A scenic spot and closed park-oriented automatic driving tour vehicle avoidance method

Abstract

The application discloses a kind of automatic driving sightseeing car avoidance methods for scenic spot and closed park, it is related to vehicle intelligent avoidance field of technology.The application is based on the first emergency stop event trigger, constructs chain avoidance path template and timestamp set, realizes path synchronization and time window offset in subsequent vehicle, to avoid the global traffic stagnation problem caused by local path blockage.Through lightweight broadcast protocol and sequential scheduling mechanism, each sightseeing car can dynamically adjust its own traffic strategy according to the state of the front car, forming a low communication load, decentralized avoidance coordination mechanism.Compared with the traditional avoidance strategy based on collision prediction or brake parameter modeling, the application is based on path topological structure and historical stability, optimizes chain path selection, constructs avoidance queue combining delay time window and traffic fine adjustment, effectively avoids inefficient behaviors such as blind deceleration and repeated planning.Not only has the ability to respond to emergencies, but also suitable for complex intersection section sightseeing car scheduling.

Description

Technical Field

[0001] This invention relates to the field of intelligent vehicle obstacle avoidance technology, and in particular to an obstacle avoidance method for autonomous tour vehicles in scenic areas and closed parks. Background Technology

[0002] With the rapid development of intelligent transportation and autonomous driving technologies, autonomous tour buses, as an important branch of autonomous driving applications in low-speed scenarios, have gradually been commercially deployed in closed or semi-closed locations such as scenic spots, parks, and campuses. To ensure vehicle safety and the passenger experience, how to complete tasks such as emergency avoidance and path adjustment without human intervention has become a key focus of technological research and development. Currently, most common autonomous driving avoidance solutions rely on onboard sensors to perceive obstacles in real time and reconstruct trajectories instantly. However, in closed parks or complex scenic roads, due to limited space, frequent path intersections, and dense vehicle queues, traditional individual avoidance mechanisms suffer from problems such as large response delays, weak local coordination capabilities, and broken avoidance command transmission links. Especially after the first vehicle encounters an emergency and stops, the lack of a rapid linkage mechanism for subsequent vehicles can easily lead to path congestion or chaotic avoidance order, causing localized traffic jams or even secondary safety accidents.

[0003] CN109557925B discloses a method and device for obstacle avoidance in autonomous vehicles. It primarily calculates the vehicle's future trajectory by analyzing yaw angle changes and turning radius, and infers possible collision points and times by combining the relative position information between the vehicle and obstacles. This allows for setting the obstacle avoidance initiation time and decelerating to avoid obstacles. This method focuses on modeling the relative relationship between the vehicle and dynamic obstacles and shows good application results in high-speed or urban road environments. However, this invention mainly relies on the vehicle's own kinematic characteristics for prediction and does not consider the collaborative obstacle avoidance needs arising from path intersections between multiple vehicles. In park scenarios with multiple vehicle platoons and complex path topologies, downstream vehicles lack the ability to perceive the status of the preceding vehicle after the first vehicle stops suddenly, easily leading to delayed obstacle avoidance decisions. Furthermore, this solution does not provide a coordination mechanism for the obstacle avoidance sequence between preceding and following vehicles, failing to construct a systematic obstacle avoidance chain strategy, resulting in insufficient coordination issues such as path overlap and response misalignment.

[0004] CN116729432A discloses an obstacle avoidance system and method for autonomous vehicles. Addressing the issue of relying on a single factor in braking analysis, it collects data on vehicle state changes after each braking maneuver and updates braking performance parameters in real time, thereby optimizing the accuracy of system judgments. This invention improves obstacle avoidance accuracy when vehicle conditions change, especially after active control actions occur, determining whether the vehicle remains in a safe state and intercepting unreasonable control commands when necessary, thus enhancing system safety. However, this method remains a single-vehicle-level perception and decision-making optimization, lacking coordination of obstacle avoidance behaviors among multiple vehicles. In particular, it lacks a unified scheduling mechanism for handling vehicle platooning and path intersection conflicts. Furthermore, its mechanism does not introduce structures such as broadcasting or chained path templates to optimize information transmission efficiency and command consistency, making it difficult to support rapid multi-vehicle obstacle avoidance sequence reconstruction in park-like autonomous driving scenarios. Summary of the Invention

[0005] Given that existing autonomous vehicle obstacle avoidance technologies are mainly geared towards urban roads or highway traffic scenarios, and lack the ability to address issues arising in low-speed, multi-vehicle scenarios, this invention is proposed.

[0006] Therefore, the problem to be solved by this invention is how to solve the problems of untimely scheduling of multi-vehicle sequences and difficulty in resolving path conflicts.

[0007] To solve the above-mentioned technical problems, the present invention provides the following technical solution:

[0008] In a first aspect, the present invention provides a method for autonomous tour vehicles to avoid obstacles in scenic areas and closed parks. The method includes: when any tour vehicle detects an emergency stop condition, it constructs an initial avoidance trigger data packet and encrypts and encapsulates it into a template broadcast structure; the initial avoidance trigger data packet carries a chain-like avoidance path template generated based on the current path topology and timestamp parameters; subsequent tour vehicles receiving the template broadcast structure match the chain-like avoidance path template at their current position and, combined with the timestamp offset from the previous tour vehicle, update their own path parameters with a delay window offset to form an avoidance execution reference path and an initial delay window for the current tour vehicle; within the path intersection area, the tour vehicle constructs a sequential scheduling data packet based on its own initial delay window and the received initial delay window of the preceding vehicle using a lightweight broadcast protocol; each tour vehicle determines its acceptable passage window sequence based on the received multi-vehicle sequential scheduling data packet and locks the passage starting point with the minimum time cost, completing the avoidance sequence cooperative positioning.

