Intelligent dispatching and path planning system and method for mine carry-scraper

Abstract

The invention discloses an intelligent scheduling and path planning system and method for a mine carry-scraper, and belongs to the technical field of mining, and the system comprises an environment sensing module which is used for collecting the environment data of a mine operation area in real time, and the environment data comprise the equipment state, the road condition and the dynamic obstacle position; processing the collected data to generate environment state parameters; and the task scheduling module is used for allocating operation tasks according to the equipment state, the task target and the environment state parameters, and performing global optimization on task allocation to minimize task completion time and energy consumption. Generating a task allocation scheme based on the equipment state and the task target through a task scheduling module; the path planning module realizes optimal path planning through multi-device game optimization, and the global optimization control module coordinates the work of each module, optimizes global targets such as task completion time, energy consumption and conflicts, realizes deep fusion of scheduling and path planning, and improves the working efficiency, global optimality and robustness of mining devices.

Description

Technical Field

[0001] The present invention relates to an intelligent scheduling and path planning system and method for a mine loader, belonging to the technical field of mine exploitation. Background Art

[0002] With the continuous improvement of the requirements for production efficiency and resource utilization rate in the mine exploitation industry, the intelligent scheduling and path planning of mine equipment has become a key research direction. At present, the scheduling and path planning technologies of loaders and trucks are mostly applied to the management of automated mine equipment. Although the traditional technical solutions have achieved a certain degree of automation to some extent, there are still several technical limitations and it is difficult to meet the global optimization requirements in complex dynamic environments.

[0003] In the prior art, task scheduling and path planning are usually two independent processes, and path planning is carried out after the scheduling scheme is generated. Due to the different optimization objectives of the two, it is easy to lead to local optimality. For example, task scheduling aims to minimize the task completion time, while path planning aims to find the shortest path. This separation method fails to fully consider the coupling relationship between the two, resulting in low overall operating efficiency.

[0004] In the prior art, the cooperative scheduling of multiple devices mainly relies on pre-set rules or simple priority strategies, lacking globality and dynamics. When a road collapse or equipment failure occurs in a certain area, it is impossible to dynamically re-plan the paths and tasks of multiple devices, and the system is difficult to adapt to the changes in complex environments.

[0005] In the prior art, task scheduling and path planning are mainly based on static environmental information, assuming that the task and equipment states do not change. However, in real mine operation scenarios, road congestion, dynamic obstacles (such as other equipment), and equipment state changes (such as insufficient energy or failure) are common phenomena. The existing systems cannot quickly re-plan the paths, which may cause equipment stagnation or even operation failure.

[0006] The prior art usually only performs scheduling or path planning for a single optimization objective. For example, some technologies aim to minimize the task completion time while ignoring equipment energy consumption and the balance of task completion, resulting in low overall system performance. Summary of the Invention

[0007] The purpose of the present invention is to overcome the deficiencies of the above-mentioned existing technologies and provide an intelligent scheduling and path planning system and method for a mine loader, which significantly improves the operation efficiency, robustness, and global performance of the mine loader through the deep integration of scheduling and path planning, multi-device collaborative optimization, dynamic environment adaptation, and multi-objective optimization.

[0008] The technical solution provided by the present invention is as follows: An intelligent scheduling and path planning system for a mine loader, comprising:

[0009] An environmental perception module, which is used to collect environmental data of the mine operation area in real time, including equipment status, road conditions, and dynamic obstacle positions, and process the collected data to generate environmental status parameters;

[0010] A task scheduling module, which is used to allocate operation tasks according to equipment status, task objectives, and environmental status parameters, and globally optimize the task allocation to minimize task completion time and energy consumption;

[0011] A path planning module, which is used to plan the optimal path of the equipment based on the operation tasks allocated by the task scheduling module and the environmental status parameters provided by the environmental perception module;

[0012] A multi-device collaboration module, which is used to coordinate the competition and collaboration of multiple devices in path planning and solve path conflicts between devices;

[0013] A real-time dynamic adjustment module, which is used to dynamically adjust task scheduling and path planning when the environmental status or equipment status changes;

[0014] A global optimization control module, which is used to coordinate the work of the environmental perception module, task scheduling module, path planning module, multi-device collaboration module, and real-time dynamic adjustment module, and optimize the global scheduling and path planning scheme.

[0015] Preferably, the environmental perception module includes:

[0016] A lidar, which is used to construct a three-dimensional map of the mine operation area;

[0017] A GPS, which is used to obtain the accurate position information of the equipment;

[0018] A camera, which is used to capture images of roads, obstacles, and the operation environment;

[0019] A data processing unit, which is used to fuse the raw data collected by the lidar, GPS, and camera, and generate environmental status parameters for path planning and task scheduling.

[0020] Preferably, the task scheduling module includes:

[0021] A task allocator, which is used to allocate global task objectives to multiple devices;

[0022] A global optimization unit, which is used to optimize the task allocation scheme to minimize task completion time and equipment energy consumption, where the optimization objective is the weighted sum of task completion time and energy consumption, and generate a task scheduling scheme based on the status and task objectives of all devices.

[0023] Preferably, the task allocation optimization of the global optimization unit is based on the following constraints:

[0024] The distance between the current position of the device and the target task is the shortest;

[0025] The remaining energy of each device meets the task execution requirements;

[0026] Each task is executed by a unique device.

[0027] Preferably, the path planning module includes:

[0028] A single-device path optimization unit for calculating the optimal path to the target position according to the task target and current position of the device, where the path optimization considers the device speed, steering angle, and obstacles in the environment;

[0029] A multi-device game optimization unit for solving multi-device path conflicts through a game model and optimizing the multi-device path plan to ensure the collaborative optimality of all device paths.

[0030] Preferably, the multi-device game optimization unit constructs a path competition model between devices based on differential game theory and solves the optimal path of the device through the Nash equilibrium of path optimization.

