A dynamic vehicle avoidance planning method, electronic equipment, and computer-readable storage medium for underground coal mines based on multi-source data fusion.

Abstract

This invention relates to the field of intelligent traffic control technology in mines, and in particular to a method, electronic device, and computer-readable storage medium for dynamic vehicle avoidance planning in coal mines based on multi-source data fusion. The method includes: S1, real-time acquisition of multi-source data; S2, constructing an avoidance demand assessment model, calculating the avoidance demand index of a first and second vehicle traveling in opposite directions based on the multi-source data, the avoidance demand index being obtained through a two-layer assessment mechanism; S3, comparing the avoidance demand indices of the first and second vehicles, and controlling the vehicle with the lower avoidance demand index to perform an avoidance action; S4, generating and issuing dynamic avoidance commands, which are simultaneously sent to the on-board terminal of the target vehicle and the guidance display screen of the target chamber, and triggering a vibration alarm on the personnel positioning card when personnel are detected approaching. By constructing a technical closed loop of multi-dimensional real-time data → quantitative assessment of avoidance demand → generation of dynamic passage commands, adaptive allocation of roadway resources is achieved.

Description

Technical Field

[0001] This invention relates to the field of intelligent traffic control technology in mines, and in particular to a dynamic vehicle avoidance planning method, electronic equipment, and computer-readable storage medium for underground coal mines based on multi-source data fusion. Background Technology

[0002] Currently, underground coal mine roadways are characterized by narrow spaces and complex environments, making vehicle transportation safety and efficiency paramount. However, existing underground traffic control systems still face numerous technical bottlenecks, primarily in the following aspects: (1) The problem of information silos is serious. Vehicle positioning systems (such as UWB / RFID), personnel positioning systems, traffic light control systems and chamber occupancy status sensors usually operate independently. There is a lack of effective integration and coordination between data, which cannot provide a unified and comprehensive information view for scheduling decisions, forming decision blind spots.

[0003] (2) The scheduling strategy is static and rigid, generally relying on fixed time periods for release or manual experience scheduling, which cannot respond in real time to dynamic changes such as sudden congestion and vehicle convergence in the lane, and lacks an efficient emergency avoidance decision-making mechanism.

[0004] (3) Safety risks are prominent, especially since heavy engineering vehicles have large blind spots. When personnel accidentally enter the vehicle passageway, the lack of linkage between systems makes it impossible to provide real-time and effective early warnings, which can easily lead to safety accidents. (4) There is a bottleneck in the operation efficiency. As a key resource, the avoidance chamber lacks an intelligent priority allocation strategy. It is often first-come, first-served, which leads to vehicles queuing in an disorderly manner at the entrance of the chamber and waiting time is too long, which seriously restricts the overall efficiency of underground logistics transportation.

[0005] Therefore, there is an urgent need for a new method that can break down data silos, make collaborative intelligent decisions, and achieve dynamic obstacle avoidance command, so as to comprehensively improve the safety, intelligence, and operational efficiency of underground transportation in coal mines. Summary of the Invention

[0006] The technical problem to be solved by this invention is to provide a dynamic obstacle avoidance planning method for underground vehicles in coal mines based on multi-source data fusion, in order to address the problems of information silos, static scheduling defects, security risks and efficiency bottlenecks in the existing technologies mentioned above.

[0007] The technical solution adopted by this invention to solve its technical problem is: a dynamic obstacle avoidance planning method for underground vehicles in coal mines based on multi-source data fusion, characterized in that it includes: S1. Real-time acquisition of multi-source data, including vehicle positioning data, vehicle status data, personnel positioning data, chamber occupancy status data, and roadway signal light status data; S2. Construct an avoidance demand assessment model, and calculate the avoidance demand index of the first vehicle and the second vehicle traveling in opposite directions based on the multi-source data. The avoidance demand index is obtained through a two-layer assessment mechanism. S3. Compare the avoidance demand index of the first vehicle and the second vehicle, and control the vehicle with the lower avoidance demand index to perform the avoidance action. S4. Generate and issue a dynamic avoidance command. The dynamic avoidance command is simultaneously sent to the vehicle terminal of the target vehicle and the guidance display screen of the target chamber, and triggers the vibration alarm of the personnel positioning card when personnel are detected approaching.