[0009] As a preferred embodiment of the autonomous tour vehicle obstacle avoidance method for scenic areas and closed parks described in this invention, the construction of the initial obstacle avoidance trigger data packet includes: extracting the shortest feasible transfer path node sequence extending backward from the emergency stop point based on the spatial topology of the current tour vehicle's trajectory buffer segment and the forward passable area, and constructing a directional topology vector set between path nodes; constructing a time stability map based on the directional topology vector set and the tour vehicle's historical travel stability parameters between nodes, and selecting a path sequence from the shortest feasible transfer path node sequence to generate a chain-like obstacle avoidance path template according to the principle of minimum change rate of stability distribution; determining the timestamp set of the path node sequence in the chain-like obstacle avoidance path template based on the actual length of each path segment in the chain-like obstacle avoidance path template and the current tour vehicle's historical average travel speed; and binding and encapsulating the chain-like obstacle avoidance path template and the timestamp set into an initial obstacle avoidance trigger data packet.

[0010] As a preferred embodiment of the autonomous tour vehicle obstacle avoidance method for scenic areas and closed parks described in this invention, the construction of the shortest feasible transfer path node sequence includes: based on the map node number N0 corresponding to the last trajectory point in the current tour vehicle's trajectory cache segment, constructing a weighted directed graph search model in the spatial topology of the forward passable area, with map node number N0 as the starting point and the nearest valid turning path node as the target, to extract the shortest feasible transfer path node sequence that simultaneously satisfies path transfer legality, valid turning, and path passability; the construction of the chain-like obstacle avoidance path template includes: in the shortest feasible transfer path node sequence, constructing a direction vector set between adjacent path nodes, and calling the passability of the corresponding path segment in the tour vehicle's historical multiple passes. The stability parameters are used to construct a two-dimensional map by binding the direction change angle of each path segment with the traffic stability parameters. By calculating the stability difference between adjacent path segments in the two-dimensional map, the path segment sequence with the smallest stability change is extracted to construct a chain-like avoidance path template. The generation of the timestamp set includes: numbering the path segments between each path node of the chain-like avoidance path template, extracting the actual length set of each path segment, calling the historical average speed of the current tour vehicle in each path segment, and calculating the travel time of the corresponding path segment based on the ratio between the actual length of the path segment and the historical average speed of the tour vehicle in the corresponding path segment; accumulating the travel time segment by segment according to the path segment number and superimposing it on the current emergency stop time to obtain the timestamp value corresponding to each path node and constructing a timestamp set.

[0011] As a preferred embodiment of the autonomous tour vehicle obstacle avoidance method for scenic areas and closed parks described in this invention, the delayed window offset update of its own path parameters includes: after subsequent tour vehicles receive the chained obstacle avoidance path template and timestamp set in the initial obstacle avoidance trigger data packet, they extract matching path segments at the current positioning coordinates, construct a path alignment vector by combining the structural similarity index in the spatial topology, and determine the reference path node index interval that coincides with the starting path segment of the chained obstacle avoidance path template; based on the path node timestamp sequence corresponding to the previous vehicle in the reference path node index interval, they calculate the time window offset of the current vehicle. The offset of the tour vehicle's current position from the start time in the chain avoidance path template is set as the delay start time. Based on the travel time increment corresponding to each path segment in the chain avoidance path template, a set of path segments equidistant from the current position is constructed. Combining the tour vehicle's historical travel offset factor on the current path segment, the expected delay value of the corresponding path segment is corrected, generating a travel window sequence updated with the delay offset. According to the travel window sequence updated with the delay offset, the target travel time of each path segment is sequentially accumulated to the delay start time to form the current tour vehicle's avoidance execution reference path and starting delay window.

[0012] As a preferred embodiment of the autonomous tour vehicle obstacle avoidance method for scenic areas and closed parks described in this invention, the following is defined: the travel time increment is the average travel time between two consecutive path nodes in the chain-like obstacle avoidance path template, statistically analyzed based on historical travel data; the historical travel offset factor is the offset ratio of the current tour vehicle's historical travel time on a specified path segment to the average travel time in the chain-like obstacle avoidance path template; and the correction of the expected delay value of the corresponding path segment is described. The calculations include:

[0013]

[0014] in, α represents the travel time increment for the i-th path segment in the chain template; α is a non-linear adjustment weighting factor. is the historical travel offset factor of the current tour vehicle in segment i; tanh is the hyperbolic tangent function.

