Mine truck unloading method and system

Abstract

The present application discloses a method and system for unloading a mine truck, which relates to the field of unmanned driving, including: collecting chute status data and underground mining traffic data; using a task matching algorithm to generate the optimal unloading task of the corresponding mine truck; generating the optimal planned path from the current position to the target chute through a path planning algorithm; calculating the position deviation of the target mine truck and the unloading port of the target chute according to the position of the target mine truck and the target chute, and generating a control instruction according to the position deviation, and controlling the alignment of the target mine truck and the unloading port according to the control instruction; in view of the low unloading efficiency of mine trucks in underground mining in the prior art, the present application uses a task matching algorithm to generate the optimal unloading task for the mine truck. At the same time, the path planning algorithm is used to plan the optimal driving path from the current position to the target chute for the mine truck, thereby improving the driving efficiency of the mine truck.

Description

Technical Field

[0001] The present application relates to the field of unmanned driving, and in particular to a mining truck unloading method and system. Background Art

[0002] Underground mining is a common mining method. Compared with open-pit mining, underground mining is limited by terrain and mining space, which puts higher requirements on ore transportation. In underground mines, after the ore is mined from the mining area, it is usually transported to the chute by mining trucks, and then unloaded into the ore bin or transferred to the concentrator through the chute. The unloading efficiency of mining trucks directly affects the overall transportation efficiency and production capacity of the mine.

[0003] However, the current mining trucks generally have problems of low efficiency and insufficient intelligence in the underground unloading process. On the one hand, due to the narrow underground roads, complex environment, and scattered unloading points, it is difficult for mining truck drivers to quickly find a suitable unloading chute and complete the unloading action. Transportation delays often occur due to waiting in line or taking a detour to find an unloading point. On the other hand, there is a lack of unified scheduling for the unloading of mining trucks, and the allocation of chutes is very arbitrary, which can easily cause some chutes to be overloaded while others are underutilized, resulting in low unloading efficiency. At the same time, due to manual experience, safety accidents such as collisions and overloading of mining trucks during the unloading process often occur.

[0004] In order to improve the transportation efficiency of underground mine trucks and reduce empty driving and waiting time, some mining companies have tried to optimize the unloading process of mine trucks. Common practices include: rationally planning transportation routes, adding unloading points, introducing unloading management systems, etc. These measures have alleviated the efficiency bottleneck of the "last mile" of unloading to a certain extent, but there are still shortcomings: first, there is a lack of real-time perception means, and it is impossible to obtain the real-time status of the chute and the road. The scheduling decision relies on static planning and manual experience; second, there is a lack of effective information interaction between the mine truck and the scheduling system, and the navigation instructions and path planning cannot be accurately delivered to the vehicle end; third, the unloading link is not highly automated, and there is a lack of precise guidance and control means for the unloading action of the mine truck. The above problems restrict the overall efficiency improvement of the mine transportation system.

[0005] In recent years, the new generation of information and communication technologies represented by the Internet of Vehicles and the Internet of Things have flourished, bringing new technical support for the intelligent upgrade of mining areas. V2X (Vehicle to Everything) communication technology can realize direct communication and information interaction between vehicles and roads, and vehicles and infrastructure, and roadside sensing equipment can perform real-time detection and environmental perception of vehicles and road targets. Based on V2X communication and roadside perception, building a comprehensive interconnection and information integration of vehicles, roads, environments, and management platforms can provide real-time, accurate, and comprehensive data support for mining truck scheduling decisions, and provide strong technical support for vehicle navigation control, thereby breaking through the bottleneck of traditional unloading mode and realizing the intelligent reconstruction of the entire process of mining truck unloading. Summary of the invention

[0006] In view of the problem of low unloading efficiency of mine trucks in underground mines in the prior art, this application provides a mine truck unloading method and system, which collects chute status data and underground mine traffic data through roadside sensing equipment, combines the mine truck status data, and uses a task matching algorithm to generate the optimal unloading task for the mine truck, guiding the mine truck to the most suitable chute for unloading. At the same time, the path planning algorithm is used to plan the optimal driving path from the current position to the target chute for the mine truck, thereby improving the driving efficiency of the mine truck.

[0007] The purpose of this application is achieved through the following technical solutions.

[0008] One aspect of the present application provides a mine truck unloading method, comprising: S1, collecting chute status data and underground mine traffic data through roadside sensing equipment, and uploading the collected data to a central control system; S2, after the mine truck is loaded with ore, it sends an unloading task request to the central control system through a V2X communication module; S3, the central control system generates an optimal unloading task for the corresponding mine truck based on the chute status data, underground mine traffic data and mine truck status data using a task matching algorithm, and sends the optimal unloading task to a target mine truck and a target chute; S4, the target mine truck generates an optimal planned path from a current position to a target chute through a path planning algorithm; S5, the target mine truck drives to the target chute according to the optimal planned path, the central control system calculates the position deviation of the target mine truck and the unloading port of the target chute according to the positions of the target mine truck and the target chute, generates a control instruction based on the position deviation, and controls the target mine truck to align with the unloading port according to the control instruction; S6, after the target mine truck completes the alignment, it unloads the ore into the target chute.

[0009] Furthermore, the chute status data includes the chute workload, working status and expected unloading cycle; the mine traffic data includes traffic flow and traffic congestion;

[0010] Furthermore, the unloading task request includes the mine truck status data, which includes the location, speed and load of the mine truck.

[0011] Further, S3, the central control system uses a task matching algorithm to generate the optimal unloading task of the corresponding mine truck according to the chute status data, the underground mine traffic data and the mine truck status data, and sends the optimal unloading task to the target mine truck and the target chute, including: S31, according to the chute status data, the underground mine traffic data and the mine truck status data, a weighted scoring method is used to determine the priority of each unloading task; S32, the Hungarian algorithm is used to establish a task allocation matrix between the mine truck and the chute, and the elements in the task allocation matrix are the priority of the unloading task, as well as the estimated driving time of the mine truck and the estimated working time of the chute; S33, the Hungarian algorithm is used to solve the task allocation matrix to obtain the task allocation matrix containing each The initial unloading task matching result of the target chute and the expected arrival time of the mining truck; S34, obtaining the chute working status in the chute status data. When the chute working status indicates that the corresponding chute has a fault, cancel the unloading task assigned to the chute, and re-incorporate the unloading task into the task allocation matrix, and use the Hungarian algorithm to solve the updated task allocation matrix to obtain the adjusted unloading task matching result; S35, compare the initial unloading task matching result and the adjusted unloading task matching result, and select the optimal unloading task matching result; S36, according to the optimal unloading task matching result, extract the target chute number and expected arrival time corresponding to each mining truck, and generate the optimal unloading task for each mining truck.

[0012] S33, according to the task allocation matrix established in S32, the mine trucks are regarded as rows of the matrix, and the chutes are regarded as columns of the matrix. The matrix elements are triplets consisting of the unloading task priority score between the corresponding mine trucks and the chutes, the estimated travel time of the mine trucks, and the estimated working time of the chutes; the Hungarian algorithm is applied to the task allocation matrix to perform matrix zero operations: the minimum element value of each row of the matrix is ​​subtracted from the minimum element value of the row to obtain a row zero matrix; the minimum element value of each column of the row zero matrix is ​​subtracted from the minimum element value of the column to obtain a total zero matrix; unmarked zero elements are searched in the total zero matrix: if there are unmarked zero elements, the row and column where the zero element is located are marked, indicating that the corresponding mine truck is temporarily assigned to the corresponding chute; if all zero elements have been marked, the initial unloading task matching result is obtained according to the marked zero elements, including the target chute number and estimated arrival time of each mine truck, wherein the estimated arrival time of the mine truck is determined according to the estimated travel time of the mine truck between the corresponding mine truck and the chute in the task allocation matrix.