[0031] Preferably, the multi-device collaboration module includes:

[0032] A path conflict detector for detecting the path intersection points between devices and calculating the risk value of path conflicts;

[0033] A collaboration optimization unit for adjusting the device priorities of the conflicting paths and dynamically updating the control strategy of the path planning module.

[0034] Preferably, the real-time dynamic adjustment module includes:

[0035] A dynamic monitoring unit for monitoring the current position, remaining energy, task completion degree, and real-time environmental status of the device;

[0036] A dynamic adjustment unit for re-optimizing the task scheduling plan and path planning plan according to the status information provided by the dynamic monitoring unit.

[0037] Preferably, the global optimization control module uses the variational method to optimize the global objective function of task scheduling and path planning. The global objective function includes the weighted sum of task completion time, device energy consumption, and path conflicts, and determines the global optimal scheduling and path planning plan by solving the Euler-Lagrange equation.

[0038] The present invention also provides a method for intelligent scheduling and path planning of a mine loader, including the following steps:

[0039] S1. Real-time collect the environmental data of the mine operation area through the environmental perception module, construct a three-dimensional map using lidar, GPS, and cameras, and generate environmental state parameters.

[0040] S2. Based on the equipment status, task objectives, and environmental state parameters, allocate operation tasks through the task scheduling module, and optimize the task allocation scheme through the global optimization unit. The optimization objective is the weighted sum of task completion time and energy consumption.

[0041] S3. Through the path planning module, according to the operation tasks assigned by the task scheduling module and the environmental state parameters, plan the optimal path for each device to reach the task objective, and solve path conflicts through multi-device game optimization to ensure the collaborative optimality of path planning.

[0042] S4. Through the multi-device collaboration module, detect the path intersection points between devices, calculate the path conflict risk value, and adjust the device priorities of the conflicting paths through the collaboration optimization unit to dynamically update the path control strategy of the path planning module.

[0043] S5. Through the real-time dynamic adjustment module, when the environmental state or device state changes, monitor the current position, remaining energy, and task completion degree of the device, and dynamically adjust the task scheduling plan and path planning plan.

[0044] S6. Through the global optimization control module, optimize the global objective function of task scheduling and path planning based on the variational method to achieve intelligent scheduling and path planning of mine operations.

[0045] The beneficial effects of the present invention are as follows: 1. The present invention adopts a global optimization technical solution combining differential game and variational method, achieving the technical effect of deep integration of scheduling and path planning. Compared with the prior art technical solutions in which task scheduling and path planning are independent of each other and lack global optimal control, it solves the problems of path conflicts, uneven task allocation, and insufficient global optimality easily occurring in multi-device collaboration in the prior art, and significantly improves the overall operation efficiency of mine equipment.

[0046] 2. The present invention adopts an optimization strategy based on differential game in the multi-device collaboration module, achieving the technical effect of multi-device path conflict detection and collaborative optimization. Compared with the prior art technical solutions that only rely on heuristic algorithms for path planning and cannot dynamically handle multi-device competition, it solves the path conflicts and resource competition problems of devices in complex dynamic environments, ensuring the path coordination and real-time performance of multi-devices when executing tasks.

[0047] 3. The present invention adopts an optimization mechanism that combines a real-time dynamic adjustment module with environmental perception data, achieving the technical effect of quickly responding and real-time adjusting task planning in a dynamic mine environment. Compared with the prior art solutions that only rely on static environmental data for planning, it solves the problems of poor system robustness and planning failure caused by environmental changes (such as road congestion, equipment failures, etc.), enhancing the adaptability and reliability of the system to complex mine operation environments.

[0048] 4. The global objective function optimization method based on the variational method in the present invention achieves the technical effect of minimizing the task completion time and energy consumption. Compared with the prior art solutions with a single optimization objective, it solves the deficiencies of the existing methods that cannot simultaneously consider task efficiency, energy consumption reduction, and path coordination, realizes the unified modeling of multi-objective optimization, and improves the global performance of the mine loader scheduling and path planning system. BRIEF DESCRIPTION OF THE DRAWINGS

[0049] Figure 1 is the system framework diagram of the present invention;

[0050] Figure 2 is the method flow diagram of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0051] The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Apparently, the described embodiments are only a part of the embodiments of the present invention, rather than all of them. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0052] As Figure 1 shown, a mine loader intelligent scheduling and path planning system includes:

[0053] An environmental perception module for real-time collecting environmental data of the mine operation area, including equipment status, road conditions, and dynamic obstacle positions, and processing the collected environmental data to generate environmental status parameters;

[0054] A task scheduling module for allocating operation tasks according to equipment status, task objectives, and environmental status parameters, and globally optimizing the task allocation to minimize the task completion time and energy consumption;

[0055] A path planning module for planning the optimal path of the equipment based on the operation tasks allocated by the task scheduling module and the environmental status parameters provided by the environmental perception module;

[0056] A multi-device cooperation module for coordinating the competition and cooperation of multiple devices in path planning and solving path conflicts between devices;

[0057] A real-time dynamic adjustment module, which is used to dynamically adjust task scheduling and path planning when the environmental state or device state changes;

[0058] A global optimization control module, which is used to coordinate the work of the environmental perception module, task scheduling module, path planning module, multi-device cooperation module and real-time dynamic adjustment module, and optimize the global scheduling and path planning scheme.