[0015] As a preferred embodiment of the autonomous tour vehicle obstacle avoidance method for scenic areas and closed parks described in this invention, the step of constructing a sequential scheduling data packet via a lightweight broadcast protocol includes: extracting the current path segment number and the corresponding delay start time from the chained obstacle avoidance path template, and generating a corresponding path time identifier unit; aligning the received previous tour vehicle path time identifier unit with the path segment number, and calculating the time difference between the delay start times of adjacent path segments as the time offset difference with the previous tour vehicle; based on the time offset difference, and combined with the updated passage window sequence of the current path segment's delay offset, calculating the passage fine-tuning amount of the target path segment within the current tour vehicle's passable window, and encapsulating the passage fine-tuning amount and the vehicle number together into the sequential scheduling data packet; the passage fine-tuning amount is the difference between the current tour vehicle's passage time on the target path segment and the timestamp of the previous tour vehicle on the same path segment, minus the minimum safe passage interval.

[0016] As a preferred embodiment of the autonomous tour vehicle obstacle avoidance method for scenic areas and closed parks described in this invention, the completion of the obstacle avoidance sequence collaborative positioning includes: the tour vehicle summarizing the path segment numbers, corresponding start delay times, and passage fine-tuning amounts from multiple sequential scheduling data packets received, constructing a delay window set at the path segment level, and using it as a candidate passage window sequence; for each candidate passage window, calculating the expected cost of the passage start point based on a nonlinear time cost function; selecting the sequence with the minimum corresponding cost from all candidate passage window sequences, locking the delay start time of the first path segment as the target passage start point of this tour vehicle, and using it as the final start time for the obstacle avoidance execution reference path, thus completing the collaborative positioning.

[0017] In a second aspect, the present invention provides a computer device, including a memory and a processor, wherein the memory stores a computer program, wherein: when the computer program instructions are executed by the processor, the steps of the autonomous tour vehicle avoidance method for scenic areas and closed parks as described in the first aspect of the present invention are implemented.

[0018] Thirdly, the present invention provides a computer-readable storage medium having a computer program stored thereon, wherein: when the computer program instructions are executed by a processor, the steps of the autonomous tour vehicle avoidance method for scenic areas and closed parks as described in the first aspect of the present invention are implemented.

[0019] The beneficial effects of this invention are as follows: Based on the initial emergency stop event, this invention constructs a chain-like avoidance path template and a timestamp set, achieving path synchronization and time window offset in subsequent vehicles, thereby avoiding global traffic stagnation caused by local path congestion. Through a lightweight broadcast protocol and sequential scheduling mechanism, each tour bus can dynamically adjust its own traffic strategy according to the status of the vehicle in front, forming a low-communication-load, decentralized avoidance cooperation mechanism. Compared with traditional avoidance strategies based on collision prediction or braking parameter modeling, this invention optimizes chain-like path selection based on path topology and historical stability, and constructs an avoidance queue by combining delay time windows and traffic fine-tuning, effectively avoiding inefficient behaviors such as blind deceleration and redundant planning. It not only has the ability to respond to emergencies, but is also suitable for tour bus scheduling in complex intersection sections, providing a highly robust autonomous driving cooperative avoidance solution for closed environments such as parks. Attached Figure Description

[0020] To more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0021] Figure 1 A flowchart illustrating the obstacle avoidance method for autonomous tour vehicles operating within scenic areas and closed parks. Detailed Implementation

[0022] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings.

[0023] Many specific details are set forth in the following description in order to provide a full understanding of the invention. However, the invention may also be practiced in other ways different from those described herein, and those skilled in the art can make similar extensions without departing from the spirit of the invention. Therefore, the invention is not limited to the specific embodiments disclosed below.

[0024] Secondly, the term "one embodiment" or "embodiment" as used herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation of the present invention. The phrase "in one embodiment" appearing in different places in this specification does not necessarily refer to the same embodiment, nor is it a single or selective embodiment that is mutually exclusive with other embodiments.

[0025] As mentioned in the background section, most common autonomous driving obstacle avoidance solutions currently rely on onboard sensors to perceive obstacles in real time and reconstruct trajectories instantly. However, in closed parks or complex scenic roads, due to factors such as limited space, frequent path intersections, and dense vehicle queues, traditional individual obstacle avoidance mechanisms suffer from problems such as large response delays, weak local coordination capabilities, and broken obstacle avoidance command transmission links. In particular, after the first vehicle encounters an obstacle and stops abruptly, the lack of a rapid linkage mechanism for subsequent vehicles can easily lead to path congestion or disordered obstacle avoidance order, causing traffic jams in local areas or even secondary safety accidents.

[0026] Figure 1 This is a flowchart illustrating a method for autonomous tour vehicles to avoid obstacles in scenic areas and closed parks, according to an embodiment of the present invention. Figure 1 As shown, the methods for autonomous tour vehicles to avoid obstacles in scenic areas and closed parks include:

[0027] S1: When any vehicle in the tour bus detects an emergency stop condition, it constructs the first avoidance trigger data packet, which is encrypted and encapsulated into a template broadcast structure; the first avoidance trigger data packet carries the chained avoidance path template generated based on the current path topology and the timestamp parameter.

[0028] 1. Based on the spatial topology of the current tour vehicle's trajectory buffer segment and forward passable area, extract the shortest feasible transfer path node sequence extending backward from the emergency stop point, and construct the directional topology vector set between path nodes.