[0013] If there are unmarked mine trucks or chutes, continue to look for marked zero elements; if there are unmarked mine trucks or chutes, it means that the currently marked zero elements cannot completely cover the task allocation matrix, and adjust the unmarked row and column elements: find the minimum value among the unmarked elements, recorded as δ; subtract δ from all unmarked row elements and add δ to all marked column elements to obtain the adjusted total zero matrix; return to continue looking for unmarked zero elements; based on the marked zero elements, obtain the initial unloading task matching results, including the target chute number and estimated arrival time of each mine truck, where the estimated arrival time of the mine truck is determined based on the estimated travel time of the mine truck between the corresponding mine truck and the chute in the task allocation matrix.

[0014] This application encapsulates the three factors of unloading priority, driving time, and working time into triplets as elements of the task allocation matrix, so that the matching results take into account both task importance and time cost; through the subtraction operation of matrix row and column elements, the task allocation matrix is ​​converted into an equivalent total zero matrix, which is convenient for finding the optimal match; an iterative strategy of marking unassigned zero elements and adjusting the minimum element of uncovered rows and columns is adopted to gradually expand the matching set until all zero elements are marked, and a complete matching solution is obtained; the matching result includes the target chute number and estimated arrival time of each mining truck, and clearly gives a one-to-one corresponding unloading task allocation solution. The estimated arrival time is directly derived from the task allocation matrix, which ensures the time optimality of vehicle scheduling.

[0015] Further, S32, the Hungarian algorithm is used to establish a task allocation matrix between the mine truck and the chute, the elements in the task allocation matrix are the priority of the unloading task, as well as the estimated driving time of the mine truck and the estimated working time of the chute, including: according to the priority of each unloading task, a priority matrix between the mine truck and the chute is established, the rows of the priority matrix represent the mine truck, the columns represent the chute, and the elements in the priority matrix are the priorities of the unloading tasks between the corresponding mine truck and the chute; according to the position and speed of the mine truck in the mine truck status data, the estimated driving time of each mine truck to each chute is calculated, and a driving time matrix between the mine truck and the chute is established, the rows of the driving time matrix represent the mine truck, the columns represent the chute, and the elements of the driving time matrix is the estimated driving time from the corresponding mine truck to the corresponding chute; according to the estimated unloading cycle of the chute in the chute status data, the estimated working time of each chute is obtained, and a chute working time matrix is ​​established. The rows and columns of the chute working time matrix represent the chutes, and the diagonal elements in the chute working time matrix are the estimated working time of the corresponding chutes; the priority matrix, the driving time matrix and the chute working time matrix are normalized; according to the normalized priority matrix, the driving time matrix and the chute working time matrix, a task allocation matrix is ​​established, the rows in the task allocation matrix represent mine trucks, the columns represent chutes, and the elements are triples, which contain the priority of the unloading task, as well as the estimated driving time of the mine truck and the estimated working time of the chute.

[0016] Further, S35, compares the initial unloading task matching result and the adjusted unloading task matching result, and selects the optimal unloading task matching result, including: respectively calculating the overall priority scores of the initial unloading task matching result and the adjusted unloading task matching result: the overall priority score = Σ (the priority score of each unloading task), respectively calculating the overall estimated driving time of the initial unloading task matching result and the adjusted unloading task matching result: the overall estimated driving time of the mine truck = Σ (the estimated driving time of each mine truck), comparing the overall priority scores of the initial unloading task matching result and the adjusted unloading task matching result, and taking the unloading task matching result with a higher overall priority score as the candidate optimal matching result; if the overall priority scores of the two are the same, continue to compare the overall estimated driving time of the mine truck, and take the unloading task matching result with a shorter overall estimated driving time of the mine truck as the optimal unloading task matching result.

[0017] Further, S4, the target mine truck generates an optimal planned path from the current position to the target chute through a path planning algorithm, including: S41, the target mine truck extracts the position coordinates of the target chute in the optimal unloading task; S42, the current position of the target mine truck is used as the starting point, and the position coordinates of the target chute are used as the target point; S43, the underground mine map is obtained, and the candidate path from the starting point to the target point is searched on the underground mine map through the A* algorithm; S44, according to the traffic flow and traffic congestion in the underground mine traffic data, the travel time and travel cost of each candidate path are calculated; S45, according to the travel time and travel cost of each candidate path, a weighted scoring method is used to select the candidate path with the highest score as the optimal planned path;

[0018] Further, S5, generates a control instruction according to the position deviation, and controls the target mine truck to align with the unloading port according to the control instruction, including: S51, the target mine truck travels according to the optimal planned path, and uploads its own position and speed to the central control system through the V2X communication module; S52, the target chute obtains the matching target mine truck according to the optimal unloading task, and uploads the unloading port position of the target chute to the central control system through the roadside sensing device; S53, the central control system uses the Kalman filter algorithm to integrate the position, speed and posture of the target mine truck, as well as the unloading port position of the target chute, and calculates the lateral deviation, longitudinal deviation and posture deviation of the target mine truck relative to the unloading port of the target chute; S54, generates an adjustment instruction according to the lateral deviation, longitudinal deviation and posture deviation using the model predictive control algorithm, and the adjustment instruction includes adjusting the direction, adjusting the distance and adjusting the posture.

[0019] Another aspect of the present application also provides a mine truck unloading system based on V2X communication and roadside sensing equipment, which is used to execute a mine truck unloading method of the present application.

[0020] Compared with the prior art, the advantages of this application are:

[0021] This application realizes real-time perception and information interaction of chute status, traffic conditions and mine truck status by deploying roadside sensing equipment and V2X communication network in the mining area. The central control system uses the collected data to intelligently generate mine truck unloading tasks through the task matching algorithm, and guides the mine truck to the optimal chute for unloading. This avoids inefficient behaviors such as drivers blindly looking for unloading points and over-concentration, thereby reducing unloading waiting time and improving unloading efficiency. At the same time, the task matching algorithm fully considers dynamic factors such as chute workload and expected unloading cycle, which can achieve chute unloading load balancing and further improve unloading efficiency and chute utilization. In addition, by combining real-time perception data for task matching, unloading scheduling is transformed from static planning to dynamic optimization, and the intelligence level of the unloading process is greatly improved.

[0022] After the unloading task of the mine truck is determined, this application uses the path planning algorithm to plan an optimal path for the mine truck to the target chute. The path planning process fully considers factors such as the underground road network topology, real-time traffic flow and traffic events. The mine truck can effectively avoid congested sections and travel on the path with the best road conditions and the shortest time. At the same time, the positional relationship between the mine truck and the target chute has been considered during task matching, so the empty driving distance of the mine truck from the loading point to the unloading point is also greatly shortened.

[0023] This solution uses V2X communication and roadside sensing equipment to achieve precise positioning of the mine truck and the chute unloading port. When the mine truck approaches the chute, the central control system calculates the position deviation between the mine truck and the unloading port, quantifies the deviation into lateral deviation and longitudinal deviation. When the deviation exceeds the safety threshold, it generates adjustment instructions in time and sends them to the mine truck to guide it to accurately align. The adjustment instructions give clear adjustment directions and distances, which help the mine truck to quickly complete the alignment action. The entire alignment process is carried out automatically under the command of the central control system, avoiding the risk of collision caused by the driver's lack of experience and significantly improving the safety of unloading.