[0008] This method constructs an avoidance demand assessment model and comparison mechanism based on multi-source data, overcoming the defects of static scheduling. The instruction issuance mechanism in step S4, through the multi-terminal linkage of vehicle-mounted terminals, chamber guidance screens, and personnel positioning cards, ensures that decision instructions are executed efficiently and accurately, providing a reliable execution guarantee for subsequent solutions to safety risks and efficiency bottlenecks.

[0009] According to one embodiment of the present invention, the two-layer evaluation mechanism in step S2 includes a first-layer static evaluation and a second-layer dynamic evaluation, wherein the vehicle weight W in the first-layer static evaluation is... v The calculation formula is: W v =k1×T+k2×E, Where k1 is the load capacity coefficient, T is the load capacity, k2 is the mission urgency coefficient, and E is the mission urgency. The formula for calculating the avoidance demand index D in the second-level dynamic assessment is as follows: D = α × (1 / d p )+β×C+γ×R, Where α, β, and γ are all weighting coefficients, and satisfy α + β + γ = 1, d p C is the distance between the vehicle and the nearest person, C is the roadway congestion coefficient, and R is the chamber availability rate, R=N free / N total N free N represents the number of currently available chambers. total Let R be the total number of chambers, and 0 ≤ R ≤ 1; The avoidance demand index D is calculated by coupling the weighting coefficient γ with the chamber availability rate R to determine the allocation of target chambers.

[0010] The first layer of static assessment, based on vehicle load capacity and mission urgency, identifies vehicle priorities to ensure efficient passage for important tasks and heavy vehicles. The second layer of dynamic assessment introduces three real-time changing environmental factors: distance between vehicles and personnel, roadway congestion coefficient, and chamber availability. This enables the decision-making model to keenly perceive and respond to dynamic changes in the underground environment. This dual-layer fusion assessment ensures that the final avoidance decision considers both the mission itself and the real-time situation, greatly improving the rationality and accuracy of the decision.

[0011] According to one embodiment of the present invention, the task urgency E is an 8-bit field written by the production scheduling server and broadcast via the CAN bus when the vehicle departs, and normalized by the formula k2=E / 255; if the received E value is interrupted for more than 3 broadcast cycles, the system defaults to E=0.

[0012] By writing to the production scheduling server and broadcasting via the CAN bus, the authority of task urgency information and its seamless integration with the vehicle control system are ensured, achieving deep linkage between production scheduling and traffic control. Normalization processing enables the unified quantification of urgency values ​​from different sources, facilitating model calculations. If an interruption exceeds three cycles, the default value E=0 is set, effectively avoiding the problem of misjudging task priorities due to brief communication interruptions, which could lead to the system continuously waiting for high-priority instructions and getting stuck. This enhances the system's stability and availability in the complex electromagnetic environment downhole.

[0013] According to one embodiment of the present invention, the formula for calculating the lane congestion coefficient C is as follows: C = (N) vehicle / L lane )×(1- V avg / V limit )×λ; Where, N vehicle L represents the current number of vehicles in the roadway section. lane V is the length of the tunnel section. avg V represents the average speed of vehicles within the segment. limit The speed limit for the roadway is given by λ, which is the vehicle type correction factor.

[0014] By combining vehicle density, speed efficiency, and vehicle type, the system can accurately reflect the actual traffic load of a lane. Even if the number of vehicles is small, but the speed is slow, or there are many heavily loaded vehicles, the system can identify potential congestion risks. This quantitative perception provides accurate data input for subsequent dynamic decision-making.

[0015] According to one embodiment of the present invention, Every 200ms, count all vehicles in the current roadway section; When the calculated lane congestion coefficient C > 0.7, the weighting coefficient β is increased by 30%. When the calculated congestion coefficient C of the alleyway is greater than 0.9, the system will forcibly activate the one-way mode, instructing oncoming vehicles in the congested section to enter the nearest vacant chamber to give way.