[0059] Among them, the environmental perception module in the present invention is the basic part of the intelligent scheduling and path planning system for mine loaders. Its main function is to collect environmental data of the mine operation area in real time through a variety of sensors and data processing devices, including equipment status, road conditions, and the positions of dynamic obstacles, etc., to provide basic data support for the task scheduling module and path planning module. This module closely cooperates with the task scheduling module and path planning module. Through the collected and processed environmental state parameters, the global optimization control can adapt to the complex and changeable dynamic environment of the mine. The specific technical details are as follows:

[0060] In this embodiment, the environmental perception module includes the following main parts:

[0061] A lidar unit, which is used to scan the three-dimensional space information of the mine operation area. Generally, the lidar emits laser beams at a fixed frequency, captures the reflection signals of the laser and environmental objects, and generates point cloud data. As an option, the point cloud data can be further processed to construct a high-precision three-dimensional map of the mine operation area, which is used to identify the boundaries of operation areas such as roads, obstacles, and ore accumulation areas. Specifically, the scanning range and resolution of the lidar can be adjusted according to the scale of the mine. For example, in a small mine environment, a device with a small scanning range but high resolution can be selected, while in a large mine, a lidar with a larger coverage range is preferred.

[0062] GPS, which is used to accurately locate the positions of loaders and other equipment and obtain accurate position information of the equipment. As an implementation method, the GPS receives signals from multiple satellites to provide the real-time coordinates of the equipment. Specifically, the positioning accuracy is related to the terrain complexity of the mine. For complex terrains, in some embodiments, an inertial navigation system (INS) can be combined to improve the positioning reliability. In some relatively extreme scenarios, the GPS signal may be blocked or interfered. In this case, the inertial navigation data can be fused with the GPS signal through the Kalman filtering algorithm to maintain the continuity of the position data.

[0063] The camera unit is used to capture images of the mine road, obstacles, and the operating environment. Generally, the camera processes the environmental images through a Convolutional Neural Network (CNN) to achieve the recognition and classification of key areas such as obstacles and mine road signs in the mine scene. In a possible implementation, the camera data can be combined with the point cloud data of the lidar to form an environment perception model based on vision-point cloud fusion. This fusion method can improve the accuracy of obstacle recognition, especially when the obstacle has a low reflectivity or the lidar signal is difficult to capture. Specifically, the fusion algorithm can adopt Bayesian filtering or deep neural network technology to output the accurate position and category information of the obstacle.

[0064] The data processing unit is used to fuse the raw data collected by the lidar, GPS, and camera, and generate environmental state parameters for path planning and task scheduling. Generally speaking, the data processing unit includes a preprocessing module, a data fusion module, and an output module. Specifically, the preprocessing module filters, denoises, and normalizes the raw data. For example, the random noise in the lidar data can be removed by the median filtering method, while the noise in the camera image can be processed by Gaussian filtering. As a possible implementation, the data fusion module adopts an algorithm based on the Extended Kalman Filter (EKF) to fuse multi-source data into unified environmental state parameters. In some cases, a multi-sensor data processing framework based on graph optimization can be adopted to improve the robustness and real-time performance of environmental perception.

[0065] In the specific implementation of the present invention, the definition of the environmental state parameters includes:

[0066] The relative position between devices:

[0067] d ik (t) = ||x i，pos (t) - x k，pos (t)||

[0068] where, d ik (t) represents the Euclidean distance between device i and device k at time t, reflecting the relative position relationship between the two devices in three-dimensional space; x i，pos (t) and x k，pos (t) represent the current position coordinates of device i and device k; ||·|| represents the Euclidean norm (L2 norm), that is, the length of the vector or the straight-line distance between two points.

[0069] The spatial position and motion state of dynamic obstacles. As an implementation, the center point, boundary, and speed of dynamic obstacles can be analyzed through lidar point clouds, and a linear motion model is used to predict their future positions. For example, the motion state of an obstacle can be expressed as:

[0070] pobstacle (t + Δt) = p obstacle (t) + v obstacle (t)·Δt

[0071] Where p obstacle (t + Δt) represents the position coordinates of the dynamic obstacle at time t + Δt, p obstacle (t) is the current position, v obstacle (t) is the current speed, and Δt is the time interval.

[0072] The accessibility index of the road traffic condition includes slope, curvature, and surface friction coefficient. In some embodiments, the accessibility index can be obtained by fusing data from lidar and cameras. For example, the slope can be calculated by fitting the road plane with point cloud data, and the friction coefficient can be inferred by analyzing the image texture features.

[0073] Environmental risk assessment data, such as the location and extent of the collapse area or low visibility area. This information can be obtained by combining historical data and real-time collected data. Generally, through a convolutional neural network model to perform regional segmentation on the mine camera images, detect the boundaries of the risk areas, and combine with the point cloud depth information of the lidar to generate an environmental risk assessment map.

[0074] As a basic module, the environmental perception module is core to providing accurate, real-time, and efficient environmental data support for task scheduling and path planning. The collaborative work of the lidar, GPS, camera, and data processing unit in the environmental perception module is a key technical link for realizing the intelligent scheduling and path planning of the mine loader.

[0075] The task scheduling module is an important part of the intelligent scheduling and path planning system of the mine loader. Its function is to generate an optimal task allocation plan based on the equipment status, road conditions, and dynamic obstacle positions provided by the environmental perception module, and provide task inputs for the path planning module. The task scheduling module needs to comprehensively consider time, energy consumption, and task priorities in a complex scenario of multiple devices and multiple tasks, and achieve reasonable resource allocation through optimization methods to maximize the utilization efficiency of the equipment. Collaborating closely with other modules of the system, the task scheduling module provides preconditions for path planning and at the same time accepts real-time feedback from the dynamic adjustment module to correct the scheduling plan.

[0076] In this embodiment, the task scheduling module includes two main parts: a task allocator and a global optimization unit.

[0077] The role of the task allocator is to initially allocate task objectives. Generally, the task allocator first matches the possibility of each device with the task to be allocated based on the device's current location, current task status, and remaining energy. The core of this step lies in the initial screening based on geographical distance and task attributes. For example, in some embodiments, devices that are closer will be preferentially allocated to the target task, especially when the task has strict time requirements. This approach provides a range of feasible solutions for global optimization by initially defining the matching relationship between devices and tasks, reducing the complexity of subsequent optimization.