[0024] The task allocation algorithm of this application introduces a priority evaluation mechanism, which assigns different priorities to unloading tasks according to the grade and process requirements of different ores. In the task allocation process, the system prioritizes high-priority ores to idle chutes, ensuring the priority unloading of key materials and reducing logistics delays from the mine site to the unloading point. At the same time, through dynamic matching of task priorities, the flexibility of ore flow organization is improved, the balance of ore grades can be achieved, and material accumulation at the unloading point can be reduced, thereby meeting the balanced feeding requirements of mineral processing operations.

[0025] This solution monitors the status of the chute in real time. When a chute failure is found, the system can automatically cancel the unloading task assigned to the faulty chute and assign the affected tasks to other available chutes by recalculating the task allocation matrix. The dynamic adjustment of the unloading task effectively avoids the unloading interruption problem caused by the chute failure, localizes the impact of the failure, and improves the fault tolerance and continuity of the unloading operation. In addition, the unloading tasks and driving paths generated by the system are all based on the results of global information optimization. The reliability of navigation instructions is much higher than the driver's personal experience judgment, thereby improving the overall reliability of mine truck dispatching and command. BRIEF DESCRIPTION OF THE DRAWINGS

[0026] The present application will be further described in the form of exemplary embodiments, which will be described in detail by the accompanying drawings. These embodiments are not restrictive, and in these embodiments, the same number represents the same structure, wherein:

[0027] Figure 1is a schematic diagram of an exemplary application scenario of a mine truck unloading method according to some embodiments of the present application;

[0028] Figure 2 is an exemplary flow chart of a mine truck unloading method according to some embodiments of the present application. DETAILED DESCRIPTION

[0029] The method and system provided in the embodiments of the present application are described in detail below with reference to the accompanying drawings.

[0030] like Figure 1 and Figure 2 As shown, this application uses V2X communication and roadside sensing equipment to build a comprehensive information collection and transmission system. Mine truck status data: The mine truck is equipped with a V2X communication module, which collects the position, speed, load and other status data of the mine truck in real time through the on-board sensor, and transmits the data to the roadside unit (RSU) through the V2X module or directly uploads it to the central control system. Position data can be obtained by GPS positioning or relative positioning with the RSU; speed data can be measured by wheel speed sensor; load data can be obtained by cargo box pressure sensor or cargo weight sensor. In addition, the V2X module can also calculate the expected arrival time of the mine truck to the destination as a reference for scheduling decisions. The low latency characteristics of V2X ensure the real-time nature of status data.

[0031] Chute status data: Roadside sensing equipment is installed at the entrance of each chute to monitor the status of the chute in real time. The sensors equipped with the sensing equipment can identify the idle / occupied status of the chute and detect the current workload of the chute, such as the amount of ore in the chute, the status of the unloading equipment, etc. The sensor can also estimate the expected available time of the chute based on the unloading capacity of the chute and the current amount of ore, that is, the time when a new mining truck is expected to enter for unloading. When a chute fails, the sensor can obtain the fault information in time and report it. The chute status data is uploaded to the central control system through the communication module of the road test equipment.

[0032] Traffic status data: Traffic flow monitoring sensors are installed in the main transportation lanes of underground mines to collect data such as vehicle flow and speed to determine the congestion of the road. For sections that are prone to bottlenecks, real-time images can also be obtained through video monitoring equipment to analyze traffic conditions. RSUs can be set up at intersection nodes to obtain traffic information of vehicles at intersections through V2X communication, and identify priority and yield relationships. The traffic event detection module can identify traffic accidents, abnormal parking, etc. All traffic data is uploaded to the central control system through wired or wireless communication networks for path planning and navigation and obstacle avoidance.

[0033] Data upload: Communication base stations and RSUs in the mining infrastructure are responsible for data upload. Through wireless communication technologies such as DSRC and LTE-V, the data collected by sensing equipment and mining trucks are transmitted to edge computing nodes and central control systems. The edge computing nodes located on site clean, fuse and preliminarily analyze the original sensing data, which can reduce the computing pressure of the central node on the one hand, and compress the data on the other hand to reduce the transmission bandwidth. The data processed by the edge node will be transmitted to the central system at a certain frequency according to the timeliness requirements of the data, and follow the deterministic network transmission protocol to strictly control the transmission delay.

[0034] On the basis of information collection and transmission, this application further conducts intelligent analysis on the acquired data such as the status of the mine truck, the status of the chute, and the traffic conditions, and performs unloading task matching and driving route planning based on this, so as to realize the dynamic optimization of the unloading scheduling of the mine truck. Mine truck status analysis: After the central control system receives the real-time position, speed, load and other status data uploaded by the mine truck, it first performs map matching on the location of the mine truck to determine the road section where it is located and the distance from the destination. Combining the real-time speed of the mine truck and the speed limit requirements of the road, the system can predict the estimated time (ETA) of the mine truck to arrive at the destination. At the same time, the system divides the mine truck into different states such as empty, half-loaded, and fully loaded according to the load capacity of the mine truck. For fully loaded or heavily loaded mine trucks, the system will give priority to assigning them to chutes with large loads to complete unloading as soon as possible, reduce the weight of the vehicle, and improve driving safety. In addition, the continuous working time of the mine truck can be comprehensively considered according to the oil and power status of the mine truck to avoid shutdown due to energy exhaustion.

[0035] Chute status analysis: The chute status analysis module receives the chute status data uploaded by the roadside sensing device and evaluates the working status of each chute in real time. By analyzing the occupied / idle status of the chute and combining the unloading capacity of the chute design, the system can determine which chutes are currently idle and have a large surplus of unloading capacity. For these chutes, the system marks them as priority unloading points and gives priority to them when allocating tasks. For chutes in the occupied state, the system will predict their estimated available time to provide a reference for the arrival time scheduling of mine trucks. When the chute monitoring data shows that the equipment is overloaded or there is a fault, the system can immediately remove the chute from the available unloading points and trigger a maintenance operation instruction. After the fault is eliminated, the system will restore the chute to an available state. Through real-time monitoring of the health status of the chute, the risk of equipment overload and unplanned downtime can be effectively avoided.

[0036] Traffic condition analysis: The traffic condition analysis module summarizes the data such as traffic volume, speed, and traffic events collected by road monitoring sensors and RSUs, and evaluates the real-time traffic conditions of the road network in the mining area. By comparing traffic volume with road capacity, the system can identify sections prone to congestion and predict the duration of congestion. At the same time, the system will also predict the distribution of traffic flow in different time periods based on historical data, and judge the traffic pressure that may occur during peak hours in advance. When traffic accidents, abnormal weather and other conditions are monitored, the system can update the road network status in real time and change the original path. The traffic analysis results can be directly fed back to the path planning module, so that it can actively avoid congested sections when generating driving routes, choose smooth and efficient routes, and guide mining trucks to pass during peak hours when necessary to relieve traffic pressure. In addition, the traffic analysis results can also be used for the dynamic evaluation of the unloading capacity of the chute. For example, when the road near the chute is not smooth, the unloading efficiency of the chute will be reduced accordingly, and its priority will be reduced when assigning tasks.

[0037] After obtaining real-time data such as the status of the mine truck, the status of the chute, and the traffic conditions, this application uses a task matching algorithm to dynamically generate and optimize unloading tasks, and assign the optimal task to the mine truck: S31. In the evaluation of the priority of unloading tasks, the system comprehensively considers multiple factors that affect the unloading efficiency and urgency, and finally calculates the comprehensive priority score of each unloading task by weighting and scoring each evaluation indicator.