[0016] The statistical frequency of every 200ms ensures that the system can perceive changes in traffic flow at a near real-time speed. When C > 0.7 triggers moderate congestion, the system automatically guides vehicles to smoother routes or tunnels by increasing the β value in the dynamic weights to prevent congestion from worsening. When C > 0.9 triggers severe congestion, the system forcibly activates a one-way traffic mode and forcibly clears one side of the lane, thereby breaking the congestion deadlock in the shortest possible time, improving the efficiency of traffic flow in extreme situations, and avoiding long-term disorderly queuing of vehicles.

[0017] According to one embodiment of the present invention, when the lane congestion coefficient C > 0.7, the estimated arrival time t of the first vehicle and the second vehicle at the conflict point is calculated. A and t B , like Then, compare D + 0.2 × C and reorder. If D A >D B Then the second vehicle enters the nearest available chamber; If there is no available chamber, the second vehicle will remain in place with its hazard lights on, and the first vehicle will reduce its speed to 0.8 × V. limit ; like If so, the original state will be maintained.

[0018] In specific scenarios where two vehicles are about to meet at a conflict point, the system assesses conflict risk by predicting the time difference of arrival. When the time difference is less than 20 seconds, the system recognizes the risk and does not simply use the original priority. Instead, it recalculates the congestion coefficient C as an additional bonus D+0.2×C. This allows for faster congestion relief in congested environments. Subsequent processes such as waiting and slowing down in open chambers ensure that there are corresponding safety instructions output under any circumstances, enhancing the reliability and safety of the system.

[0019] According to one embodiment of the present invention, in step S3, if valid data on the availability of the chamber cannot be obtained within a preset time, the system defaults to γ=0 and forces all vehicles to give way to pedestrians, prioritizing the safety of personnel.

[0020] When the system is unable to obtain the status of the chamber due to sensor failure, network interruption or other reasons, the system will default to all chambers being unavailable and force the system to enter a pedestrian priority safety mode. This means that any vehicle avoidance decision will give way to pedestrian safety. As long as personnel are detected approaching, vehicles must give way, providing the highest level of safety guarantee for underground personnel and solving the systemic risks of large blind spots of heavy vehicles and lack of early warning for personnel accidentally entering the mine.

[0021] According to one embodiment of the present invention, in step S4, the instruction sent to the chamber guidance display screen includes the chamber number, vehicle number, and reversing guidance distance information; the instruction sent to the vehicle's on-board terminal includes voice broadcast and screen animation path guidance information.

[0022] The specific instructions sent to the guidance screen in the tunnel provide clear and accurate guidance for the driver, avoiding misoperation or hesitation caused by unclear instructions; the multimodal instructions sent to the vehicle terminal can effectively attract the driver's attention and guide their operation in the noisy and attention-limited environment underground, ensuring that the decisions generated by all the aforementioned intelligent algorithms can be executed correctly and efficiently.

[0023] An electronic device is also provided, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the program, it implements the steps of the dynamic avoidance planning method for underground vehicles in coal mines based on multi-source data fusion as described in the above scheme.

[0024] A computer-readable storage medium is also provided, on which a computer program is stored, characterized in that, when the program is executed by a processor, it implements the steps of the dynamic avoidance planning method for underground vehicles in coal mines based on multi-source data fusion described in the above scheme.

[0025] The beneficial effects of this invention are: This invention constructs an avoidance demand assessment model and comparison mechanism based on multi-source data, overcomes the defects of static scheduling, and ensures that decision-making instructions are executed efficiently and accurately through multi-terminal linkage of vehicle terminals, chamber guidance screens, and personnel positioning cards, providing reliable execution guarantee for subsequent solutions to safety risks and efficiency bottlenecks; The first-level static evaluation considers the inherent attributes of vehicles and task attributes, ensuring the basic fairness and rationality of the scheduling strategy and guaranteeing that important transportation tasks are given priority. The second-level dynamic evaluation introduces real-time variables such as the distance between vehicles and personnel, the roadway congestion coefficient, and the availability of the tunnel. When the congestion coefficient C increases, the system can automatically issue an early warning and adjust the strategy. When C > 0.9, the one-way mode is forcibly activated. When the tunnel data is abnormal, the system can automatically downgrade to the safe mode, flexibly responding to various complex scenarios such as congestion, intersections, and sudden accidents, reducing vehicle queuing time, and breaking through the efficiency bottleneck of underground transportation. The abnormal degradation strategy establishes pedestrian safety as the highest priority; the introduction of the vehicle type correction coefficient λ enables the system to distinguish between different risk levels of heavy vehicles and light vehicles; and the vibration alarm triggered by the personnel positioning card realizes proactive and direct early warning for personnel, rather than relying solely on the passive avoidance of vehicles, which greatly improves the safety level of underground operations. By incorporating the availability of tunnels as a dynamic decision factor into the algorithm model, when deciding which vehicle should enter a tunnel to give way, the system considers not only the vehicle's priority but also the current vacancy status of the tunnel. Combined with the one-way traffic mode, the system can efficiently direct vehicles to use tunnel resources in batches and in an orderly manner during congestion, avoiding vehicles crowding at tunnel entrances, competing for lanes, and waiting in an disorderly manner. This allows the limited tunnel resources to play their maximum role in traffic flow and optimizes the overall traffic capacity of the tunnel. Attached Figure Description