[0078] As an option, the task allocator can adopt a scoring-based task matching strategy. Specifically, the task allocator calculates the matching score S of each device i with task j ij , and the scoring formula is:

[0079]

[0080] where S ij represents the matching score of device i for task j, d ij represents the Euclidean distance between device i and task j; E i represents the current remaining energy of device i; E i，min is the minimum energy required for the device to complete the task; ω j represents the priority weight of task j, which is determined by the urgency and importance of the task; represents the relative adequacy of the device's remaining energy, and the more sufficient the energy, the higher the priority.

[0081] Through the above scoring formula, the task allocator can quickly complete the initial allocation and at the same time provide input data for the global optimization unit.

[0082] In a possible implementation, the global optimization unit optimizes the task allocation scheme through variational methods. Generally, the goal of the global optimization unit is to minimize the global task completion time and energy consumption of the system. The optimization objective function can be expressed as:

[0083]

[0084] C ij =α 1 T ij +α 2 E ij

[0085] where J represents the global objective function, which is used to describe the comprehensive optimization objective of all devices, and C ij represents the cost of device i executing task j; T ij is the estimated time for device i to travel from its current location to task j; E ijThe energy consumption for device i to execute task j; α 1 Is the weight factor for the task time cost, used to balance the time required to complete the task with other optimization objectives, α 2 Is the weight factor for the task energy consumption, used to balance the energy consumed when the task is completed with other objectives; X ij Is a decision variable, taking values of 0 or 1, indicating whether device i executes task j; N represents the total number of devices; M represents the total number of tasks.

[0086] As an option, the optimization process needs to satisfy the following constraints:

[0087] Each task must be executed by one and only one device:

[0088]

[0089] Among them, Indicates that task j must be executed by one of the N devices and only one device; X ij Is the decision variable for task allocation, indicating whether device i is assigned to execute task j; N represents the total number of devices; M represents the total number of tasks; Indicates that the constraint of the formula applies to all tasks j.

[0090] The total energy consumption of each device shall not exceed its remaining energy:

[0091]

[0092] Among them, Indicates that the total energy consumption for device i to execute the tasks assigned to it cannot exceed the current remaining energy E of device i i ; E ij Indicates the energy consumption required for device i to execute task j; X ij Indicates the task allocation decision variable; E i Indicates the current remaining energy of device i, reflecting the current power or fuel reserve status of the device; N represents the total number of devices; M represents the total number of tasks; Indicates that this constraint applies to all devices i.

[0093] The travel path length of the device needs to be within a predetermined range:

[0094] d ij ≤D max ,

[0095] Among them, d ij Indicates the Euclidean distance between device i and task j; D max Indicates the maximum distance limit that device i can accept; Xij is a decision variable for task allocation; indicates that this constraint only applies when device i has been assigned to execute task j.

[0096] Through the above objective function and constraints, the global optimization unit can generate an optimal task allocation scheme within the search space. The optimization method can use the gradient descent method to solve, and by iteratively updating the decision variable X ij value, gradually approaching the optimal solution.

[0097] The path planning module is a key component of the present invention, responsible for calculating the optimal path from the current position to the task target for each device. Combining the task target assigned by the task scheduling module and the environmental state parameters provided by the environmental perception module, the path planning module not only has to plan the optimal path for a single device, but also solve the problem of path conflicts among multiple devices. This module constructs a dynamic state evolution equation, combines differential game theory and optimization algorithms to ensure the global feasibility and local optimality of the path, and at the same time adapts to the real-time changes of the dynamic environment.

[0098] Generally, the inputs of the path planning module include task targets, device states, and environmental states. After these data are dynamically modeled, an optimized path plan is generated. As an option, the path planning module can also handle dynamic environmental changes, such as road blockages or the appearance of obstacles, and adjust the path in real time. Specifically, the implementation of the path planning module involves two major parts: single-device path optimization and multi-device collaborative optimization. The following is a detailed description in combination with specific implementation methods.

[0099] The path planning of each device is modeled by a state evolution equation, and its state change is described by the following formula:

[0100]

[0101] Wherein, represents the state change rate of device i at time t, and x i (t) is the state variable of device i at time t, including the position x i,pos (t), the task completion degree x i,task (t), and the remaining energy x i,energy (t); u i,path (t) is the path control variable of device i, which determines the speed and driving direction of the device; e(t) is the external dynamic environmental state parameter, including road traffic conditions, dynamic obstacle positions, weather conditions, etc.; f i represents the dynamic evolution function of the device state.

[0102] Specifically, the position change of the device is defined by the following equation:

[0103]

[0104] Among them, represents the rate of change of the position of device i at time t, and v i (t) represents the speed of device i; θ i (t) is the driving direction angle of the device; cos(θ i (t)) and sin(θ i (t)) respectively represent the horizontal and vertical components corresponding to the device motion direction angle θ i (t) in the two-dimensional plane, which are used to decompose the velocity vector.

[0105] The energy consumption of the device is calculated by the following formula:

[0106]

[0107] Among them, represents the rate of change of the energy of device i at time t, and h i (u i，path (t)) represents the instantaneous energy consumption function of device i under the path control u i，path (t); u i，path (t) represents the path control variable of device i; γ 1 and γ 2 are constants of the device energy consumption model; v i represents the current speed of device i.

[0108] The change in the task completion degree is described by the following equation:

[0109]

[0110] Among them, represents the rate of change of the task completion degree of device i executing the task at time t; β is the task completion rate constant; u i，task (t) represents the task execution state, taking values of 0 or 1.

[0111] In some embodiments, the single-device path optimization is solved by the optimal control method. The optimization goal is to minimize the path length and energy consumption. Generally, the optimal path control variable can be calculated iteratively by the gradient descent algorithm.