[0038] Idleness of the chute: The idleness of the chute reflects the current availability of the chute. The higher the idleness, the faster the chute can accept new unloading tasks. The system divides the idleness into multiple levels according to the real-time occupancy status and estimated available time of the chute, such as completely idle, about to be idle, unloading, etc., and assigns different scores to different levels. The idleness weight can be set to w1. Unloading capacity of the chute: The unloading capacity reflects the unloading efficiency of the chute per unit time. The stronger the unloading capacity, the faster the unloading task can be completed. The system divides the unloading capacity into multiple gears according to the designed unloading speed of the chute, the current equipment status, etc., and assigns different scores to different gears. The unloading capacity weight can be set to w2. Cargo type of mine truck: Different types of ore may have different priorities. For example, high-grade ore or urgently needed special materials may need to be transported first. The system presets the priority level of the ore type and maps the cargo type carried by the mine truck to the corresponding priority score. The cargo type weight can be set to w3. Fullness of mine truck: The fullness reflects the carrying efficiency of the mine truck. The higher the fullness, the greater the transport capacity of the single vehicle, and the scheduling priority can be appropriately increased. According to the ratio of the mine truck load to the rated load, the system divides the full load into empty, half-loaded, and full-loaded levels, and assigns different scores. The full load weight can be set to w4. Estimated driving time of the mine truck: The driving time affects the efficiency of the mine truck in completing the unloading task. The shorter the driving time, the faster the transportation can be completed. The system estimates the estimated driving time of the mine truck based on the distance between the current position of the mine truck and the target chute and the average speed of the mine truck, and divides the time into multiple intervals, assigning different scores to different intervals. The driving time weight can be set to w5. Road smoothness: Road smoothness reflects the traffic status of the mine truck's driving route. The higher the smoothness, the higher the transportation efficiency. According to the real-time traffic flow, average driving speed, etc. of the road, the system divides the smoothness into multiple levels and assigns different scores to different levels. The smoothness weight can be set to w6.

[0039] Based on the above indicators, the system scores each unloading task to be assigned and calculates the weighted comprehensive score. Assume that the score of the i-th unloading task on each indicator is , then its comprehensive priority score It can be expressed as: ,in, arrive is the weight coefficient of each indicator, which must satisfy The weight coefficient can be set according to the actual situation of the mining area and scheduling preferences. Comprehensive priority score The higher the value, the higher the urgency and importance of the unloading task, and the higher the scheduling priority should be given. All the unloaded tasks to be assigned are sorted to form a priority queue, and the sorting result is used for subsequent task allocation optimization.

[0040] S32: Construction of the task allocation matrix. After determining the task priority, the system uses the Hungarian algorithm to establish the task allocation matrix between the mine truck and the chute. The task allocation matrix is ​​a two-dimensional matrix. The rows of the matrix represent the mine trucks, the columns represent the chute, and the matrix elements are a triple, including the priority of the unloading task between the mine truck and the chute, the estimated driving time of the mine truck, and the estimated working time of the chute. To construct the task allocation matrix, the system performs the following processing: According to the priorities of each unloading task obtained in S31, a mine truck-chute priority matrix is ​​established. This matrix represents the priority of the unloading task between each mine truck and each chute. According to the position and speed of the mine truck, the estimated driving time of each mine truck to each chute is estimated, and a mine truck-chute driving time matrix is ​​established. According to the estimated unloading cycle in the chute status data, the estimated working time of each chute is obtained, and a chute-chute working time matrix is ​​established. The diagonal elements of the matrix are the estimated working time of the chute. Due to the different dimensions of priority, driving time and working time, the system normalizes the above three matrices and unifies them to the range of 0-1. The three normalized matrices are fused into a task allocation matrix, where the matrix elements are triplets containing priority, travel time, and working time.

[0041] S33, in the initial task matching stage, the system uses the Hungarian algorithm to solve the task allocation matrix and obtain an initial unloading task matching solution. The Hungarian algorithm is a classic algorithm for dealing with the weighted maximum matching problem of bipartite graphs, which can find the global optimal matching in multiple cost matrices. Applying this algorithm to the unloading task allocation of mining trucks can obtain the matching result with the best global scheduling efficiency while taking into account multiple factors.

[0042] The Hungarian algorithm requires a non-negative integer matrix as input, so the task allocation matrix G obtained in step S32 needs to be transformed. Assume that the size of matrix G is N*M, where N is the number of mining trucks to be assigned and M is the number of available chutes. Define the transformed matrix as G'. For any element in matrix G , according to the following rules: ,in, Represents elements The normalized values ​​of priority, estimated travel time, and estimated unloading period in the triplet; are the proportional coefficients of the three components, which must satisfy The setting of the proportional coefficient is related to the scheduling target. A larger setting indicates that more priority is attached to the task. The smaller the value of the matrix element is, the lower the comprehensive cost of matching the mine truck with the chute is, and it should be selected first. The cost matrix used as input to the Hungarian algorithm.

[0043] The transformed cost matrix The Hungarian algorithm is applied to solve the minimum weight matching. The basic idea of ​​the Hungarian algorithm is to perform subtraction transformation on the rows and columns of the matrix respectively, so that each row and column has at least one 0 element, and then continuously modify and optimize the distribution of 0 elements until a complete match is found. Specifically, Subtract the minimum element of each row from the row to get ;right Subtract the smallest element from each column of ;right Perform row and column coverage to minimize the total number of coverage lines and cover every 0 element. If the number of coverage lines is equal to the matrix order N, the optimal match has been obtained and the algorithm ends. Otherwise, proceed to the next step: find the smallest uncovered element, subtract all elements not covered by rows, and add it to all elements covered by rows and columns. The time complexity of the Hungarian algorithm is , where N is the matrix order, i.e. the number of mine trucks or chutes. This algorithm guarantees the global optimal solution under multiple constraints.

[0044] According to the minimum weight matching obtained by the Hungarian algorithm, the mine trucks and chutes are matched one by one to obtain the initial unloading task matching result. For each mine truck, its matching chute is its unloading destination. At the same time, according to the distance and speed between the mine truck and the chute, the estimated arrival time of the mine truck can be estimated. The matching result is formally represented as a triple ,in represents the i-th mining truck, Indicates the target well number that matches it. Represents the estimated arrival time of the mining truck.

[0045] By solving the task allocation matrix using the Hungarian algorithm, an initial, globally optimal unloading task matching scheme can be obtained. This scheme achieves the best match between mine trucks, chutes, and tasks, can maximize the transportation efficiency of mine trucks and the utilization efficiency of chutes, and ensures that the overall execution time of unloading tasks is the shortest under constraints.

[0046] S34, in actual mine production, the chute equipment may stop working due to failure or maintenance. At this time, the unloading task previously assigned to the chute cannot be executed normally and needs to be rearranged as soon as possible to avoid the mine trucks being empty or waiting near the chute and wasting transportation capacity. The system monitors the working status of each chute by acquiring the chute status data in real time. The chute status data can be collected and uploaded by sensors or equipment control systems at the chute site. When the status data of a chute changes abnormally, such as a sudden drop in unloading efficiency, equipment shutdown, fault alarm, etc., the system can determine that the chute has failed and mark it as a "faulty chute".

[0047] To improve the accuracy and reliability of fault identification, the system can use multi-sensor data fusion methods to comprehensively analyze multiple state parameters of the chute, and identify and locate the fault in combination with preset fault diagnosis rules or fault mode libraries. At the same time, it can also use trend analysis of time series data to provide early warning of possible chute failures and provide buffer time for task reallocation.