[0026] The present invention will be further described below with reference to the accompanying drawings and embodiments.

[0027] Figure 1 This is a flowchart of Embodiment 1 of the present invention.

[0028] Figure 2 This is a control block diagram of Embodiment 1 of the present invention.

[0029] Figure 3 This is a flowchart of the dynamic evaluation in Embodiment 1 of the present invention. Detailed Implementation

[0030] The present invention will now be described in further detail with reference to the accompanying drawings. These drawings are simplified schematic diagrams, illustrating only the basic structure of the invention, and therefore only show the components relevant to the invention.

[0031] Example 1 like Figure 1 and Figure 2 As shown, a dynamic vehicle avoidance planning method for coal mines based on multi-source data fusion includes: S1. Real-time collection of multi-source data, including vehicle positioning data, vehicle status data, personnel positioning data, chamber occupancy status data, and roadway signal light status data; S2. Construct an avoidance demand assessment model, calculate the avoidance demand index of the first and second vehicles traveling in opposite directions based on multi-source data, and obtain the avoidance demand index through a two-layer assessment mechanism. S3. Compare the avoidance demand index of the first vehicle and the second vehicle, and control the vehicle with the lower avoidance demand index to perform the avoidance action. S4. Generate and issue dynamic avoidance commands. The dynamic avoidance commands are simultaneously sent to the vehicle terminal of the target vehicle and the guidance display screen of the target chamber. When personnel are detected approaching, the vibration alarm of the personnel positioning card is triggered.

[0032] The two-layer evaluation mechanism includes a first-layer static evaluation and a second-layer dynamic evaluation. The vehicle weight W in the first-layer static evaluation is... v The calculation formula is: W v =k1×T+k2×E, Wherein, k1 is the load capacity coefficient, with a default value of k1=0.6, which can also be adjusted by the mine in the ground dispatch software within the range of 0.4-0.8; T is the load capacity, which is automatically written into the CAN frame on-site through the vehicle-mounted weighing sensor or dispatch order, with an accuracy of ±0.5t; k2 is the task urgency coefficient, and E is the task urgency; the task urgency E is an 8-bit field written by the production dispatch server and broadcast via the CAN bus when the vehicle departs, and is normalized by the formula k2=E / 2 55; if the received E value is interrupted for more than 3 broadcast cycles, the system defaults to E=0.

[0033] By broadcasting task urgency via the CAN bus, deep integration between the vehicle control system and the production scheduling system is achieved, ensuring the authority and real-time nature of scheduling instructions. Normalization processing enables tasks of different priorities to be compared under the same quantitative system.

[0034] Specifically, E∈[0,255], 0-63 corresponds to routine material transportation; 64-127 corresponds to equipment maintenance transportation; 128-191 corresponds to gas inspection escort vehicles; and 192-255 corresponds to emergency rescue. The UWB vehicle tag encapsulates the task urgency E field within the location data frame, with a period of ≤200ms.

[0035] During the normalization process, it is necessary to ensure that 0 ≤ k2 ≤ 1 and that k2 is of the same order of magnitude as k1. To ensure this requirement, a CAN-UWB joint test bench was built in the laboratory, and 1000 sets of E values ​​were randomly injected to verify that k2 is linear and has no jumps between 0 and 1. The 7-day operation log showed a 100% correspondence with the dispatch order.