[0029] The construction of the shortest feasible transfer path node sequence includes: based on the map node number N0 corresponding to the last trajectory point in the current tour vehicle's trajectory cache segment, a weighted directed graph search model is constructed in the spatial topology of the forward passable area. This model starts at map node number N0 and targets the nearest valid turning path (e.g., with a U-turn area and fork width sufficient for safe vehicle turning). The model extracts the shortest feasible transfer path node sequence that simultaneously satisfies path transfer legality, valid turning, and path width tolerance. Each edge in the graph corresponds to an actual travel path, and the weight of each edge is calculated based on a comprehensive weighted average of path width, safety level, and historical travel frequency (the weight calculation can be set according to actual conditions; the above is just an example).

[0030] The legality of a path transfer refers to whether, when the current tour vehicle moves from its current location to the next path node, the path connection has a clear boundary in the map topology, and whether the transfer does not violate road traffic rules (such as one-way traffic, turning restrictions, or terrain closures). In other words, the path transfer operation must conform to the definition of the geographic topology and be actually accessible to the vehicle.

[0031] Effective turning refers to the space and trajectory feasibility for a tour bus to perform a turning maneuver when adjusting its direction at a path node (such as making a U-turn, turning, or switching to an alternate route).

[0032] Path width tolerance refers to the compatibility between the current tour vehicle's width and the minimum effective passage width of the path segment, determining whether vehicles can pass without collision. Generally, the effective width of the path segment should be greater than or equal to the maximum outer width of the vehicle plus a safety margin (e.g., 0.3 meters on each side). If the path is too narrow, it is deemed incompatible and cannot be added to the sequence of transferable path nodes.

[0033] This invention employs an improved Dijkstra shortest path search algorithm to perform feasibility testing and path topology continuity analysis on all paths, and selects the shortest feasible transfer path node sequence in the path transfer structure that has legality of passage, path width that meets the passage requirements, and target with effective turnaround capability.

[0034] The path node sequence obtained in the above manner will serve as the basic structure for constructing the subsequent directional topology vector set.

[0035] 2. Based on the directional topology vector set, and combined with the historical stability parameters of the sightseeing vehicle between nodes, a time stability map is constructed. According to the principle of minimum change rate of stability distribution, a path sequence is selected from the shortest feasible transfer path node sequence to generate a chain-like avoidance path template.

[0036] The construction of the chained obstacle avoidance path template includes:

[0037] In the shortest feasible transfer path node sequence, a direction vector set is constructed between adjacent path nodes (each direction vector consists of the starting path node and the ending path node, and direction data is generated based on the two-dimensional spatial coordinate difference; multiple direction vectors together constitute a direction vector set). The travel stability parameters of the corresponding path segments in the tour vehicle's multiple historical passages are called (including but not limited to the variance of the actual travel time of the path segment, braking frequency, success rate of passage, ratio of sharp turns within the path segment, and number of obstructions, etc.). The direction change angle (vector angle) of each path segment is bound to the travel stability parameters to construct a two-dimensional map.

[0038] By calculating the stability difference between adjacent path segments in the two-dimensional map, the path segment sequence with the smallest stability change is extracted, and a chain-like avoidance path template is constructed.

[0039] Based on the above map analysis results, the set of path segments with the smallest rate of change in the stability distribution is selected as the backbone path of the chain-like obstacle avoidance path template. This template possesses characteristics such as continuity, directional consistency, and good historical traffic performance, effectively avoiding the shortcomings of traditional shortest path methods that may select paths with sharp turns or frequent congestion. The final output chain-like obstacle avoidance path template is a sequence of path segments that is structurally continuous, directionally coordinated, and historically controllable.

[0040] 3. Determine the set of timestamps for the path node sequence in the chain avoidance path template based on the actual length of each path segment in the chain avoidance path template and the historical average speed of the current tour vehicle.

[0041] The generation of the timestamp set includes:

[0042] The path segments between each path node in the chain-like obstacle avoidance path template are numbered. The actual length set of each path segment is extracted. The historical average speed of the current tour vehicle in each path segment is retrieved. Based on the ratio between the actual length of the path segment and the historical average speed of the tour vehicle in the corresponding path segment, the travel time of the corresponding path segment is calculated. The travel time is accumulated segment by segment according to the path segment number and superimposed on the current emergency stop time to obtain the timestamp value corresponding to each path node, thus constructing a timestamp set. For example, if the emergency stop trigger time is taken as the starting time T0 of the timestamp set, then T1 = T0 + Δt1, T2 = T1 + Δt2, where Δt1 and Δt2 are the required travel times for the first and second path segments, respectively; T1 and T2 are the timestamps of the tour vehicle's expected arrival at path nodes 1 and 2, respectively; and so on, until the timestamps of all path segments are generated.

[0043] 4. Bind and encapsulate the chained avoidance path template and timestamp set into the first avoidance trigger data packet.

[0044] Specifically, after completing the construction of the path structure template and timestamp set, the first vehicle binds and encapsulates these two sets of data to generate the first-vehicle avoidance trigger data packet. The encapsulation format of the first-vehicle avoidance trigger data packet adopts a two-level structure, where the first level is the path structure field and the second level is the timestamp field. The structure supports CRC checksums, has data integrity verification capabilities, and avoids information loss or misalignment during communication.

[0045] To prevent data leakage and tampering during transit, the data packet content is encrypted using a lightweight symmetric encryption algorithm, and the key can be periodically updated by the park's main control system.