[0048] For the chute marked as "faulty chute", the system immediately cancels all unloading tasks that have been assigned to the chute. These tasks may be in the following different states: the mine truck has arrived at the faulty chute and started unloading: at this time, the system instructs the mine truck to stop unloading immediately and drive away from the faulty chute; the mine truck is on the way to the faulty chute: the system issues a task cancellation instruction to notify the mine truck to change the destination; the mine truck has not set off yet: directly cancel the task matching between the mine truck and the faulty chute. When revoking a task, the system needs to synchronously update the status information of the mine truck, reset its unloading status to "unassigned", and clear the destination and estimated arrival time. These revoked tasks will re-enter the task pool and wait for reallocation.

[0049] After revoking the assigned tasks, the system needs to update the task allocation matrix to reflect the current task allocation status. The update steps are as follows: remove the column corresponding to the faulty chute from the task allocation matrix so that it no longer participates in the task allocation; for the mine trucks whose tasks are revoked due to the faulty chute, reset the elements of the corresponding rows to the initial values; for other elements in the original allocation matrix, recalculate their priorities, estimated travel times, etc. according to the latest status data, and update the matrix element values; renormalize the updated matrix. The updated task allocation matrix reflects the new situation that the faulty chute cannot continue to work and some mine trucks have re-entered the "pending allocation" state. This matrix will serve as the input for a new round of task allocation optimization.

[0050] The system reapplies the Hungarian algorithm to the updated task allocation matrix to solve the new optimal task matching. The calculation process is the same as the initial matching in S33, except that the input cost matrix has changed. The algorithm solves the adjusted unloading task matching results, including the new target chute and estimated arrival time for each mining truck to be assigned. Compared with the initial matching, the adjusted matching scheme no longer assigns tasks to the faulty chute, but re-optimizes the available chutes.

[0051] The method for reallocating the faulty shaft task proposed in this application uses the idea of ​​real-time state perception and online optimization to dynamically respond to shaft fault events. By revoking the assigned tasks, updating the task allocation matrix, and re-optimizing the matching, it minimizes the impact of the fault while ensuring the stability of the mine truck scheduling, thereby improving the robustness and fault tolerance of the scheduling system. This method makes full use of the V2X network to realize real-time information interaction between the vehicle, road, and cloud, and extends the Hungarian algorithm to online scheduling optimization in a dynamic environment, solving the problems of discontinuous task execution and untimely response in traditional scheduling schemes. At the same time, the system adopts a modular design concept, decoupling functions such as fault diagnosis, task revocation, and task reallocation, thereby improving the maintainability and scalability of the system.

[0052] S35, after obtaining the initial task matching results and the adjusted task matching results, the system needs to select the best solution from the two alternative solutions, and convert the best matching results into specific unloading tasks and send them to the mine truck. Since the initial matching and the matching after fault adjustment have differences in constraints and input data, the matching results obtained may differ in priority distribution and time efficiency. In order to select the globally optimal scheduling plan, the system adopts a two-stage comparison method: Calculate the overall priority score: For each matching result, the system first extracts the priority scores of each unloading task, and then adds them up to obtain the overall priority score of the entire matching result. Suppose the kth matching result contains n unloading tasks, and the priority score of each task is , then the overall priority score of the matching result It can be expressed as: ,The higher the overall priority score is, it means that the matching results contain more high priority tasks and the urgency of the ,overall tasks is higher.

[0053] Calculate the overall travel time. For each matching result, the system extracts the estimated travel time of the mining truck in each unloading task, adds them up, and obtains the overall travel time of the mining truck under this matching result. Assume that the kth matching result contains m mining trucks, and the estimated travel time of each mining truck is , then the overall travel time of the matching result It can be expressed as: ,The shorter the overall driving time is, the higher the overall scheduling efficiency of the matching result is,and the shorter time it takes for the mining truck to complete the unloading task.

[0054] Based on the calculated overall priority score and overall travel time, the system compares the initial matching results and the adjusted matching results. In the first stage, the system compares the overall priority Pk of the two matching results, and the result with the higher priority score is selected as the first candidate. If the Pk of the two results is the same, the second stage is entered to compare the overall travel time Tk, and the result with the shorter travel time is selected as the optimal matching result. Through the two-stage comparison of first comparing the overall priority and then comparing the overall travel time, the system comprehensively considers the two factors of task urgency and scheduling efficiency to ensure that a better scheduling solution is selected under multiple objectives. This comparison method is easy to implement, has low computational complexity, and can effectively balance task response quality and scheduling time performance.

[0055] S36, after determining the optimal matching result, the system needs to convert the matching result into a specific unloading task and send it to each mining truck for execution. The system scans the optimal matching result matrix row by row, and each row corresponds to an unloading task of a mining truck. For the i-th row, the following task elements are extracted: mining truck number: ; Target chute number: ; Estimated time of arrival: ; Encapsulate the extracted elements into a task object, such as The process is repeated until all rows are scanned and n offloading tasks are obtained.

[0056] For each uninstall task , the system sends it to the corresponding mining truck through the V2X network . Tasks can be issued using vehicle networking communication technologies such as DSRC and LTE-V, and the message content includes the target chute address, driving path, estimated arrival time, etc. After receiving the task message, the mine truck sets the target chute as the new navigation destination, and plans the driving path and speed according to the estimated arrival time to ensure that it arrives at the unloading location on time. During driving, the mine truck can also report its own position and status in real time through V2I communication, which facilitates the system to track the progress of task execution. The optimal matching result selection and task generation and issuance method of this application, while making full use of real-time scheduling results, takes into account the balance between task urgency and scheduling efficiency. Through fine-grained optimization and V2X network distribution, the scheduling instructions are quickly and accurately delivered to the mine truck, forming an end-to-end task execution closed loop, and ensuring the global optimality and dynamic adaptability of mine truck scheduling. At the same time, the system uses lightweight processing such as matrix scanning and task encapsulation to reduce the computational overhead of matching result parsing and conversion, and improve the efficiency of task generation.

[0057] S4, the target mining truck generates the optimal planning path from the current position to the target chute through the path planning algorithm, including: S41, extracting the target chute position, the target mining truck extracts the position coordinates of the target chute from the optimal unloading task , as the target point for path planning. The location coordinates of the chute can be obtained on the mine map and sent to the mine truck along with the unloading task. S42, the current location of the mine truck As the starting point of path planning, the mining truck can obtain its own position coordinates in real time through the on-board GPS positioning module.

[0058] S43, obtain the underground mine map, and search for the candidate path from the starting point to the target point on the underground mine map through the improved A* algorithm; specifically, convert the underground mine map into a vector road network map, in which the nodes represent intersections or feature points, and the edges represent road sections. Each edge contains attribute information such as road length, speed limit, and number of lanes. The starting point and target chute of the mining truck are abstracted as nodes on the road network map. According to the volume, load and other parameters of the mining truck, determine whether it meets the traffic restrictions of each road. For roads that do not meet the restrictions, set their edge weights to infinity to avoid searching for infeasible paths.

[0059] When estimating the node cost, not only the geometric distance from the target point is considered, but also the impact of real-time traffic conditions on the travel time. The system estimates the travel time of the cada road section based on traffic flow and congestion, and uses it as part of the heuristic function h(n): ,in, Represents the geometric distance from node n to the target point, It represents the estimated travel time considering the road condition, and α and β are balance factors. In order to improve the search efficiency, a multi-objective pruning strategy is introduced. During the search process, the path that has reached the target point is recorded, and its travel time is used as the evaluation upper limit of other paths. For nodes whose cost estimate is higher than the optimal path, they are directly pruned without further searching. It is possible to search for the path with the optimal travel time in a dynamic traffic environment, and improve the search efficiency through pruning and road condition estimation.