[0036] The formula for calculating the avoidance demand index D in the second-level dynamic assessment is as follows: D = α × (1 / d p )+β×C+γ×R, Where α, β, and γ are all weighting coefficients, and satisfy α + β + γ = 1. The system defaults to α = 0.4, β = 0.4, and γ = 0.2, which can be adaptively fine-tuned after calibrating 10 conflict scenarios on-site during engineering deployment; C is the tunnel congestion coefficient; R is the chamber availability rate, R = N.free / N total N free N represents the number of currently available chambers. total Let be the total number of chambers, and 0 ≤ R ≤ 1; d p d represents the distance between the vehicle and the nearest person. p The extraction algorithm is to locate the ZigBee cluster to d p The conversion is as follows: a. Deploy ZigBee anchor nodes every 50m underground to form a precision grid of 3-5m; b. The positioning card worn by the personnel sends an RSSI broadcast packet every 500ms. The server obtains the three-dimensional coordinates using trilateration and Kalman filtering: (x p ,y p ,z p ); c. Real-time output of vehicle UWB tag coordinates of the vehicle's front center: (x v ,y v ,z v ); d. Euclidean distance: If d p If the distance is less than 50m and the predicted trajectory intersects after 5 seconds (cubic spline extrapolation), an early warning will be triggered. When ZigBee signal strength fluctuations cause d p When the jitter is greater than 2m, the inertial navigation IMU fusion algorithm is activated until the drift is less than 0.5m / 30s, which meets the 1m safety threshold requirement downhole.

[0037] The formula for calculating the alleyway congestion coefficient C is: C = (N) vehicle / L lane )×(1- V avg / V limit )×λ; Where, N vehicle L represents the current number of vehicles in the roadway section. lane V is the length of the tunnel section. avg V represents the average speed of vehicles within the segment. limit The speed limit for the lane is set, and λ is the vehicle type correction coefficient. Generally, the coefficients are set as follows: heavy-duty vehicles: λ = 1.2, empty vehicles: λ = 1.0, and pedestrians: λ = 0.5. This congestion coefficient model integrates three major factors: vehicle density, speed efficiency, and vehicle type influence, achieving a refined and quantitative perception of lane congestion status, with an accuracy far exceeding that of human experience-based judgment.

[0038] The two-tiered evaluation mechanism overcomes the limitations of a single evaluation dimension. The first tier, static evaluation, is based on the inherent attributes of vehicles, ensuring the basic fairness of the scheduling strategy and guaranteeing the priority right-of-way for high-value transportation tasks. The second tier, dynamic evaluation, introduces environmental variables, enabling the decision-making system to have real-time perception and dynamic response capabilities. It can make scientific and reasonable avoidance decisions based on the complex underground environment, achieving a balance between static fairness and dynamic efficiency.

[0039] like Figure 3 As shown, the dynamic assessment of roadway congestion coefficients is handled in a tiered manner, specifically as follows: Every 200ms, count all vehicles in the current roadway section; When the calculated lane congestion coefficient C > 0.7, the weighting coefficient β is increased by 30%. When the calculated congestion coefficient C of the alleyway is greater than 0.9, the system will forcibly activate the one-way mode, instructing oncoming vehicles in the congested section to enter the nearest vacant chamber to give way.

[0040] This tiered processing enables tiered early warning and rapid response to traffic congestion. When C > 0.7, by increasing the β weight, the system automatically guides vehicles to choose smoother routes, preventing congestion from worsening. When C > 0.9, the mandatory one-way traffic mode is an efficient emergency response mechanism that prioritizes restoring mainline traffic.

[0041] The detailed processing of conflict prediction in dynamic assessment is as follows: When the roadway congestion coefficient C > 0.7, calculate the estimated arrival time t of the first and second vehicles at the conflict point. A and t B , like Then, compare D + 0.2 × C and reorder. If D A >D B Then the second vehicle enters the nearest available chamber; If there is no available chamber, the second vehicle will remain in place with its hazard lights on, and the first vehicle will reduce its speed to 0.8 × V. limit ; like If so, the original state will be maintained.

