Vehicle positioning method based on grey wolf role and sensor mapping and related equipment

Abstract

The invention provides a vehicle positioning method and related equipment based on grey wolf role and sensor mapping, and relates to the technical field of vehicle positioning, and the method comprises the steps: constructing a physically constrained position feasible region when a GPS signal is abnormal, and introducing a hierarchical structure of a grey wolf optimization algorithm to achieve the collaborative fusion of multi-sensor data, thereby improving the positioning accuracy of a vehicle. Wherein the alpha wolf carries out dead reckoning by utilizing inertia and wheel speed information, the beta wolf combines environment perception and a high-precision map to generate a matching position, the delta wolf dynamically evaluates and adjusts confidence weights of the alpha wolf and the beta wolf, the omega wolf is guided to carry out local fine search in a high-confidence area, and finally, rapid convergence is carried out according to a road semantic feature matching degree and a high-precision positioning result is output. And the positioning accuracy, the robustness and the real-time performance of the vehicle in a GPS-free or weak-signal environment are remarkably improved.

Description

Technical Field

[0001] This invention relates to the field of vehicle positioning technology, and in particular to a vehicle positioning method and related equipment based on the mapping of the gray wolf role to sensors. Background Technology

[0002] In complex environments such as urban canyons and tunnels, GPS signals are often weakened or even completely lost due to obstruction or interference. This exposes the inherent weakness of traditional vehicle positioning technology, which heavily relies on external satellite signals. Problems such as decreased positioning accuracy, trajectory drift, and response delays occur, which are particularly pronounced under extreme weather or strong electromagnetic interference conditions, seriously affecting driving safety and the reliability of navigation systems. Existing technologies mostly use inertial navigation or single sensors for assisted positioning, but they generally suffer from shortcomings such as rapid error accumulation, poor environmental adaptability, and a lack of effective fusion mechanisms for multi-source information. As a result, it is difficult to maintain high-precision and high-stability positioning output in the absence of GPS signals for extended periods. Summary of the Invention

[0003] This invention aims to solve the technical problems existing in the above-mentioned related technologies, and proposes a vehicle positioning method and related equipment based on gray wolf role and sensor mapping. When GPS signal is lost or degraded, it can achieve high-precision and robust continuous vehicle positioning by simulating the hierarchical cooperation mechanism of gray wolf pack and fusing multi-source sensor data.

[0004] The solution to the technical problem of this invention is as follows: This invention provides a vehicle localization method based on the mapping between the gray wolf role and sensors, comprising the following steps: Acquire inertial measurement unit data, wheel speed sensor data, and environmental perception data of the vehicle at the current moment; When GPS signal loss or signal quality is detected to be below a preset threshold, the feasible motion trajectory of the vehicle within a preset time window is predicted based on the vehicle kinematics model, and a physically constrained location feasible region is generated. A gray wolf population is initialized within the feasible region of the location, and the population is divided into α wolves, β wolves, δ wolves and ω wolves according to the social hierarchy of the gray wolves; The α wolf is configured to perform dead reckoning by fusing data from the inertial measurement unit and wheel speed sensor to generate a first candidate position; The β wolf is configured to match the environmental perception data with the high-precision map to generate a second candidate location; The δ wolf is configured to evaluate the consistency between the first candidate position and the second candidate position, and dynamically adjust their respective confidence weights based on the consistency results; The position of each wolf in the gray wolf population is updated based on the confidence weight, and the ω wolf performs a local dense search in the weighted high confidence region; During the search process, if the matching degree between multiple candidate positions of ω wolves and road semantic features exceeds a preset matching threshold, the iteration is terminated early, and the candidate position with the highest matching degree is output as the final localization result of the vehicle.

[0005] Furthermore, the vehicle kinematic model includes a maximum steering angle constraint, a longitudinal acceleration limit, and a trajectory boundary determined by the wheel speed integral, wherein the feasible position region is a three-dimensional spatial region that simultaneously satisfies the constraints.

