splitting method to decompose the task at the demand point, and then a genetic algorithm is used to allocate the split demand to different vehicles.

[0089] Specifically, a splitting strategy is developed based on task demand and vehicle capacity: when the demand for a single task is greater than the maximum capacity of a single vehicle, it must be split to ensure that the task can be distributed among multiple vehicles to avoid overload.

[0090] Flexible splitting strategy based on optimization objectives: For tasks that do not exceed the capacity of a single vehicle, splitting decisions can be made based on optimization objectives. Splitting can minimize empty seats and optimize the number of vehicles dispatched. The splitting results are incorporated into the combined optimization objectives of route cost and vehicle dispatch cost, using a genetic algorithm for route planning and vehicle scheduling optimization.

[0091] The splitting decision is made automatically or manually by the scheduling system: For this type of task, whether to split is calculated by the scheduling system based on the real-time vehicle scheduling situation and task requirements. The dispatcher can make the final decision based on the suggestions provided by the system.

[0092] The vehicle orders obtained by task splitting are chromosomally encoded to obtain an ordered sequence. According to the vehicle capacity constraints, the sequence is cut to obtain the driving routes of several vehicles; then, the chromosomes are split and assigned to different vehicles to obtain the driving routes of several vehicles.

[0093] For example, the chromosome encoding process is as follows: All car orders are encoded as an ordered list of real numbers and used in optimization processes such as crossover and mutation. For example, if there are 10 car orders (after splitting), the chromosome is first constructed as an ordered sequence of [1-2-3-4-5-6-7-8-9-10]. Then, based on the vehicle capacity constraints, the sequence is segmented to obtain the travel routes of several vehicles.

[0094] The chromosome decoding process is as follows: decoding is the process of splitting the chromosome and assigning it to different vehicles to obtain the driving routes of several vehicles. This embodiment uses the labeling method for decoding.

[0095] The core idea of ​​this method is to traverse all available vehicles at each node, starting from the beginning of the chromosome and ending at the end. Labels are used to record all possible vehicles and paths that could have entered that node, including the total number of vehicles used to reach the current node and the total cost of the path. Then, through Pareto optimization, the dominated labels are removed, and the lowest-cost combined path is selected from the remaining non-dominated labels. (Entering this node indicates that the order has been assigned.)

[0096] The above genetic algorithm also includes: using a labeling method to record all vehicles and paths that may enter the current node, then deleting the dominated labels through Pareto optimization, and finally selecting a combined path with the lowest cost from the remaining non-dominated labels.

[0097] Pareto optimization means, for example, that by node 3, label 7 uses cars v1 and v2 for a total cost of 10, while label 8 uses cars v1, v2, and v3 for a total cost of 15. Because label 8 calls a newer car than label 7 and has a higher total cost than label 1, label 7 dominates label 8. In reality, the remaining capacity of cars v1 and v2, already used by the previous node, can still meet node 3's order demand. Therefore, there is no need to call v3, so label 8 is deleted.

[0098] General decoding assigns orders to vehicles according to the coding order and vehicle order. The solution efficiency and quality are very low, and it is impossible to achieve mixed scheduling of multiple models and the local deployment of scattered vehicles underground.

[0099] Fitness is assessed as follows: A fitness function, also known as an evaluation function, measures an individual's ability to adapt to its environment. It serves as a criterion for distinguishing good individuals from bad individuals within a population, determined by an objective function. The fitness function is always non-negative, while the objective function can be positive or negative, necessitating a conversion between the objective function and the fitness function. Chromosomes with lower costs have higher fitness.

[0100] The objective function of the problem studied in this embodiment is a minimization problem, so the fitness function is transformed as follows:

[0101] F(x)=f max -f(x)

[0102] Where, f max It is the upper bound of f(x), which is the historical maximum value after chromosome decoding.

[0103] The selection operation is as follows: The purpose of the selection operation is to select the best individuals from the current population so that they can reproduce. The probability model of the selection operation is usually designed so that the selection probability of an individual is proportional to its fitness, that is, individuals with high fitness are more likely to be selected.

[0104] The selection strategy in this embodiment is tournament selection. N individuals are sampled from the entire population and allowed to compete (a tournament) to select the best individual. The number of individuals participating in the tournament is called the tournament size. Typically, n = 2 is the most commonly used size, also known as binary tournament selection (BTS).

[0105] The crossover operation is as follows: In genetic algorithms, crossover refers to the process of generating offspring from two or more parent individuals. Its main function is to maintain population diversity, prevent premature convergence, and help search for the global optimal solution.