[0060] S44, according to the traffic flow and traffic congestion in the underground mining traffic data, calculate the travel time and travel cost of each candidate path; use the traffic flow and road speed limit to estimate the travel time of the cada section on the path, and then accumulate and sum them to obtain the total travel time of the entire path: ,in, and Respectively represent the length and average speed of the ith road section, and n is the number of road sections on the path. Taking into account factors such as road section length, road condition level, energy consumption, etc., the travel cost of each road section is calculated, and then the total travel cost is accumulated: ,in, It represents the travel cost of the ith road segment, which can be estimated by the weighted sum of the length, fuel consumption, road conditions, etc. By calculating the travel time and travel cost, the pros and cons of each candidate path can be comprehensively evaluated.

[0061] S45, based on the travel time and travel cost of each candidate path, a weighted scoring method is used to select the candidate path with the highest score as the optimal planning path; due to the different dimensions of travel time and travel cost, normalization is required. A linear normalization method is used to map the cada attribute values ​​to the 0-1 interval: , ,in, and is the normalized attribute value, , are the maximum and minimum values ​​of the travel time and travel cost among all candidate paths, respectively. Weights are assigned to the normalized travel time and travel cost, and the comprehensive score of each path is calculated: ,in, and is the weight coefficient of time and cost, satisfying . According to the production scheduling requirements of the mining area, the weight value can be flexibly set to give priority to the path with the shortest time or the lowest cost.

[0062] All candidate paths are sorted from high to low according to the comprehensive score S, and the path with the highest score is the optimal planning path. If there are multiple paths with the highest score, any one can be selected as the optimal path. Compared with the empirical path, this method can significantly improve the path rationality and traffic efficiency of mining trucks, alleviate the traffic pressure on local sections, and improve the completion rate of unloading tasks. At the same time, the multi-attribute weighted scoring has strong adaptability and flexibility, and the optimization strategy can be adjusted according to the actual selection preferences of the mining area.

[0063] S5, the target mine truck drives to the target chute according to the optimal planned path, and the central control system calculates the position deviation of the target mine truck and the unloading port of the target chute according to the positions of the target mine truck and the target chute, and generates a control instruction according to the position deviation, and controls the target mine truck to align with the unloading port according to the control instruction.

[0064] S51, the target mining truck drives according to the optimal planned path, and uploads its own position and speed to the central control system through the V2X communication module; the real-time position coordinates (x, y) and heading angle θ of the mining truck are obtained by fusion positioning using multi-source sensors such as GPS / Beidou, IMU, and wheel speed. The real-time driving speed v of the mining truck is measured by the wheel speed sensor. The position coordinates (x, y), heading angle θ, and speed v are encapsulated into a position speed object (x, y, θ, v), and a timestamp and vehicle ID are added. Using V2X communication technologies such as DSRC and LTE-V, the position speed object is uploaded to the central control system, and the upload frequency can be set according to the communication bandwidth and control requirements. After receiving the position data uploaded by the on-board equipment, the central control system updates the real-time position and speed information of the mining truck to provide input for subsequent alignment control.

[0065] S52, the target chute obtains the matching target mine truck according to the optimal unloading task, and uploads the unloading port position of the target chute to the central control system through the roadside sensing device; as shown in the screenshot, a laser radar is installed above the unloading port of the chute to perform a three-dimensional scan of the unloading port and generate point cloud data. Using algorithms such as ICP, the scanned point cloud is aligned with the chute CAD model to calculate the spatial coordinates (X, Y, Z) of the unloading port. The unloading port coordinates are converted to the same coordinate system as the mine truck positioning to obtain a unified position description. The unloading port coordinates are encapsulated as a position object (X, Y, Z), the target mine truck ID is marked, and it is uploaded to the central control system through the roadside communication equipment. After the central control system receives the unloading port position of the chute, it matches it according to the target mine truck ID to prepare for the position deviation calculation.

[0066] S53, the central control system continuously receives the position and speed messages of the mining truck and the position messages of the unloading port of the chute through the V2X module and road test communication. The system compares the timestamps of the two types of messages and extracts a set of time-synchronized vehicle-road data for subsequent deviation calculation: read the position (x, y), heading angle θ, and speed v from the mine truck message to form the mine truck state vector [x, y, θ, v]'; read the unloading port coordinates (X, Y, Z) from the chute message to obtain the unloading port position vector [X, Y, Z]'; based on the Kalman filter theory, establish a mine truck kinematic model, take the state vector as the estimated quantity, and the position and speed message as the observed quantity, and recursively estimate the optimal state of the mine truck [x', y', θ', v']'; coordinate system conversion, map the unloading port position [X, Y, Z]' from the chute coordinate system to the mine truck coordinate system, recorded as [X', Y', Z']'; calculate the deviation between the current posture of the mine truck and the unloading port: lateral deviation Δy = Y'-y'; longitudinal deviation Δx = X'-x'; heading angle deviation Δθ = atan2(Z', θ').

[0067] S54, input the calculated deviation (Δx, Δy, Δθ) into the model predictive controller of the central control system, and load the dynamic model and kinematic constraints of the mine truck at the same time. The controller adopts the following design process: Mine truck dynamic modeling. The node dynamic model is used longitudinally, and the two-degree-of-freedom bicycle model is used horizontally to characterize the force change law of the mine truck, such as speed, acceleration, torque, etc. Constraint setting. Including the physical parameter restrictions of the mine truck (such as maximum speed, maximum acceleration, minimum turning radius, etc.) and the geometric constraints of the unloading platform (such as channel width, ramp angle, etc.). Prediction model construction. In the discrete time scale, based on the current measurement value and future control sequence, predict the dynamic response of the system state quantity. Performance index determination. Design a multi-objective optimization problem, such as minimizing the heading deviation, minimizing the braking distance, and the shortest adjustment time. Rolling optimization solution. Using numerical optimization methods such as sequential quadratic programming, the optimal control quantity at the current moment is solved online, mainly including the vehicle speed adjustment Δv, the front wheel angle adjustment Δφ, and the adjustment time Δt. Control instruction generation. The solution results are encapsulated into control instructions and sent to the mining truck through V2X communication. The on-board actuator adjusts the throttle / brake, steering wheel and gearbox at the same time according to the received control instructions to control the mining truck to move along the desired trajectory. Periodic rolling optimization. Based on the newly collected vehicle-road information, the model predictive control problem is updated to generate new optimal control instructions to guide the mining truck to continuously adjust until the alignment requirements are met.

[0068] Task execution and monitoring, automatic driving of mining trucks. After receiving the unloading task, the mining truck extracts the location coordinates and estimated arrival time of the target chute from the task information as the driving target and time limit. The on-board control system plans a preliminary driving route based on the current position and target position, combined with the on-board electronic map. During driving, the mining truck communicates with the roadside base station and other mining trucks in real time through the V2X module to receive dynamic information such as road traffic flow and congestion. When congestion is detected on the road ahead, the on-board system will automatically re-plan the route, select an alternative route, avoid the congested section, and reduce driving time. When planning the route, factors such as road capacity, expected queuing time, and route length will be comprehensively considered to minimize the time it takes for the mining truck to reach the target chute. When the mining truck arrives at the target chute, the visual sensor and laser radar are used to identify the exact location of the chute, and the body posture is controlled to dock with the chute unloading port. Through machine vision guidance and servo control, the automatic positioning and high-precision docking of the carriage and the chute are achieved without human intervention.