[0046] S2: The subsequent tour vehicle receiving the template broadcast structure matches the chained avoidance path template at the current position, and combines it with the timestamp offset of the previous tour vehicle to update its own path parameters with a delay window offset, forming the avoidance execution reference path and the starting delay window of the current tour vehicle.

[0047] Once the first tour vehicle completes the construction of the first avoidance trigger data packet, which includes a chain-like avoidance path template and a set of path node timestamps, and propagates it through the regional broadcast structure, any subsequent tour vehicle within that path area must respond to the broadcast data structure in order to reconstruct the passage sequence of vehicles and synchronize the avoidance plan throughout the entire area.

[0048] S2.1: After the subsequent tour vehicle receives the chained avoidance path template and timestamp set in the initial avoidance trigger data packet, it extracts the matching path segment at the current positioning coordinates, constructs the path alignment vector by combining the structural similarity index in the spatial topology, and determines the reference path node index range that coincides with the starting path segment of the chained avoidance path template.

[0049] It should be noted that traditional path alignment techniques typically rely on absolute coordinate projection or simple node adjacency matching, which struggles to handle situations where path direction deviations or node overlaps are not significant in complex scenarios. This invention enhances the accuracy and fault tolerance of path matching by using a spatial structure similarity index based on directional topology.

[0050] Specifically, the tour vehicle first extracts the map node number corresponding to its current position, combines it with the direction vector formed by the current forward path direction of the vehicle, and matches the cosine value of the angle between the direction vectors of each path segment in the chain avoidance path template to construct a set of path alignment vectors.

[0051] The path alignment vector set is used to quantify the directional structural consistency between the current vehicle's path segment and the chain template path segment. The alignment cosine values ​​of all path segments are sorted, and the index value corresponding to the path segment with the smallest directional angle is extracted as the starting node index for path structure matching.

[0052] When the path direction changes drastically or the angle between path segments approaches 90°, the cosine of the angle approaches zero. To avoid path matching loss, the continuity of path segment numbers and the continuity of geometric node positions provide redundant judgment criteria, thereby improving the robustness and accuracy of the path alignment mechanism in intersection areas or right-angle path segment scenarios.

[0053] Based on this, a path node index interval is constructed by searching a set of consecutively similar path segments in the chain template. The path node index interval represents a path region that can be continuously aligned in the chain template starting from the current vehicle position, and is an important reference for subsequent calculation of temporal offset.

[0054] The method of constructing path node index intervals by means of the present invention can effectively solve the matching failure problem caused by differences in map sampling accuracy and slight differences in path structure.

[0055] S2.2: Based on the timestamp sequence of the path nodes corresponding to the preceding vehicle in the reference path node index interval, calculate the offset of the current tour vehicle's current position from the start time in the chain avoidance path template, and set the offset as the delayed start time.

[0056] Specifically, after extracting the path node index range, the timestamp set in the initial avoidance trigger data packet is called, and based on the correspondence between the path nodes of the preceding vehicle and the current path nodes, the node timestamp subsequence of the preceding vehicle in the path index range is extracted. The starting timestamp in the node timestamp subsequence indicates the estimated travel time of the first tour vehicle at the starting node of the corresponding path segment.

[0057] The tour bus then uses its own location data to obtain the current time and calculates the difference between the current time and the start time of the timestamp subsequence to obtain the time offset.

[0058] The above time offset is defined as the initial delay start time of the current tour vehicle, representing the earliest acceptable time point for subsequent tour vehicles to initiate avoidance scheduling within the same structural path. This avoids the shortcomings of traditional static sorting that arbitrarily sets a fixed delay time for subsequent vehicles. By accurately matching the path structure index and calculating the difference between the timestamp and the target time, it achieves dynamic and personalized assignment of the avoidance scheduling start point.

[0059] S2.3: Based on the travel time increment corresponding to each path segment in the chain avoidance path template, construct a set of path segments equidistant from the current position. Combine the historical travel offset factor of the sightseeing vehicle on the current path segment, correct the expected delay value of the corresponding path segment, and generate the travel window sequence after delay offset update.

[0060] The passage time increment is the average passage time between two consecutive path nodes in the chain avoidance path template. It is based on historical passage data and is obtained by statistically analyzing the average passage time recorded by the first vehicle for the path segment in multiple historical operating cycles, thus possessing path-level stability.

[0061] Among them, the historical travel offset factor is the offset ratio of the current tour vehicle's historical travel time and the average travel time in the chain avoidance path template on the specified path segment.

[0062] Among them, the expected delay value of the corresponding path segment is corrected. The calculations include:

[0063]

[0064] in, α represents the travel time increment for the i-th path segment in the chain template; α is a non-linear adjustment weighting factor. is the historical travel offset factor of the current tour vehicle in segment i; tanh is the hyperbolic tangent function.

[0065] S2.4: Based on the updated passage window sequence after the delay offset, the target passage time of each path segment is sequentially added to the delay start time to form the current tour vehicle's avoidance execution reference path and starting delay window.

[0066] S3: Within the path intersection area, the tour vehicle constructs sequential scheduling data packets using the Lightweight Broadcast protocol based on its own starting delay window and the starting delay window of the preceding vehicle.