[0042] In abnormal states, security degradation procedures are implemented, specifically as follows: If valid chamber availability data cannot be obtained within a preset time (e.g., 5 seconds), the system defaults to γ=0 and forces all vehicles to yield to pedestrians, prioritizing personnel safety. When personnel enter a 50-meter radius of a vehicle and their movement trajectories intersect, the onboard terminal's audible and visual alarm and the personnel positioning card's vibration are triggered simultaneously.

[0043] When the system encounters uncertainty due to sensor failure, network interruption, or any other reason, it forces all vehicles to yield to pedestrians, improving the safety level of underground operations and effectively solving the safety risks caused by the large blind spots of heavy vehicles. Specific Implementation The main transport roadway of a coal mine is 2.3km long and 4.5m wide. There are 3 avoidance chambers along the entire line. UWB positioning base stations, personnel positioning beacons, chamber occupancy sensors, wireless communication equipment, etc. have been installed in the roadway. The vehicle attributes of heavy-duty vehicle A are broadcast via the CAN bus. After being parsed by the system, according to W... v The static weight W is calculated as k1 × load capacity + k2 × E. v =0.92, because of its large load capacity and high mission urgency; similarly, the static weight W of the empty vehicle B is calculated. v =0.35; Two vehicles are traveling towards each other at a distance of 300 meters. The system, through real-time processing of vehicle positioning data, predicts a point of conflict in the trajectories of the two vehicles and automatically triggers an avoidance decision process. The system collects and integrates multi-source data in real time to obtain the precise real-time position and velocity vectors of vehicles A and B. It detects that there are maintenance personnel 50 meters behind vehicle B and finds that the chamber in front of vehicle B is idle.

[0045] The avoidance demand assessment model is activated to calculate the dynamic avoidance demand index D for both vehicles. For unloaded vehicle B, the presence of a person 50 meters behind it reduces the value of the distance term between the vehicle and the nearest person, significantly increasing the value of the (1 / distance) term. After weighted calculation, its dynamic index D... B The value was increased by 0.3, resulting in D. B =0.87; For heavy-duty vehicle A, the static weight W v =0.92, and with no other significant adverse factors, its dynamic index D is calculated. A =1.42; therefore, D A ＞ D B The system controls vehicle B, which has a lower avoidance demand index, to perform an avoidance maneuver; it also generates and synchronously issues dynamic avoidance commands. Send instructions to the guidance display screen in the target chamber, such as "Enter chamber #3, vehicle B, reversing XX meters remaining." The guidance screen will light up and display the corresponding information, providing visual guidance to the driver of vehicle B. At the same time, send instructions to the onboard terminal of the target vehicle B, such as "Please proceed to the chamber ahead to avoid it," and issue the instructions in the form of screen animation path guidance to instruct the driver to perform the operation safely and accurately. Send instructions to the personnel positioning card: if personnel are detected entering within 50 meters of the vehicle and are in a potentially risky area, the system will trigger the vibration alarm of the positioning card to remind maintenance personnel to avoid the vehicle.

[0046] Send a command to the on-board terminal of vehicle A with priority passage: display a "Priority passage ahead" prompt, inform the driver of the decision result, and ensure its smooth passage.

[0047] Example 2 An electronic device includes a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the program, it implements the steps of the dynamic avoidance planning method for underground vehicles in coal mines based on multi-source data fusion as described in Embodiment 1.

[0048] Example 3 A computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of the dynamic avoidance planning method for underground vehicles in coal mines based on multi-source data fusion as described in Embodiment 1.

[0049] Based on the above-described preferred embodiments of the present invention, and through the foregoing description, those skilled in the art can make various changes and modifications without departing from the inventive concept. The technical scope of this invention is not limited to the contents of the specification, but must be determined according to the scope of the claims.