[0069] The chute equipment is also equipped with a V2X communication module, which can receive unloading tasks issued by the central dispatching system. According to the target mining truck in the task and the estimated arrival time, the chute equipment prepares for unloading in advance, such as adjusting the chute angle and opening the unloading door. During the unloading process of the mining truck, the control system of the chute equipment continuously collects equipment operating parameters through various sensors, including unloading rate, cumulative unloading volume, silo material level, etc. The system monitors the working condition of the equipment in real time to determine whether there is any abnormality. At the same time, according to the real-time unloading volume of the mining truck, the unloading speed of the chute is dynamically adjusted to synchronize with the unloading speed of the mining truck, ensuring unloading efficiency while avoiding overflow. When the mining truck completes unloading, the chute equipment automatically closes the unloading door and uploads the unloading data to the central dispatching system for subsequent task optimization. The fully automatic control of the chute reduces the manual operation link and improves unloading efficiency and safety.

[0070] Real-time monitoring and exception handling, exception detection and adjustment, during the unloading operation, the system should have the ability to detect and handle equipment anomalies and task deviations in real time. The system obtains the status data reported by various sensing devices and V2X communication, and detects abnormal situations in time by setting reasonable thresholds. For chute equipment, the monitored parameters mainly include equipment fault signals, unloading rate, material level height, etc. When a fault signal is detected, or parameters such as unloading rate and material level exceed the normal range, the system determines that the chute is abnormal, immediately sends an alarm to the central control center, and notifies the mining truck that is unloading or preparing to unload to suspend operations to avoid unloading obstructions or accidents due to chute failures. For mining trucks, their driving trajectory and vehicle status are mainly monitored. If the mining truck deviates from the planned path, the system will issue a timely warning and issue a correction instruction to guide the mining truck back to the correct route. If the mining truck fails or has other abnormalities, the system will notify the mining truck to stop, and dispatch other mining trucks to take over the task in time to ensure the continuous unloading operation. When the monitoring system finds an abnormal situation, the central dispatching system will immediately start the emergency dispatching plan. For chute failures, the system automatically reallocates tasks to other available chutes. If a mining truck fails, the system will adjust the dispatch strategy and dispatch a mining truck from the spare fleet to replace it. The emergency dispatch plan can quickly respond to emergencies and minimize the impact of abnormalities on unloading operations.

[0071] Data feedback and optimization: After each unloading task is completed, the control systems of the mine truck and chute will transmit the task execution data back to the central dispatching system, including the actual driving path, driving time, unloading volume, energy consumption, etc. The dispatching system uses data analysis algorithms to evaluate the efficiency of this dispatching plan, and optimizes subsequent task allocation, path planning and other strategies in combination with the overall production status of the current mining area. During the data analysis process, the dispatching system establishes a correlation model between task efficiency and influencing factors, including the relationship between task allocation and vehicle utilization, the relationship between road conditions and mine truck waiting time, etc. Through the machine learning algorithm, the optimization model is continuously fitted to find the best combination of dispatching parameters to improve the accuracy of task allocation. At the same time, the system summarizes the experience of handling various abnormal situations, iteratively optimizes the emergency dispatch plan, and improves the system's ability to respond to emergencies. The continuous feedback and analysis of task completion data and abnormal handling experience enable the dispatching system to have the ability of self-learning and optimization, adapt to the dynamically changing mining environment, and continuously improve dispatching efficiency.

[0072] Task completion and reporting. When the mining truck completes the unloading task, the on-board system generates a task completion report, compares the actual execution status with the task plan, and reports it to the central dispatch system through V2X communication. The content of the task completion report includes: the difference between the actual driving path and the planned path, the difference between the actual driving time and the expected time, the difference between the unloading amount and the planned amount, energy consumption, vehicle status and other data. After receiving the task completion report, the central dispatch system stores the relevant data in the task execution database for generating mining area production reports and performance evaluation. At the same time, the dispatch system analyzes the difference data in the completion report, judges the efficiency and accuracy of task execution, and identifies the key factors affecting task completion, such as road congestion and vehicle performance, to provide data support for subsequent dispatch optimization. The task completion report enables the dispatch system to grasp the complete life cycle data of each unloading task, objectively evaluate the feasibility of the dispatch plan, and also provides mine area managers with intuitive production data display and analysis tools, so as to facilitate timely discovery and resolution of problems in production organization.

[0073] Dynamic optimization and self-learning, real-time data feedback, the intelligent scheduling system has a complete data collection and feedback mechanism, which can continuously obtain various real-time data during the unloading operation. Through the correlation analysis of multi-source heterogeneous data such as on-board sensors, chute monitoring equipment, V2X communication, etc., the system fully perceives the production status of the mining area and grasps the key factors affecting the scheduling effect. The system focuses on collecting the following real-time data: status data such as the position, speed, load, and power of the mining truck; production data such as the equipment condition, material level height, and unloading volume of the chute; environmental data such as road traffic flow, congestion, and road conditions; performance data such as the completion of scheduling tasks, task efficiency, and abnormal conditions. After cleaning, fusion, and correlation analysis, the massive real-time data is converted into a knowledge base for scheduling optimization. Through data mining algorithms, the system learns the optimal scheduling strategy under different scenarios from historical data, and applies it to dynamically adjust the current scheduling plan so that scheduling decisions can adapt to real-time changes in production conditions.

[0074] Algorithm optimization, the scheduling algorithm is the core of the intelligent scheduling system, and its quality directly determines the scheduling efficiency and accuracy. In view of the characteristics of large-scale dynamic scheduling of open-pit mines, this application has designed a variety of optimization algorithms, and continuously improved the algorithm performance through a self-learning mechanism. Multi-objective optimization: In scheduling decisions such as task allocation and path planning, multiple scheduling objectives such as time, energy consumption, and equipment utilization are comprehensively considered to establish a multi-objective optimization model, solve the Pareto optimal solution, and balance various scheduling indicators. Adaptive parameter adjustment: For different production conditions, the system automatically adjusts the key parameters of the scheduling algorithm, such as the time weight of task allocation, the energy consumption weight of path planning, etc., so that the algorithm can adapt to the dynamic environment. The parameter adjustment process adopts a reinforcement learning framework to accumulate experience through continuous trial and error and learn the optimal parameter combination.

[0075] Hierarchical scheduling: The entire scheduling problem is divided into the strategic layer, the task layer, and the operational layer, and a hierarchical optimization strategy is adopted. The strategic layer is responsible for the formulation of global scheduling strategies, the task layer is responsible for task decomposition and optimization, and the operational layer is responsible for the execution of specific tasks. Each layer is dynamically adjusted based on real-time feedback and works together. Active learning: In the case of unsatisfactory scheduling results, the system diagnoses the weak links in scheduling and optimizes them in a targeted manner through an active learning mechanism. The system learns from experts and collects manual scheduling experience; learns from the optimal scheduling samples through imitation learning; and applies scheduling strategies for other scenarios through knowledge transfer. The scheduling algorithm continues to evolve in production practice through self-learning and dynamic optimization, transitioning from expert experience to data intelligence, and ultimately forming a set of adaptive, efficient, and accurate mine intelligent scheduling optimization systems, laying a solid technical foundation for the realization of unmanned and intelligent mining production processes.

[0076] Execution and monitoring of tasks, automatic driving of mining trucks: After receiving the task, the mining truck receives the location and estimated arrival time of the target chute through V2X communication, and automatically adjusts the driving route to avoid congestion and reduce waiting time. When the mining truck arrives at the target chute, it automatically docks for unloading; chute equipment coordination: After receiving the task, the chute equipment is ready to receive the mining truck for unloading. During the unloading process, the system continuously monitors the working status of the chute equipment to ensure the normal operation of the equipment, and automatically adjusts the working speed of the equipment according to the unloading amount.