[0067] It should be noted that when multiple tour vehicles are scheduled to pass through the intersection area according to the chain-like avoidance path template, a broadcast scheduling mechanism is needed to coordinate the updating of the passage windows of each vehicle in order to ensure the sequentiality, stability, and spatial safety of the passage process. Traditional broadcast mechanisms suffer from large data volumes and high transmission delays, making it difficult to meet the rapid synchronization requirements of scheduling data in high-density tour scenarios. To solve this problem, this invention designs a lightweight broadcast protocol to construct a minimum effective passage control data unit structure. Using path segments as the basic unit and passage time offset and fine-tuning as the core content, it achieves efficient synchronization of the sequential scheduling information of tour vehicles within the intersection area.

[0068] S3.1: Extract the current path segment number and the corresponding delay start time from the chained avoidance path template, and generate the corresponding path time identifier unit.

[0069] Specifically, the tour vehicle first extracts the set of path segment numbers from the current chain avoidance path template and locates the path segment number to which the current location point belongs as the core structure of the path time identifier unit; then, it extracts the delay start time corresponding to the path segment number from the passage time window sequence after it has completed the delay offset update, and binds the two to generate the path time identifier unit.

[0070] The path time identifier unit is a core component of the minimized scheduling broadcast data structure. Each vehicle only needs to construct a single identifier unit for the current path segment, without transmitting full path information, thus significantly reducing data broadcast duration and channel occupancy. This structure can accurately represent the current spatiotemporal state of a vehicle within the path intersection area, possesses good decoding and recognizability, and provides a basic reference for priority comparison in subsequent sequential scheduling.

[0071] S3.2: Align the received previous tour vehicle path time identifier unit with the path segment number, and calculate the time difference between the delayed start time of adjacent path segments as the time offset difference with the previous tour vehicle.

[0072] Once the tour bus has completed the construction of its own path time identifier unit, it receives the same path time identifier unit broadcast by the tour bus in the preceding position via the Lightweight Broadcast Protocol.

[0073] To achieve sequential planning and coordination on the path segments, the current tour vehicle needs to align the path segment number of the previous tour vehicle with its own current path segment number to determine whether they are consecutive passable segments within the same chain-like obstacle avoidance path template structure.

[0074] If the routes are not adjacent or continuous, the current tour bus will terminate the subsequent calculations and wait for route replanning.

[0075] If the numbering is aligned, the starting time of the previous vehicle's delay and the starting time of the current vehicle's own delay are extracted. The starting time of the delay indicates the degree of delay of the current vehicle relative to the preceding vehicle's travel plan on the same path segment, and is the basic parameter for dynamically establishing the priority ranking of yielding time. Through this parameter, it is possible to assess whether there are potential traffic conflicts or time overlaps for the current vehicle, thereby taking measures to fine-tune the time or rearrange the traffic order in advance.

[0076] S3.3: Based on the time offset difference and combined with the updated passage window sequence of the current path segment's delay offset, calculate the passage fine-tuning amount of the target path segment within the current tour vehicle's passage window, and encapsulate the passage fine-tuning amount and vehicle number together into the sequential scheduling data packet.

[0077] Specifically, in order to achieve high-precision timing coordination in the avoidance queue, the current tour vehicle, based on the delayed start time in the path time identifier of the previous tour vehicle and combined with its own delayed offset passage window sequence, fine-tunes the passage timing of the current path segment and generates path passage fine-tuning amount.

[0078] The passage fine-tuning amount is the difference between the current tour bus's passage time on the target route segment and the timestamp of the previous tour bus on the same route segment, minus the minimum safe passage interval.

[0079] The passage fine-tuning amount represents the amount by which the current tour bus needs to delay (if the passage fine-tuning amount is less than 0) or advance (if the passage fine-tuning amount is greater than 0) the time required based on the passage window of the preceding tour bus. When the passage fine-tuning amount is less than the preset adjustment threshold (usually 0), the fine-tuning delay compensation operation is automatically triggered, which delays the start time of the current tour bus's target path segment by the value of the passage fine-tuning amount to avoid conflicts with the preceding tour bus on the path segment.

[0080] Accordingly, the start time of passage for subsequent path segments is delayed synchronously to ensure consistency of the obstacle avoidance sequence throughout the entire path.

[0081] Furthermore, after generating the current vehicle path time identifier unit, calculating the time offset difference, and constructing the traffic fine-tuning amount, these structured scheduling contents are packaged and encapsulated into a sequential scheduling data packet. This sequential scheduling data packet is then periodically broadcast to the local path intersection area via a lightweight broadcast protocol. The lightweight broadcast protocol is a fixed-length structured broadcast mechanism constructed in this invention, employing a fixed-field byte structure and encoding compression mechanism, enabling data frame formats smaller than 64 bytes, significantly reducing channel occupancy and collision probability.

[0082] Upon receiving the dispatch data packet, the receiving vehicle immediately updates its passage order, reconstructs its passage path, or adjusts its waiting strategy, forming a self-organizing cooperative obstacle avoidance network. This mechanism enables multi-vehicle sequential cooperative passage within path intersection areas with minimal communication overhead, providing stable and efficient basic communication capabilities for a park-level barrier-free dispatch system.

[0083] S4: Each tour bus determines its acceptable passage window sequence based on the received multi-vehicle sequence scheduling data packet, and locks the passage starting point with the minimum time cost to complete the avoidance sequence cooperative positioning.