[0006] Furthermore, the environmental perception data includes visual images or lidar point cloud data, and the high-precision map contains geometric and semantic information of lane lines, curbs, traffic signs, and parking areas.

[0007] Furthermore, the road semantic features include lane line continuity, curb geometry, traffic sign location, or parking area boundary. The matching degree is obtained by calculating the spatial overlap or structural similarity between the multi-sensor features at the candidate location and the corresponding semantic elements in the high-precision map.

[0008] Furthermore, during the frame-by-frame localization of the vehicle, the optimal gray wolf position obtained from the previous frame search is retained, and a trajectory smoothness energy function is constructed based on the candidate positions of the current frame and the previous frame. The trajectory smoothness energy function is used as the regularization term of the fitness function to penalize candidate trajectories with sudden acceleration changes or discontinuous path curvature.

[0009] Furthermore, when the confidence level of the environmental perception data is lower than a preset perception threshold, the weight of the α wolf is dynamically increased and the weight of the β wolf is decreased.

[0010] On the other hand, this application provides a vehicle positioning system based on the mapping of gray wolf roles and sensors, the system comprising: The data acquisition module is used to acquire data from the vehicle's inertial measurement unit, wheel speed sensor, and environmental perception. The signal detection module is used to determine whether the GPS signal is lost or its quality is below a preset threshold. The feasible region generation module is used to predict the feasible motion trajectory of the vehicle within a preset time window based on the vehicle kinematics model when the GPS signal is abnormal, and generate a physically constrained location feasible region. The gray wolf initialization module is used to initialize a gray wolf population within the feasible region of the location and to divide the population into α wolves, β wolves, δ wolves and ω wolves according to the social hierarchy of the gray wolves. The role mapping module is used for: The α wolf is configured to perform dead reckoning by fusing data from the inertial measurement unit and wheel speed sensor to generate a first candidate position; The β wolf is configured to match the environmental perception data with the high-precision map to generate a second candidate location; The δ wolf is configured to evaluate the consistency between the first candidate position and the second candidate position, and dynamically adjust their respective confidence weights based on the consistency results; The collaborative search module is used to update the position of each wolf in the gray wolf population based on the confidence weight, and drive the ω wolf to perform a local dense search in the weighted high confidence region; The semantic verification and output module is used to terminate the iteration in advance when the matching degree between multiple candidate positions of ω wolves and road semantic features exceeds a preset matching threshold during the search process, and output the candidate position with the highest matching degree as the final positioning result of the vehicle.

[0011] Furthermore, the role mapping module is further configured to: dynamically increase the weight of α wolf and decrease the weight of β wolf when the confidence level of the environmental perception data is lower than a preset perception threshold.

[0012] On the other hand, this application provides an electronic device, which includes a memory and a processor. The memory stores a computer program, and the processor executes the computer program to implement the aforementioned vehicle positioning method based on the mapping of gray wolf roles and sensors.

[0013] On the other hand, this application provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the aforementioned vehicle localization method based on the mapping between the gray wolf character and sensors.

[0014] The beneficial effects of this invention are as follows: This application provides a vehicle positioning method based on the mapping of gray wolf roles and sensors. This method constructs a physically constrained feasible region when GPS signals are abnormal, and introduces a hierarchical structure of the gray wolf optimization algorithm to achieve collaborative fusion of multi-sensor data. Specifically, the α wolf uses inertia and wheel speed information for dead reckoning, the β wolf combines environmental perception and high-precision maps to generate a matching position, and the δ wolf dynamically evaluates and adjusts the confidence weights of both, guiding the ω wolf to conduct a local fine-grained search in a high-confidence area. Finally, based on the matching degree of road semantic features, the method quickly converges and outputs a high-precision positioning result, significantly improving the positioning accuracy, robustness, and real-time performance of vehicles in environments without GPS or with weak signals. This application also provides corresponding systems, devices, and media. The beneficial effects of the systems, devices, and media are the same as the technical effects of the above method, and will not be elaborated upon here.