[0112] To solve the path conflicts among multiple devices, this module uses the differential game model to describe the competitive and cooperative behaviors of the paths of multiple devices. The path optimization goal of multiple devices is:

[0113]

[0114] Among them, min indicates that the optimization goal is to find the minimum value; J i is the cost function of device i, which includes the path length, energy consumption, and conflict risk; N is the total number of devices; Denotes the sum of the cost functions J for all N devices. i of.

[0115] The path conflict risk of the device is calculated by the following formula:

[0116] I ik (x i , x k ) = α 3 exp(-||x i，pos (t) - x k，pos (t)|| 2 )

[0117] where I ik is the conflict risk between the paths of device i and device k; α 3 is the weight factor of the path conflict cost; exp(·) represents the exponential function; ||x i，pos (t) - x j，pos (t)|| represents the Euclidean distance between device i and device k at time t.

[0118] As an option, the multi-device path optimization is solved by the Nash equilibrium of differential game theory. Generally, the optimal path control strategy of the device satisfies the following conditions:

[0119]

[0120] where u ， represents the control input of device i; H ， is the Hamiltonian function of device i; C i，path (x i , u i ) is the weighted cost of path length and energy consumption; ∑ k≠i I ik (x i , x k ) is the sum of the interaction costs between device i and other devices; I ik (x i , x k ) represents the conflict risk between the paths of device i and device k; x i represents the state vector of device i; x k represents the state vector of device k.

[0121] In some embodiments, the resolution of path conflicts is achieved by adjusting the priorities of the devices. For example, the paths of devices with higher priorities remain unchanged, and the paths of devices with lower priorities are adjusted to avoid conflicts.

[0122] The multi-device collaboration module mainly solves the conflict problems of multi-device paths and task assignments in mining operations, ensuring coordinated operations among multiple loaders or haul trucks under limited resources. This module is closely related to the path planning module and the task scheduling module. The path planning module provides the optimal path plan for a single device, while the task scheduling module assigns task goals. The responsibility of this module is to detect potential path conflicts between devices and dynamically adjust path and task assignments based on the conflict risk. By introducing game theory and conflict detection mechanisms, this module can operate effectively in a complex dynamic environment with the assistance of the real-time dynamic adjustment module.

[0123] Generally, the path of each device is generated by the path planning module, and the task is assigned by the task scheduling module. However, due to the large number of devices and the complex operating environment, path conflicts may occur among multiple devices in the same area. Specifically, a path conflict means that the planned paths of two or more devices cross or overlap, resulting in devices competing for resources in a similar space at the same time. Therefore, the multi-device collaboration module designs a path conflict detector and a collaboration optimization unit to dynamically adjust path and task priorities.

[0124] In some embodiments, the multi-device collaboration module performs path conflict detection based on device status, task goals, and path planning results. Generally, the device status includes its current position x i，pos (t), speed v i (t), and device travel direction angle θ i (t). The path conflict detector evaluates whether there is an intersection point between the paths of two devices, device i and device k. As an option, the calculation of the path conflict risk I ik can be performed using the following formula:

[0125] I ik (x i , x k ) = α 3 ·exp(-||x i，pos (t) - x k，pos (t)|| 2 )

[0126] where x i，pos (t) represents the current position of device i at time t; x k，pos (t) represents the current position of device k at time t; ||x i，pos (t) - x k,pos (t)|| is the Euclidean distance between the current positions of the two devices; α is a weight factor used to adjust the influence degree of the conflict risk; exp(·) represents the exponential function.

[0127] The output of the path conflict detector is a conflict risk matrix among all devices, and each element I of the matrixij Indicates the path conflict risk between device i and device k.

[0128] In a possible implementation, the collaborative optimization unit introduces differential game theory and solves the path allocation problem among multiple devices through a game model. Each device i is regarded as a player in the game, and its goal is to minimize its own path and task costs in the presence of other devices. Specifically, the objective function of each device is:

[0129]

[0130] Among them, C i，time (x i , u i ) is the time cost of device i, which depends on the distance between the current position and the target position; C i，energy (x i , u i ) is the energy consumption cost of device i, which is usually a function of the square of the speed; I ik (x i , x k ) is the conflict risk between the paths of device i and device k. The possibility of conflict is determined according to the relative positions x i and x k and is calculated by the foregoing formula.

[0131] Generally, the collaborative optimization unit determines the optimal paths of each device by solving the Nash equilibrium. The conditions for the Nash equilibrium are:

[0132]

[0133] Among them, represents the optimal control strategy of device i at time t, represents minimizing the subsequent objective function H i by choosing the control input u i , H i is the Hamiltonian function of device i, p i is the co-state variable, which is used to describe the sensitivity of the device state to the objective function; u i represents the control input of device i; C i，time (x i , u i ) represents the time cost of device i; C i，energy (x i , u i ) represents the energy consumption cost of device i; I represents the conflict risk between the paths of device i and device k; represents the transposed form of the co-state variable; f i (x i, u i , x -i ) is the equipment state evolution equation provided by the path planning module; x -i represents the state set of all other equipment except equipment i.

[0134] The real-time dynamic adjustment module of the present invention aims to solve the problem of unpredictable changes in equipment state or external environment in the mine environment, ensure that the system can quickly adapt to environmental dynamics, equipment failures or other emergencies, and optimize task scheduling and path planning schemes. By setting up this module, the dynamic response ability of the entire mine scraper scheduling and path planning system can be significantly improved, enabling it to still achieve efficient operation under complex operating conditions. The real-time dynamic adjustment module relies on the dynamic data provided by the environmental perception module and interacts with the task scheduling module and the path planning module at the same time to correct the task and path planning schemes in real time.