[0077] Real-time monitoring and exception handling, frequent detection and adjustment: During the unloading process, the system monitors the working status of the mining truck and the chute through sensing devices and V2X communication. If the chute equipment fails or the mining truck deviates from the task path, the system will immediately issue an alarm and automatically adjust the operation plan to avoid interruption of the operation. Data feedback and optimization: After the unloading task is completed, the system will analyze the task execution results, evaluate the efficiency of the scheduling, and optimize the subsequent task allocation and scheduling strategy based on the feedback information.

[0078] Task completion and reporting: When a task is completed, the completion status is reported. The system records the task execution data, which can be used for subsequent optimization and statistical management. Dynamic optimization and self-learning, real-time data feedback: Through continuous collection and feedback of task execution data, traffic conditions, mining truck operation data, etc., the system can learn the optimal scheduling strategy under different circumstances. Algorithm optimization: The system will adjust the scheduling algorithm according to job feedback to improve the accuracy of task scheduling and resource utilization. Especially in high load and complex environment, the system can automatically adjust the strategy to ensure the timely completion of tasks.

Claims

1. A mining truck unloading method, characterized in that: include: S1, collects chute status data and underground mining traffic data through roadside sensing equipment, and uploads the collected data to the central control system; S2, after the mining truck is loaded with ore, it sends an unloading task request to the central control system through the V2X communication module; S3, the central control system generates the optimal unloading task of the corresponding mine truck using the task matching algorithm according to the chute status data, underground mine traffic data and mine truck status data, and sends the optimal unloading task to the target mine truck and target chute; S4, the target mining truck generates the optimal planning path from the current position to the target chute through the path planning algorithm; S5, the target mine truck drives to the target chute according to the optimal planning path, and the central control system calculates the position deviation of the target mine truck and the unloading port of the target chute according to the positions of the target mine truck and the target chute, and generates a control instruction according to the position deviation, and controls the target mine truck to align with the unloading port according to the control instruction; S6, after the target mine truck is aligned, the ore is unloaded into the target chute; S3, using task matching algorithm to generate the optimal unloading task of the corresponding mining truck, including: S31, determining the priority of each unloading task by using a weighted scoring method according to the chute status data, the underground mine traffic data and the mine truck status data; S32, using the Hungarian algorithm to establish a task allocation matrix G between the mine truck and the chute, the elements in the task allocation matrix are the priority of the unloading task, the estimated driving time of the mine truck and the estimated working time of the chute; the size of the task allocation matrix G is NxM, N is the number of mine trucks to be allocated, and M is the number of available chutes; the task allocation matrix is ​​a two-dimensional matrix, the rows of the matrix represent the mine trucks, the columns represent the chutes, and the matrix elements are a triple, including the priority of the unloading task between the mine truck and the chute, the estimated driving time of the mine truck and the estimated working time of the chute; S33, using the Hungarian algorithm to solve the task allocation matrix G, to obtain the initial unloading task matching result including the target chute and the estimated arrival time of each mining truck; for any element in the task allocation matrix G , and transform according to the following rules: ,in, Represents elements The normalized values ​​of priority, estimated travel time, and estimated unloading period in the triplet; are the proportional coefficients of the three components; As input cost matrix to the Hungarian algorithm; S34, obtaining the working status of the chute in the chute status data, and when the working status of the chute indicates that the corresponding chute fails, canceling the unloading task assigned to the chute, and re-incorporating the unloading task into the task allocation matrix, using the Hungarian algorithm to solve the updated task allocation matrix, and obtaining an adjusted unloading task matching result; S35, comparing the initial offloading task matching result and the adjusted offloading task matching result, and selecting the optimal offloading task matching result; S36, extracting the target chute number and estimated arrival time corresponding to each mining truck according to the optimal unloading task matching result, and generating the optimal unloading task for each mining truck; S32, using the Hungarian algorithm to establish the task allocation matrix between the mine truck and the chute, including: According to the priority of each unloading task, a priority matrix between the mine truck and the chute is established, wherein the rows of the priority matrix represent the mine truck, the columns represent the chute, and the elements in the priority matrix are the priorities of the unloading tasks between the corresponding mine truck and the chute; According to the position and speed of the mine truck in the mine truck status data, the estimated driving time from each mine truck to each chute is calculated, and a driving time matrix between the mine truck and the chute is established, wherein the rows of the driving time matrix represent the mine trucks, the columns represent the chute, and the elements of the driving time matrix are the estimated driving time from the corresponding mine truck to the corresponding chute; According to the expected unloading cycle of the chute in the chute status data, the expected working time of each chute is obtained, and a chute working time matrix is ​​established, wherein the rows and columns of the chute working time matrix represent chutes, and the diagonal elements in the chute working time matrix are the expected working time of the corresponding chute; Normalize the priority matrix, travel time matrix and chute working time matrix; According to the normalized priority matrix, travel time matrix and chute working time matrix, a task allocation matrix is ​​established. The rows in the task allocation matrix represent mine trucks, the columns represent chutes, and the elements are triples. The triplet contains the priority of the unloading task, the estimated travel time of the mine truck, and the estimated working time of the chute.

2. The mine truck unloading method according to claim 1, characterized in that: The chute status data includes chute workload, working status and expected unloading cycle; the underground mining traffic data includes traffic flow and traffic congestion.

3. The mine truck unloading method according to claim 2, characterized in that: The unloading task request includes mine truck status data including the position, speed and load of the mine truck.

4. The mining truck unloading method according to claim 1, characterized in that: S35, selecting the optimal offloading task matching result, including: Calculate the overall priority scores of the initial offloading task matching result and the adjusted offloading task matching result respectively; Calculate the overall estimated travel time of the mining truck for the initial unloading task matching result and the adjusted unloading task matching result respectively; Compare the overall priority scores of the initial unloading task matching result and the adjusted unloading task matching result, and take the unloading task matching result with a higher overall priority score as the candidate optimal matching result; if the overall priority scores of the two are the same, continue to compare the overall estimated driving time of the mining truck, and take the unloading task matching result with a shorter overall estimated driving time of the mining truck as the optimal unloading task matching result.

5. The mine truck unloading method according to claim 3, characterized in that: S4, the target mining truck generates the optimal planning path from the current position to the target chute through the path planning algorithm, including: S41, the target mining truck extracts the position coordinates of the target chute in the optimal unloading task; S42, taking the current position of the target mining truck as the starting point and the position coordinates of the target chute as the target point; S43, obtaining an underground mine map, and searching for a candidate path from a starting point to a target point on the underground mine map by using an A* algorithm; S44, calculating the travel time and travel cost of each candidate path according to the traffic flow and traffic congestion in the underground mining traffic data; S45, according to the travel time and travel cost of each candidate path, a weighted scoring method is used to select the candidate path with the highest score as the optimal planning path.

6. The mining truck unloading method according to claim 3, characterized in that: S5, generating a control instruction according to the position deviation, and controlling the target mining truck to align with the unloading port according to the control instruction, including: S51, the target mining truck drives according to the optimal planned path and uploads its position and speed to the central control system through the V2X communication module; S52, the target chute obtains a matching target mining truck according to the optimal unloading task, and uploads the unloading port position of the target chute to the central control system through the roadside sensing device; S53, the central control system uses the Kalman filter algorithm to integrate the position, speed and posture of the target mining truck and the unloading port position of the target chute, and calculates the lateral deviation, longitudinal deviation and posture Euclidean deviation of the target mining truck relative to the unloading port of the target chute; S54, generating an adjustment instruction according to the lateral deviation, the longitudinal deviation and the attitude deviation by using a model predictive control algorithm, wherein the adjustment instruction includes adjusting the direction, adjusting the distance and adjusting the attitude.

Images