[0106] The crossover method used in this example is Order Crossover (OX). The starting and ending locations are randomly selected on the two parent chromosomes. The genes in the region of parent chromosome 1 are copied to the same locations on progeny 1. The missing genes in progeny 1 are then sequentially inserted on parent chromosome 2. Another progeny is obtained in a similar manner. The specific steps are as follows:

[0107] Step 1: Randomly select the start and end positions of several genes in a pair of chromosomes (parent generation) (the selected positions of the two chromosomes are the same).

[0108] Step 2: Generate two offspring and ensure that the position of the selected genes in the offspring is the same as that in the parent.

[0109] Step 3: First find the position of the gene selected in the first step in the other parent generation, and then put the remaining genes into the offspring generated in the previous step in order.

[0110] Mutation is a random change in certain genes within an individual with a certain probability. This is done to maintain genetic diversity in the population and prevent the algorithm from prematurely converging to a local optimum. The design of the mutation operator involves setting the mutation probability and selecting the mutation method. This randomly changes certain genes within an individual to increase population diversity. Common mutation operations include single-point mutation, multi-point mutation, and uniform mutation.

[0111] This example uses a swap mutation, which is suitable for permutation problems. In this mutation, two genes are randomly selected and their positions are swapped to introduce randomness.

[0112] Optionally, the priority evaluation and task allocation mechanism includes: establishing a task priority model to sort tasks based on the urgency of the tasks and a model trained with historical data; using a decision tree algorithm to predict task completion time and vehicle utilization, and adjusting the allocation strategy in real time to dynamically allocate tasks.

[0113] This embodiment adds coal mine gas monitoring sensors to detect gas concentrations (such as methane and carbon monoxide) underground in real time and incorporates this data into the scheduling algorithm. It also adds data processing capabilities for underground cameras, using visual sensors to perceive the environment and assist in determining whether underground passages are unobstructed. A self-diagnostic mechanism is incorporated into the sensor module to regularly assess device status and predict failure risks in advance, thereby preventing scheduling system malfunctions caused by sensor failure.

[0114] Specifically, taking carbon monoxide as an example:

[0115] First Threshold (T1): The safe lower concentration limit (e.g., 50 ppm). Within this concentration range, carbon monoxide concentrations are considered safe. Vehicle Scheduling: There will be no restrictions on vehicle scheduling, and all vehicles can be dispatched normally.

[0116] Second threshold (T2): upper limit of safe concentration (for example, 100ppm). Carbon monoxide concentration begins to rise, but is still within a relatively safe range, but may have some impact on people's health. Vehicle dispatch: restrict vehicle dispatch in some potentially dangerous areas, and issue reminders to drivers, reminding them to wear necessary safety equipment (such as respirators, gas detectors, etc.). For example, restrict some vehicles from entering areas with higher concentrations, or require drivers to check gas concentrations in advance.

[0117] The third threshold (T3): Dangerous concentration (e.g., 200 ppm). When the concentration reaches this threshold, the gas concentration has entered the dangerous range, posing a serious threat to miners and vehicles. Vehicle dispatch: Strictly limit the number of vehicles entering the high-concentration area, allowing only vehicles with emergency missions to enter. Strict safety measures are required, such as equipping vehicles with rescue equipment and limiting mission execution time, to ensure the safe return of vehicles.

[0118] The intelligent dispatching method and system for trackless rubber-tyred vehicles in underground coal mines provided by the embodiments of the present invention can intelligently dispatch trackless rubber-tyred vehicles in underground coal mines through information technology means, thereby realizing safe, efficient and orderly operation of the trackless rubber-tyred vehicles.

[0119] The embodiment of the present invention further provides a computer-readable storage medium having a computer program stored thereon. When the computer program is executed by a processor, the various processes of the embodiment of the intelligent dispatching method for trackless rubber-tyred vehicles in underground coal mines described above are implemented, and the same technical effects are achieved. To avoid repetition, the details are not described here. The computer-readable storage medium is, for example, a read-only memory (ROM), a random access memory (RAM), a magnetic disk, or an optical disk.

[0120] Of course, those skilled in the art will understand that all or part of the processes in the above-mentioned embodiment methods can be implemented by instructing the control device through a computer program, and the program can be stored in a computer-readable storage medium. When the program is executed, it may include the processes of the above-mentioned method embodiments, wherein the storage medium may be a memory, a disk, an optical disk, etc.

[0121] In this document, relational terms such as first and second, etc., are used only to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "comprises," "comprising," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus comprising a series of elements includes not only those elements, but also other elements not explicitly listed, or elements inherent to such process, method, article, or apparatus. In the absence of further limitations, an element defined by the phrase "comprising a ..." does not exclude the presence of additional identical elements in the process, method, article, or apparatus comprising the element.