[0135] In practical applications, unpredictability is the norm in mine operations. For example, problems such as equipment energy consumption exceeding expectations, roads being suddenly blocked, or task execution progress lagging behind. These situations require the system to quickly adjust the original task allocation and path planning. To meet these requirements, this module adopts a combined design of equipment state monitoring, environmental change perception, and dynamic optimization methods. By updating the state evolution equation and recalculating the dynamic optimization target, the rationality and efficiency of equipment scheduling and path planning are ensured.

[0136] In this embodiment, the real-time dynamic adjustment module mainly includes two core units: a dynamic monitoring unit and a dynamic adjustment unit. The following describes the technical implementation methods of the two units in detail: The dynamic monitoring unit is used to collect the operating state of the equipment, environmental state parameters, and operation progress information in real time to form the basic data source required for adjustment.

[0137] In some embodiments, the dynamic monitoring unit obtains the equipment state x i (t), where the equipment state includes the following three components:

[0138] Current position x i，pos (t), which is used to describe the precise position of the equipment in three-dimensional space.

[0139] Task completion degree x i，task (t), which is used to monitor the execution progress of the current equipment task, and the value range is [0,1].

[0140] Remaining energy x i，ebergy (t), which is used to evaluate whether the equipment can complete the current task.

[0141] Generally, the dynamic monitoring unit will also receive the external dynamic environmental state e(t) from the environmental perception module. These parameters include:

[0142] The road traffic conditions may restrict the progress of the equipment due to congestion, landslides or other factors.

[0143] The position of dynamic obstacles, such as the movement trajectories of other equipment or temporarily emerging obstacles.

[0144] Weather conditions, such as rain or snow, may increase the risk of equipment sliding.

[0145] As an option, the dynamic monitoring unit can set a timing trigger or an event-driven mechanism. When a specific event is detected (such as the remaining energy of the equipment being lower than the threshold, the task execution progress lagging behind or the road being blocked), the dynamic adjustment unit is immediately triggered to re-optimize the task scheduling and path planning.

[0146] After receiving the adjustment trigger information transmitted by the dynamic monitoring unit, the dynamic adjustment unit starts a real-time optimization process to correct the task scheduling and path planning scheme.

[0147] Specifically, the dynamic adjustment unit recalculates the state of the equipment based on the following dynamic evolution equation:

[0148]

[0149] where is the rate of change of the state of equipment i at time t; u i (t) is the control variable, including the path planning control u i，path (t) and the task allocation control u i，task (t); x -i (t) represents the state of other equipment; e(t) is the external dynamic environment state parameter.

[0150] The dynamic evolution function f i of the equipment state is specifically expressed as:

[0151]

[0152] where f i (x i (t), u i (t), x -i (t), e(t)) represents the state dynamic function of equipment i; x i (t) represents the state variable of equipment i; u i (t) represents the control variable of equipment i; x -i (t) represents the set of states of all other equipment except equipment i; e(t) represents the external dynamic environment state parameter; v i (t) is the speed of the equipment, which is determined by the path planning control variable u i，path (t); θ i(t) is the driving direction of the device; β is the task completion efficiency constant, representing the increment of task completion per unit time; u i，task (t) represents the task control variable; h i (u i，path , x i，energy ) is the energy consumption function of the device; γ 1 and γ 2 are constants of the device energy consumption model.

[0153] In a possible implementation, the dynamic adjustment unit uses the reinforcement learning method to dynamically optimize the above evolution equation, and adapts to environmental changes by updating the control variable in real time.

[0154] The global optimization control module is the core module of this system, which is used to coordinate the operation of each module and achieve the global optimum of task scheduling and path planning. It generates the scheduling and path planning schemes of multiple devices through the unified optimization of the device state, environmental state and task objectives. As the coordination center, the design focus of this module is on the completion efficiency of task objectives, the rationality of path planning and the optimal utilization of device resources. Generally, the system will dynamically collect environmental data and device states, and jointly solve the scheduling and path through the global optimization algorithm to ensure the collaborative and efficient operation of the whole system in a complex dynamic environment.

[0155] In some embodiments, this module is mainly responsible for integrating the inputs of each sub-module in a complex dynamic environment, including the environmental parameters of the environmental perception module, the task assignment results generated by the task scheduling module, and the preliminary path optimization scheme generated by the path planning module. It converts all inputs into a global optimization problem and solves it through a unified objective function. Specifically, under dynamic task requirements, this module should not only ensure the global consistency of task scheduling and path planning, but also meet the real-time requirements of the system.

[0156] In this embodiment, the global optimization control module takes the dynamic optimization model as the core, constructs a global objective function for the task scheduling and path planning problems, and solves the optimal task assignment and path planning schemes of the device. The objective function consists of multiple sub-objectives, including the weighted sum of task completion time, device energy consumption and path conflict. The core problem of optimization is expressed as:

[0157]

[0158] Among them, J is the global objective function, which is used to describe the comprehensive optimization objectives of all devices; α 1 is the weight factor of the task time cost, which is used to balance the relationship between the time required to complete the task and other optimization objectives, α 2 is the weight factor of the task energy consumption, which is used to balance the relationship between the energy consumed when completing the task and other objectives, α3 is the weight factor for the path conflict risk, used to balance the relationship between the path conflict risk and other optimization objectives; T i (x i , u i ) is the task completion time of device i, related to the path length and speed; E i (x i , u i ) is the energy consumption model of device i, related to the path planning and the speed control of the device; I ik (x i , x k ) is the path conflict cost between device i and device k, and the possibility of conflict is determined according to the relative positions x i and x k of device i and k, reflecting the path coordination between devices; N represents the total number of devices; T represents the planning time range of the optimization objective; t 0 represents the starting time of the optimization; represents the time accumulation of the device cost within the planning time interval (from t 0 to T).

[0159] Generally, the path conflict cost will significantly affect the scheduling and planning of devices. Therefore, the global optimization control module will dynamically adjust the path planning priorities of each device according to the risk value of the path conflict.