[0015] Other features and advantages of this application will be set forth in the description which follows, and will be apparent in part from the description, or may be learned by practicing the application. The objectives and other advantages of this application may be realized and obtained by means of the structures particularly pointed out in the description, claims and drawings. Attached Figure Description

[0016] The accompanying drawings are provided to further understand the technical solutions of the present invention and constitute a part of the specification. They are used together with the embodiments of the present invention to explain the technical solutions of the present invention, and do not constitute a limitation on the technical solutions of the present invention.

[0017] Figure 1 This is a flowchart of the vehicle localization method based on the mapping between the gray wolf role and sensors provided in this application; Figure 2 This is a structural diagram of the vehicle positioning system based on the mapping between the gray wolf role and sensors provided in this application. Detailed Implementation

[0018] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application.

[0019] The present application will be further described below with reference to the accompanying drawings and specific embodiments. The described embodiments should not be considered as limitations on the present application, and all other embodiments obtained by those skilled in the art without inventive effort are within the scope of protection of the present application.

[0020] In the following description, references are made to “some embodiments,” which describe a subset of all possible embodiments. However, it is understood that “some embodiments” may be the same subset or different subsets of all possible embodiments and may be combined with each other without conflict.

[0021] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application belongs. The terminology used herein is for the purpose of describing embodiments of this application only and is not intended to limit this application.

[0022] With the development of intelligent driving and advanced driver assistance systems (ADAS), the demand for high-precision and high-reliability positioning capabilities for vehicles is becoming increasingly urgent. Under ideal conditions, the Global Positioning System (GPS) can provide relatively accurate location information and is widely used in navigation and route planning. However, in real-world road environments, especially in complex scenarios such as urban canyons, tunnels, underground parking lots, tree-lined roads, or in adverse weather conditions, GPS signals are highly susceptible to factors such as building obstruction, multipath effects, and electromagnetic interference, resulting in signal attenuation, abrupt changes, or even complete interruption. In such cases, positioning systems relying solely on GPS will fail to function properly, leading to vehicle position estimation failures or severe drift, thereby affecting the safety and stability of critical functions such as autonomous driving decision-making, lane keeping, and automatic parking.

[0023] To address the unreliability of GPS signals, existing technologies typically incorporate onboard sensors such as inertial measurement units (IMUs), wheel speed sensors, visual cameras, and lidar as supplementary means. Among these, the INS-based method calculates position using integral acceleration and angular velocity data, offering advantages such as high short-term accuracy and independence from external signals. However, its errors accumulate rapidly over time, leading to significant increases in positioning deviation after extended operation. Wheel speed sensors combined with dead reckoning (DR) technology can alleviate this problem to some extent, but are susceptible to tire slippage, road surface gradient, and calibration errors, making it difficult to independently support high-precision positioning requirements. Another approach utilizes environmental perception data, such as visual images or lidar point clouds, to perform feature matching with high-precision maps for positioning, such as visual SLAM or lidar SLAM technologies. These methods perform well in structured road environments but are sensitive to changes in lighting, dynamic obstacles, map gaps, or semantically ambiguous areas, and have high computational complexity, posing challenges in real-time performance.

[0024] Furthermore, existing fusion localization schemes mostly employ traditional filtering algorithms such as Kalman filtering, particle filtering, or extended Kalman filtering for multi-source information fusion. While these methods can improve localization robustness to some extent, they typically assume that system noise follows a Gaussian distribution and have limited adaptability to nonlinear and non-Gaussian scenes. More importantly, they lack an active perception and adjustment mechanism for dynamic changes in the confidence levels of different sensors, making them prone to producing erroneous fusion results when sensor data conflicts or partial failures occur. For example, if the vision system fails to extract features due to strong light or rain / fog, assigning them high weights will severely mislead the overall localization output. Simultaneously, traditional methods often handle all conditions with fixed strategies, failing to adaptively adjust search or optimization strategies based on the dynamic characteristics of the environment. This results in lag in response, slow convergence, or even getting trapped in local optima in complex and abrupt scenarios.