[0084] First, the tour bus aggregates the path segment number, corresponding start delay time, and passage fine-tuning amount from multiple sequential scheduling data packets it receives, constructs a set of delay windows at the path segment level, and uses it as a candidate passage window sequence.

[0085] The delay window set records the potential traffic occupancy of the current route segment within the local route network, serving as the boundary for the tour bus to formulate its own travel plan. If the delay window set for a certain route segment is dense or has a high degree of time overlap, it indicates that there is a risk of congestion on that route segment in the short term, and the tour bus should avoid it by delaying or adjusting its route starting point.

[0086] Secondly, for each candidate passage window, the expected cost of the passage starting point is calculated based on the nonlinear time cost function.

[0087] The better nonlinear time cost C i The calculation process of the function is as follows:

[0088]

[0089] Where M is; ΔD i,j γ is the delay offset time of the current tour bus on the j-th path segment; γ is a nonlinear adjustment factor that adjusts the sensitivity of the bus's willingness to avoid obstacles to the degree of path congestion; δ i,j The time difference between the overlapping delay windows of the preceding vehicle and the current vehicle on the j-th path segment is denoted as .

[0090] Finally, the sequence with the lowest cost is selected from all candidate passage window sequences, and the start time of the delay of the first path segment is locked as the target passage start point of this tour vehicle, which serves as the final start time for the avoidance execution reference path, thus completing the cooperative positioning.

[0091] The delay start time is the starting point of the avoidance execution reference path that the current tour vehicle finally locks onto, and it is the starting reference point for the passage sequence of the entire path segment.

[0092] Once the moment is confirmed, the tour bus's internal scheduler is automatically updated, and the power system startup preparation, route segment switching detection, and control command scheduling are advanced according to the delayed start time to ensure that the tour bus can enter the queue of the chain avoidance path template in an orderly and safe manner without violating the window occupancy of any previously passing vehicles.

[0093] This embodiment also provides a computer device applicable to the obstacle avoidance method for autonomous tour vehicles in scenic areas and closed parks, including a memory and a processor; the memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions to realize the obstacle avoidance method for autonomous tour vehicles in scenic areas and closed parks as proposed in the above embodiment.

[0094] The computer device can be a terminal, comprising a processor, memory, communication interface, display screen, and input devices connected via a system bus. The processor provides computing and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system and computer programs. The internal memory provides an environment for the operation of the operating system and computer programs stored in the non-volatile storage media. The communication interface is used for wired or wireless communication with external terminals; wireless communication can be achieved through Wi-Fi, carrier networks, NFC (Near Field Communication), or other technologies. The display screen can be an LCD screen or an e-ink screen. The input devices can be a touch layer covering the display screen, buttons, a trackball, or a touchpad on the computer device's casing, or an external keyboard, touchpad, or mouse.

[0095] This embodiment also provides a storage medium storing a computer program that, when executed by a processor, implements the method for autonomous tour vehicles to avoid obstacles in scenic areas and closed parks as proposed in the above embodiments.

[0096] In summary, this invention, triggered by the initial emergency stop event, constructs a chain-like avoidance path template and a timestamp set, enabling path synchronization and time window offset in subsequent vehicles, thereby avoiding global traffic congestion caused by local path blockages. Through a lightweight broadcast protocol and sequential scheduling mechanism, each tour bus can dynamically adjust its own traffic strategy based on the status of the vehicle ahead, forming a low-communication-load, decentralized avoidance coordination mechanism. Compared to traditional avoidance strategies based on collision prediction or braking parameter modeling, this invention optimizes chain-like path selection based on path topology and historical stability, combining delay windows and traffic fine-tuning to construct an avoidance queue, effectively avoiding inefficient behaviors such as blind deceleration and redundant planning. It not only possesses emergency response capabilities but is also suitable for tour bus scheduling in complex intersection sections, providing a highly robust autonomous driving cooperative avoidance solution for closed environments such as parks.

[0097] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention, and all such modifications or substitutions should be covered within the scope of the claims of the present invention.

Claims

1. A method for autonomous tour vehicles to avoid obstacles in scenic areas and closed parks, characterized in that: include: When any of the tour buses detects an emergency stop condition, a first avoidance trigger data packet is constructed and encrypted and encapsulated into a template broadcast structure. The first avoidance trigger data packet carries a chain avoidance path template and a timestamp set generated based on the current path topology; The path segments between each path node in the chain avoidance path template are numbered, the actual length set of each path segment is extracted, the historical average speed of the current tour vehicle in each path segment is called, and the travel time of the corresponding path segment is calculated based on the ratio between the actual length of the path segment and the historical average speed of the tour vehicle in the corresponding path segment. The travel time is accumulated segment by segment according to the path segment number and superimposed on the current emergency stop time to obtain the timestamp value corresponding to each path node and construct a timestamp set. The subsequent tour vehicle that receives the template broadcast structure matches the chain avoidance path template at the current position, and updates its own path parameters with the delay window offset based on the timestamp offset of the previous tour vehicle, thus forming the avoidance execution reference path and the starting delay window of the current tour vehicle. Within the path intersection area, the tour bus constructs sequential scheduling data packets based on its own start delay window and the start delay window of the preceding vehicle using the Lightweight Broadcast protocol; The tour bus aggregates the path segment numbers, corresponding start delay times, and passage fine-tuning amounts from multiple sequential scheduling data packets it receives, constructs a set of delay windows at the path segment level, and uses them as candidate passage window sequences. For each candidate passage window, the expected cost of the passage start point is calculated based on a nonlinear time cost function. The sequence with the minimum corresponding cost is selected from all candidate passage window sequences, and the delay start time of the first path segment is locked as the current tour bus's target passage start point, which serves as the final start time for the avoidance execution reference path, thus completing the cooperative positioning.