[0160] In some embodiments, the optimization problem is solved by a multi-objective optimization algorithm, and the concept of Pareto optimal solution is preferably adopted. To speed up the solution, the system adopts a distributed optimization strategy. In the distributed optimization strategy, the cooperation between devices is realized through a multi-agent optimization model. Specifically, the optimization result of device i will affect the calculation of the path conflict cost of device j, and at the same time, the optimization information is shared through V2V communication.

[0161] As an option, the system can also optimize the dynamic parameters through reinforcement learning, such as the dynamic adjustment of the weight factor α 1 , α 2 , α 3 . Specifically, the reinforcement learning model can be trained to perform weighted correction on the optimization objective according to the historical task completion time, energy consumption data, and conflict occurrence probability.

[0162] To ensure the robustness of the global optimization result in a dynamic environment, the optimization result is also verified through the Lyapunov stability theory. Generally, the system constructs the following Lyapunov function:

[0163]

[0164] where \(V(x)\) represents the value of the Lyapunov function; \(x\) i is the target state of device \(i\); is the current state of device \(i\); \(N\) represents the total number of devices; \(x\) i is the current state of device \(i\); is the target state of device \(i\).

[0165] If then the optimization scheme meets the system stability requirements.

[0166] In a possible implementation, the global optimization control module also supports interface docking with other external systems, such as a mine operation management system or an energy management platform, for real-time updating of external constraint conditions. In terms of scalability, this module can be compatible with more types of devices, such as automated transport trucks, driverless mining vehicles, etc., to achieve collaborative optimization of full-mine intelligent scheduling.

[0167] As Figure 2 shown, the present invention also provides a method for intelligent scheduling and path planning of a mine loader, including the following steps:

[0168] S1. The environmental perception module is used to collect environmental data of the mine operation area in real time, and a three-dimensional map is constructed by using lidar, GPS, and cameras to generate environmental state parameters;

[0169] Step S1 realizes real-time monitoring of the mine operation area through the environmental perception module, and provides environmental state data required for task scheduling and path planning. The lidar is used to generate a high-precision three-dimensional map of the mine operation area, accurately identify road boundaries, obstacles, and dynamic environmental features; the GPS module provides real-time accurate position information of the device for path planning and monitoring of the device trajectory; the camera captures road dynamic information, such as dynamic obstacles or subsidence areas, and provides real-time support for path planning. Through data fusion and processing, multi-source data of lidar, GPS, and cameras are integrated to generate environmental state parameters for subsequent task scheduling and path planning;

[0170] S2. The task scheduling module allocates operation tasks based on device status, task objectives, and environmental state parameters, and the global optimization unit optimizes the task allocation scheme, with the optimization objective being the weighted sum of task completion time and energy consumption;

[0171] In step S2, the task scheduling module assigns task objectives to the mine loaders and other equipment to ensure the rationality and efficiency of task allocation. The task scheduling module dynamically generates a task allocation plan based on the current positions, remaining energy, and task requirements of the equipment, in combination with the environmental status parameters. The task scheduling plan is optimized by the global optimization unit with the optimization objective of minimizing the task completion time and energy consumption. The optimization process takes into account the load balance of the equipment and the task priorities simultaneously to ensure the overall efficiency and resource utilization rate;

[0172] S3. Through the path planning module, according to the operation tasks assigned by the task scheduling module and the environmental status parameters, plan the optimal path for each device to reach the task objective, and solve the path conflicts through multi-device game optimization to ensure the collaborative optimality of path planning;

[0173] In step S3, the path planning module plans the optimal path for each device to enable the device to efficiently reach the assigned task objective from the current position. In path planning, based on the task allocation plan and environmental status parameters, while designing the shortest path, the dynamic conditions of the road and the energy consumption limit of the device are considered. At the same time, the path conflict problem is solved through multi-device game optimization. In multi-device collaborative operation, the collaborative optimality of path planning is ensured through optimization strategies. This module also dynamically adjusts the path plan to cope with sudden changes in the environment and potential interference between multiple devices;

[0174] S4. Through the multi-device collaboration module, detect the path intersection points between devices, calculate the path conflict risk value, and adjust the device priorities of the conflicting paths through the collaboration optimization unit to dynamically update the path control strategy of the path planning module;

[0175] In step S4, the multi-device collaboration module detects the path conflict situation and dynamically adjusts the priorities between devices. By analyzing the path intersection points of the devices, calculate the risk level of the path conflict, and divide the operation priorities of the devices according to the risk level. The collaboration optimization unit adjusts the path planning strategies of each device according to the priorities to enable the operations of the conflicting devices to transition smoothly. At the same time, by dynamically updating the path control strategy, ensure the efficient collaborative operation between multiple devices and avoid delays or resource waste caused by path conflicts;

[0176] S5. Through the real-time dynamic adjustment module, when the environmental status or device status changes, monitor the current position, remaining energy, and task completion degree of the device, and dynamically adjust the task scheduling plan and path planning plan;

[0177] Specifically for step S5, when the environmental state or device state changes, the real-time dynamic adjustment module ensures the robustness and flexibility of the system operation. This module monitors the current position, remaining energy, and task completion degree of the device in real time, and adjusts the task allocation and path planning scheme in a timely manner through a dynamic feedback mechanism. For example, when the device has insufficient energy consumption, it preferentially adjusts to a low-energy consumption path; when the road conditions change, it replans the path in real time to ensure the achievement of the task goal. This module is particularly important in dealing with emergencies (such as device failures or road blockages) and can significantly improve the adaptability of the system

[0178] S6. Through the global optimization control module, the global objective function of task scheduling and path planning is optimized based on the variational method to achieve intelligent scheduling and path planning for mine operations;

[0179] Regarding step S6, the global optimization control module is the core of the entire method, responsible for coordinating the work of all modules to ensure the global optimality of task scheduling and path planning. By optimizing multi-objective factors such as task completion time, energy consumption, and path conflicts, the intelligent operation of the overall system is achieved. This module combines dynamic environmental changes, uniformly schedules devices, and optimizes the path scheme, and distributes the optimization results to each device for execution to ensure the efficient and stable completion of mine operations in a complex environment.