[0122] The various embodiments in this specification are described in a progressive manner, and each embodiment focuses on the differences from other embodiments. The same or similar parts between the various embodiments can be referenced to each other.

[0123] The above description of the disclosed embodiments enables one skilled in the art to implement or use the present invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention is not limited to the embodiments shown herein, but is to be construed in the widest possible manner consistent with the principles and novel features disclosed herein.

Claims

1. An intelligent dispatching method for trackless rubber-tyred vehicles in coal mines, characterized in that: The method comprises: Obtain vehicle application tasks and vehicle location status information from underground team members. The vehicle application tasks include: the number of personnel required to use the vehicle, the location of the vehicle to be boarded and disembarked, the vehicle type, and the usage time; Based on a preset intelligent scheduling algorithm, the vehicle application task and the location status information are jointly optimized with the combined minimum of path cost and vehicle call cost as the optimization goal, and the trackless rubber-tyred vehicles are dispatched according to the optimized solution; the intelligent scheduling algorithm includes: global dynamic path planning based on the optimal path algorithm, priority evaluation and task allocation mechanism, and multi-vehicle collaboration and mixed task scheduling based on the genetic algorithm.

2. The method according to claim 1, characterized in that The global dynamic path planning based on the optimal path algorithm includes: The deep Q network is used to dynamically learn the characteristics of the underground coal mine traffic network, and the real-time congestion situation and the concentration of harmful gases underground are incorporated into the calculation of the path cost to dynamically calculate the path cost. The multi-vehicle coordination and mixed task scheduling based on genetic algorithm includes: The path cost is incorporated into the genetic algorithm, and the genetic algorithm performs path optimization with the comprehensive minimization of the path cost and the vehicle calling cost as the optimization goal.

3. The method according to claim 2, characterized in that The multi-vehicle coordination and mixed task scheduling based on genetic algorithm also includes: The tasks for which the number of personnel required for the vehicle does not exceed or exceeds the capacity of a single vehicle are split, and different split results are respectively incorporated into the genetic algorithm to obtain a path plan with the minimum combined path cost and vehicle call cost.

4. The method according to claim 3, characterized in that The genetic algorithm comprises: If the number of people required for a task demand point is greater than the maximum capacity limit of a single vehicle, a strategy is first formulated in advance using the prior splitting method to decompose the task at the demand point, and then a genetic algorithm is used to allocate the split demand to different vehicles.

5. The method according to claim 4, characterized in that The genetic algorithm also includes: Chromosome encoding is performed on the vehicle orders obtained by splitting the tasks to obtain an ordered sequence, and the sequence is cut according to the capacity constraints of the vehicles to obtain the driving routes of several vehicles; The chromosomes are split and assigned to different vehicles to obtain the driving routes of several vehicles.

6. The method according to claim 5, characterized in that The genetic algorithm also includes: Use the labeling method to record all vehicles and paths that may enter the current node, then use Pareto optimization to delete the dominated labels, and finally select a combined path with the lowest cost from the remaining non-dominated labels.

7. The method according to claim 2, characterized in that The priority assessment and task allocation mechanism includes: Establish a task priority model to sort the tasks based on their urgency and the model trained with historical data; Through the decision tree algorithm, the task completion time and vehicle utilization rate are predicted, and the allocation strategy is adjusted in real time to dynamically allocate tasks.

8. An intelligent dispatching system for trackless rubber-tyred vehicles in coal mines, characterized in that: The intelligent vehicle dispatching system includes: a user-side application for submitting vehicle applications and checking vehicle status, a sensor module for providing real-time feedback on vehicle and personnel information, a storage module for storing historical vehicle operation data, a dispatching center server for receiving user applications and executing intelligent dispatching algorithms to dynamically manage vehicle resources, and a cloud computing platform for providing computing power support; The dispatching center server is used to execute the intelligent dispatching method for rubber-tyred trackless vehicles in underground coal mines as described in any one of claims 1-7.

9. The system according to claim 8, characterized in that The sensor module includes a vehicle position sensor, a vehicle load sensor, and a vehicle status sensor; The vehicle position sensor is used to obtain the vehicle's position information in real time and feed it back to the dispatch center server in real time; The vehicle load sensor is used to detect the number of passengers and load conditions in the vehicle and provide real-time feedback to the dispatch center server; The vehicle status sensor is used to monitor the operating status and safety status of the vehicle and provide real-time feedback to the dispatch center server.

10. The system according to claim 8, wherein: The storage module includes a data storage unit and a history record management unit; The data storage unit is used to store the position, load and status data fed back by the sensor module; The history record management unit is used to store the historical running track, dispatch records and running status data of the vehicle.

Images