[0025] In summary, current vehicle positioning technologies generally suffer from the following shortcomings when facing complex environments with limited GPS signals: First, they rely excessively on single sensors or fixed fusion models, lacking intelligent collaboration and dynamic weight allocation for multi-source heterogeneous information; second, error accumulation is a significant problem, especially in situations with prolonged absence of satellite signals, making it difficult to maintain positioning continuity; third, they do not adequately utilize road semantics and physical motion constraints, failing to effectively combine vehicle dynamics characteristics with prior environmental knowledge for constraint optimization; and fourth, they lack efficient global search and local fine-grained positioning collaboration mechanisms, making it difficult to balance accuracy with real-time performance and robustness. Therefore, there is an urgent need for a novel vehicle positioning method that can maintain strong adaptability, high accuracy, and continuous stable output even when signals are interrupted or degraded.

[0026] To address the aforementioned issues, this application proposes a vehicle localization method based on gray wolf roles and sensor mapping. Its core lies in the deep integration of the swarm intelligence mechanism of the gray wolf optimization algorithm with multi-source sensors. Task division and collaborative optimization are achieved by simulating the social hierarchy of gray wolves. When GPS signals are abnormal, a physically constrained feasible region is first constructed based on the vehicle's kinematics model, and a gray wolf population is initialized within this region. The α wolf is responsible for fusing inertial measurement unit and wheel speed data to generate the first candidate position through dead reckoning. The β wolf uses environmental perception data and high-precision map matching to generate the second candidate position. The δ wolf dynamically evaluates the consistency between the two and adjusts the confidence weights, thereby guiding the ω wolf to perform local dense search in high-confidence areas. Finally, the iteration is terminated early based on the matching degree between the candidate position and road semantic features, and the optimal localization result is output. The entire process realizes sensor role-based localization, hierarchical search, adaptive weighting, and semantic localization, significantly improving the accuracy, robustness, and real-time performance of vehicle localization in complex environments.

[0027] First, the vehicle localization method based on the mapping between the gray wolf role and sensors provided in the embodiments of this application will be described in detail below with reference to the accompanying drawings.

[0028] Reference Figure 1 The implementation process of the vehicle localization method based on the mapping between the gray wolf role and sensors provided in this application embodiment includes, but is not limited to, the following steps.

[0029] Step S110: Obtain the vehicle's inertial measurement unit data, wheel speed sensor data, and environmental perception data at the current moment.

[0030] In step S110, key information about the vehicle's current operating status and surrounding environment is collected in real time, including acceleration and angular velocity data from the inertial measurement unit, wheel speed information provided by wheel speed sensors, and environmental perception data acquired by devices such as cameras or lidar. This multi-source heterogeneous data forms the basic input for subsequent positioning calculations, providing a reliable alternative information source when GPS signals are abnormal, and ensuring that the system has the ability to continuously perceive vehicle motion and environmental characteristics.

[0031] Step S120: When GPS signal loss or signal quality is detected to be lower than a preset threshold, the feasible motion trajectory of the vehicle within a preset time window is predicted based on the vehicle kinematics model, and a physically constrained location feasible region is generated.

[0032] In step S120, when a GPS signal loss is detected or its quality falls below a preset threshold, a trajectory prediction mechanism based on the vehicle's kinematics model is activated. This step utilizes vehicle dynamic constraints, such as maximum steering angle, longitudinal acceleration limits, and motion boundaries determined by wheel speed integrals, to extrapolate possible paths the vehicle might take within a preset time window, thereby generating a location feasible region constrained by physical laws. This feasible region effectively limits the spatial range of subsequent searches, avoiding computational redundancy and positioning divergence risks associated with unconstrained optimization.

[0033] Step S130: Initialize the gray wolf population within the feasible region and divide the population into α wolves, β wolves, δ wolves and ω wolves according to the social hierarchy of the gray wolves.