[0180] It should be understood that the parts not elaborated in detail in this specification all belong to the prior art. The above embodiments merely describe the preferred implementation manners of the present invention and do not limit the scope of the present invention. Without departing from the design spirit of the present invention, various deformations and improvements made by those of ordinary skill in the art to the technical solution of the present invention shall fall within the protection scope determined by the claims of the present invention.

Claims

1. An intelligent scheduling and path planning system for mine loader, characterized in that: It includes: The environmental perception module is used to collect environmental data of the mine operation area in real time, including equipment status, road conditions and dynamic obstacle locations, and process the collected data to generate environmental status parameters; The task scheduling module is used to allocate job tasks according to equipment status, task objectives and environmental status parameters, and globally optimize task allocation to minimize task completion time and energy consumption; The path planning module is used to plan the optimal path of the equipment based on the job tasks assigned by the task scheduling module and the environmental state parameters provided by the environmental perception module; Multi-device collaboration module, used to coordinate the competition and collaboration of multiple devices in path planning and resolve path conflicts between devices; Real-time dynamic adjustment module, used to dynamically adjust task scheduling and path planning when the environment status or equipment status changes; The global optimization control module is used to coordinate the work of the environment perception module, task scheduling module, path planning module, multi-device collaboration module and real-time dynamic adjustment module, and optimize the global scheduling and path planning solutions.

2. The intelligent scheduling and path planning system for a mine loader according to claim 1, characterized in that: The environment perception module comprises: LiDAR, which is used to build three-dimensional maps of the mining area; GPS, used to obtain the precise location of the device; Cameras to capture images of mine roads, obstacles and working environment; The data processing unit is used to fuse the raw data collected by the lidar, GPS and camera to generate environmental state parameters for path planning and task scheduling.

3. The intelligent scheduling and path planning system for a mine loader according to claim 1, characterized in that: The task scheduling module includes: A task distributor, used to distribute global task targets to multiple devices; The global optimization unit is used to optimize the task allocation plan to minimize the task completion time and equipment energy consumption, where the optimization target is the weighted sum of the task completion time and energy consumption, and generates a task scheduling plan based on the status of all devices and the task objectives.

4. The intelligent scheduling and path planning system for a mine loader according to claim 1, characterized in that: The task allocation optimization of the global optimization unit is based on the following constraints: The distance between the device’s current location and the target task is the shortest; The remaining energy of each device meets the task execution requirements; Each task is performed by a unique device.

5. The intelligent scheduling and path planning system for a mine loader according to claim 1, characterized in that: The path planning module includes: A single device path optimization unit, for calculating an optimal path to a target location based on the mission objective and current location of the device, wherein the path optimization takes into account the device speed, steering angle, and obstacles in the environment; The multi-device game optimization unit is used to resolve multi-device path conflicts through a game model and optimize multi-device path solutions to ensure the optimal coordination of all device paths.

6. The intelligent scheduling and path planning system for mining scrapers according to claim 5, characterized in that: The multi-device game optimization unit constructs a path competition model between devices based on differential game theory, and solves the optimal path of the device through the Nash equilibrium of path optimization.

7. The intelligent scheduling and path planning system for a mine loader according to claim 1, characterized in that: The multi-device collaboration module includes: A path conflict detector is used to detect path intersections between devices and calculate the risk value of path conflicts; The collaborative optimization unit is used to adjust the device priorities of conflicting paths and dynamically update the control strategy of the path planning module.

8. The intelligent scheduling and path planning system for a mine loader according to claim 1, characterized in that: The real-time dynamic adjustment module includes: Dynamic monitoring unit, used to monitor the current location of the equipment, remaining energy, task completion and real-time environmental status; The dynamic adjustment unit is used to re-optimize the task scheduling plan and path planning plan according to the status information provided by the dynamic monitoring unit.

9. The intelligent scheduling and path planning system for mine loader according to claim 1, characterized in that: The global optimization control module uses the variational method to optimize the global objective function of task scheduling and path planning. The global objective function includes the weighted sum of task completion time, equipment energy consumption and path conflict, and determines the global optimal scheduling and path planning scheme by solving the Euler-Lagrange equation.

10. A method for intelligent scheduling and path planning of a mine loader, applied to an intelligent scheduling and path planning system for a mine loader as claimed in any one of claims 1 to 9, characterized in that: The following steps are involved: S1. Collect environmental data of the mining operation area in real time through the environmental perception module, build a three-dimensional map using laser radar, GPS and camera, and generate environmental status parameters; S2, assigning job tasks based on equipment status, task objectives and environmental status parameters through the task scheduling module, and optimizing the task assignment scheme through the global optimization unit, with the optimization target being the weighted sum of task completion time and energy consumption; S3. Through the path planning module, according to the job tasks assigned by the task scheduling module and the environmental state parameters, the optimal path to the task target is planned for each device, and the path conflict is solved through multi-device game optimization to ensure the coordinated optimization of path planning; S4, through the multi-device collaboration module, detect the path intersections between devices, calculate the path conflict risk value, and adjust the device priority of the conflicting path through the collaborative optimization unit, and dynamically update the path control strategy of the path planning module; S5. Through the real-time dynamic adjustment module, when the environmental state or device state changes, the current position, remaining energy and task completion of the device are monitored, and the task scheduling plan and path planning plan are dynamically adjusted; S6. Through the global optimization control module, the global objective function of task scheduling and path planning is optimized based on the variational method to realize intelligent scheduling and path planning of mining operations.

Images