Claims

1. A dynamic obstacle avoidance planning method for underground vehicles in coal mines based on multi-source data fusion, characterized in that, include: S1. Real-time acquisition of multi-source data, including vehicle positioning data, vehicle status data, personnel positioning data, chamber occupancy status data, and roadway signal light status data; S2. Construct a collision avoidance demand assessment model. Based on the multi-source data, calculate the collision avoidance demand index for the first and second vehicles traveling in opposite directions. The collision avoidance demand index is obtained through a two-layer assessment mechanism. The two-layer assessment mechanism includes a first-layer static assessment and a second-layer dynamic assessment. The vehicle weight W in the first-layer static assessment... v The calculation formula is: IN v =k1×T+k2×E, Where k1 is the load capacity coefficient, T is the load capacity, k2 is the mission urgency coefficient, and E is the mission urgency. The formula for calculating the avoidance demand index D in the second-level dynamic assessment is as follows: D =α×(1 / d p )+β×C+γ×R, Where α, β, and γ are all weighting coefficients, and satisfy α + β + γ = 1, d p C is the distance between the vehicle and the nearest person, C is the roadway congestion coefficient, and R is the chamber availability rate, R=N free / N total N free N represents the number of currently available chambers. total Let R be the total number of chambers, and 0 ≤ R ≤ 1; The avoidance demand index D is calculated by coupling the weighting coefficient γ with the chamber availability rate R to determine the allocation of target chambers; S3. Compare the avoidance demand index of the first vehicle and the second vehicle, and control the vehicle with the lower avoidance demand index to perform the avoidance action. S4. Generate and issue a dynamic avoidance command. The dynamic avoidance command is simultaneously sent to the vehicle terminal of the target vehicle and the guidance display screen of the target chamber, and triggers the vibration alarm of the personnel positioning card when personnel are detected approaching.

2. The method for dynamic obstacle avoidance planning of underground vehicles in coal mines based on multi-source data fusion according to claim 1, characterized in that: The task urgency E is an 8-bit field written by the production scheduling server and broadcast via the CAN bus when the vehicle departs, and normalized using the formula k2=E / 255; if the received E value is interrupted for more than 3 broadcast cycles, the system defaults to E=0.

3. The method for dynamic obstacle avoidance planning of underground vehicles in coal mines based on multi-source data fusion according to claim 1, characterized in that: The formula for calculating the congestion coefficient C of the alleyway is: C= (N vehicle / L lane )×(1- V avg / V limit )×λ; Where, N vehicle L represents the current number of vehicles in the roadway section. lane V represents the current tunnel segment length. avg V represents the average speed of vehicles within the current roadway section. limit λ represents the current speed limit in the roadway, and λ is the vehicle type correction factor.

4. The method for dynamic obstacle avoidance planning of underground vehicles in coal mines based on multi-source data fusion according to claim 3, characterized in that: Every 200ms, count all vehicles in the current roadway section; When the calculated lane congestion coefficient C > 0.7, the weighting coefficient β is increased by 30%. When the calculated congestion coefficient C of the alleyway is greater than 0.9, the system will forcibly activate the one-way mode, instructing oncoming vehicles in the congested section to enter the nearest vacant chamber to give way.

5. The dynamic obstacle avoidance planning method for underground vehicles in coal mines based on multi-source data fusion according to claim 4, characterized in that: When the roadway congestion coefficient C > 0.7, calculate the estimated arrival time t of the first and second vehicles at the conflict point. A and t B , like Then, compare D + 0.2 × C and reorder. If D A >D B Then the second vehicle enters the nearest available chamber; If there is no available chamber, the second vehicle will remain in place with its hazard lights on, and the first vehicle will reduce its speed to 0.8 × V. limit ; like If so, the original state will be maintained.

6. The method for dynamic obstacle avoidance planning of underground vehicles in coal mines based on multi-source data fusion according to claim 5, characterized in that: In step S3, if valid data on the availability of the chamber cannot be obtained within a preset time, the system defaults to γ=0 and forces all vehicles to give way to pedestrians, prioritizing the safety of personnel.

7. The method for dynamic obstacle avoidance planning of underground vehicles in coal mines based on multi-source data fusion according to claim 1, characterized in that: In step S4, the instructions sent to the chamber guidance display screen include the chamber number, vehicle number, and reversing guidance distance information; the instructions sent to the vehicle's on-board terminal include voice broadcast and on-screen animation path guidance information.

8. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the program, it implements the steps of the dynamic avoidance planning method for underground vehicles in coal mines based on multi-source data fusion as described in any one of claims 1-7.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the program is executed by the processor, it implements the steps of the dynamic avoidance planning method for underground vehicles in coal mines based on multi-source data fusion as described in any one of claims 1-7.

Images