2. The method for autonomous tour vehicles to avoid obstacles in scenic areas and closed parks as described in claim 1, characterized in that: The construction of the initial avoidance trigger data packet includes: Based on the current trajectory buffer segment and the spatial topology of the forward passable area of ​​the tour vehicle, the shortest feasible transfer path node sequence extending backward from the emergency stop point is extracted, and the directional topology vector set between the path nodes is constructed. Based on the aforementioned directional topology vector set, a time stability map is constructed by combining the historical travel stability parameters of the sightseeing vehicle between nodes. According to the principle of minimum change rate of stability distribution, a path sequence is selected from the shortest feasible transfer path node sequence to generate a chain-like avoidance path template. The chained avoidance path template and timestamp set are bound and encapsulated into the first avoidance trigger data packet.

3. The method for autonomous tour vehicles to avoid obstacles in scenic areas and closed parks as described in claim 2, characterized in that: The construction of the shortest feasible transfer path node sequence includes: Based on the map node number corresponding to the last trajectory point in the current tour vehicle's trajectory cache segment. Within the spatial topology of the forward-accessible area, a map node numbering system is constructed. A weighted directed graph search model with the starting point as the origin and the destination as the nearest valid turning path node is used to extract the shortest feasible transition path node sequence that simultaneously satisfies the path transition legality, valid turning, and path width tolerance. The construction of the chain-like obstacle avoidance path template includes: In the shortest feasible transfer path node sequence, a direction vector set is constructed between adjacent path nodes, and the travel stability parameters of the corresponding path segments in the multiple historical passages of the sightseeing vehicle are called. The direction change angle and travel stability parameters of each path segment are bound to construct a two-dimensional map. By calculating the stability difference between adjacent path segments in the two-dimensional map, the path segment sequence with the smallest stability change is extracted to construct a chain avoidance path template.

4. The method for autonomous tour vehicles to avoid obstacles in scenic areas and closed parks as described in claim 1, characterized in that: The delayed window offset update of its own path parameters includes: After the subsequent sightseeing vehicle receives the chained avoidance path template and timestamp set in the first avoidance trigger data packet, it extracts the matching path segment at the current positioning coordinates, constructs the path alignment vector by combining the structural similarity index in the spatial topology, and determines the reference path node index interval that coincides with the starting path segment of the chained avoidance path template. Based on the timestamp sequence of the path nodes corresponding to the preceding vehicle in the reference path node index interval, calculate the offset of the current tour vehicle's current position from the start time in the chain avoidance path template, and set the offset as the delayed start time; Based on the passage time increment corresponding to each path segment in the chain avoidance path template, a set of path segments equidistant from the current position is constructed. Combined with the historical passage offset factor of the sightseeing vehicle on the current path segment, the expected delay value of the corresponding path segment is corrected, and the passage window sequence after delay offset update is generated. Based on the updated passage window sequence after the delay offset, the target passage time of each path segment is sequentially added to the delay start time to form the current tour vehicle's avoidance execution reference path and starting delay window.

5. The method for autonomous tour vehicles to avoid obstacles in scenic areas and closed parks as described in claim 4, characterized in that: The passage time increment is the average passage time between two consecutive path nodes in the chain avoidance path template, which is statistically based on historical passage data. The historical travel offset factor is the offset ratio of the current tour vehicle's historical travel time and the average travel time in the chain avoidance path template on the specified path segment. The expected delay value of the corresponding path segment is corrected. The calculations include: in, For the first in the chain template Increment of travel time for a route segment; The weighting factor is adjusted non-linearly. For the current tour bus in the Historical traffic offset factor for the segment; It is the hyperbolic tangent function.

6. The method for autonomous tour vehicles to avoid obstacles in scenic areas and closed parks as described in claim 1, characterized in that: The construction of sequentially scheduled data packets via the Lightweight Broadcast Protocol includes: Extract the current path segment number and the corresponding delay start time from the chained obstacle avoidance path template, and generate the corresponding path time identifier unit; Align the received path time identifier units of the previous tour bus with the path segment numbers, and calculate the time difference between the delayed start times of adjacent path segments as the time offset difference with the previous tour bus. Based on the time offset difference, combined with the updated passage window sequence after the delay offset of the current path segment, the passage fine-tuning amount of the target path segment within the current passage window of the tour vehicle is calculated, and the passage fine-tuning amount and the vehicle number are encapsulated together into the sequential scheduling data packet. The passage fine-tuning amount represents the extent to which the current tour bus needs to delay or advance the passage time based on the passage window of the previous tour bus.

7. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that: When the processor executes the computer program, it implements the steps of the autonomous tour vehicle avoidance method for scenic areas and closed parks as described in any one of claims 1 to 6.

8. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by the processor, it implements the steps of the autonomous tour vehicle avoidance method for scenic areas and closed parks as described in any one of claims 1 to 6.

Images