[0034] In step S130, the population of gray wolf optimization algorithm individuals is initialized within the aforementioned generated feasible region. Based on the social hierarchy of gray wolves in nature, the entire population is explicitly divided into four roles: α wolf, β wolf, δ wolf, and ω wolf. This division is not random but lays the foundation for different roles to undertake specific tasks, giving the entire optimization process hierarchical and collaborative intelligent search characteristics, thereby more efficiently approximating the real vehicle location.

[0035] Step S140: Configure the α wolf to perform dead reckoning by fusing inertial measurement unit data and wheel speed sensor data to generate the first candidate position.

[0036] In step S140, the α wolf is assigned a specific positioning responsibility: fusing inertial measurement unit data and wheel speed sensor data to generate a preliminary candidate position using dead reckoning. This position represents a possible position estimate derived from the vehicle's own motion state, demonstrating the ability to autonomously extrapolate the vehicle's trajectory without external absolute reference, and providing a crucial unary input for subsequent multi-source information fusion.

[0037] Step S150: Configure the β wolf to match environmental perception data with a high-precision map to generate a second candidate location.

[0038] In step S150, β Wolf is configured to perform a matching task based on environmental perception data and a high-precision map. By comparing the currently perceived features such as road structure, lane lines, curbs, or landmarks with the corresponding semantic information in the high-precision map, another independent candidate location is generated. This location reflects the vehicle's relative positioning in the environmental space and has strong geometric and semantic consistency, serving as an effective supplement or verification to the dead reckoning results.

[0039] Step S160: Configure δwolf to evaluate the consistency between the first candidate position and the second candidate position, and dynamically adjust their respective confidence weights based on the consistency results.

[0040] In step S160, a delta wolf is introduced to evaluate the consistency between the first and second candidate positions generated by the α and β wolves, respectively. By analyzing the degree of agreement between the two in terms of spatial location, motion trend, or semantic context, the delta wolf dynamically adjusts its corresponding confidence weight, so that the more credible information source gains greater influence in the subsequent optimization process, thereby realizing the adaptive adjustment of sensor confidence and real-time optimization of the fusion strategy.

[0041] Step S170: Update the position of each wolf in the gray wolf population based on confidence weights, and have ω wolf perform a local dense search in the weighted high confidence region.

[0042] In step S170, the positions of each role in the gray wolf population are updated based on the confidence weights output by the δ wolf, and the ω wolf is guided to focus on areas with high confidence after weighting to conduct local intensive searches. This step utilizes the iterative mechanism of swarm intelligence to improve positioning accuracy while narrowing the search range, enabling the optimization process to take into account both global exploration and local fine convergence, effectively balancing efficiency and accuracy.

[0043] In step S180, during the search process, if the matching degree between the candidate positions of multiple ω wolves and the road semantic features exceeds the preset matching threshold, the iteration is terminated in advance, and the candidate position with the highest matching degree is output as the final positioning result of the vehicle.

[0044] In step S180, during the local search performed by the ω-wolf, the matching degree between its candidate positions and road semantic features is continuously monitored. Once the matching degree between the candidate positions of multiple ω-wolves and semantic elements such as lane line continuity, curb shape, or traffic signs in the high-precision map exceeds a preset threshold, it is determined that a reliable localization solution has been found. The system will terminate the iteration process in advance and select the candidate position with the highest matching degree as the final output result, thereby improving the real-time response capability of the algorithm while ensuring accuracy.

[0045] In some embodiments of this application, the vehicle kinematic model introduces kinematic constraints that conform to the actual physical characteristics of the vehicle, including maximum steering angle constraints, longitudinal acceleration limits, and trajectory boundaries determined by wheel speed integrals. The feasible region of location is a three-dimensional spatial region that simultaneously satisfies the constraints, thereby providing a reliable initial spatial range for subsequent positioning search.

[0046] Specifically, the maximum steering angle constraint reflects the minimum turning radius that the vehicle can achieve under mechanical structural limitations, preventing the predicted trajectory from making sharp turns that are impossible to complete in reality. The longitudinal acceleration constraint reflects the vehicle's motion capability during acceleration or braking, constrained by the power system and tire grip, preventing the generation of overly aggressive or unrealistic speed changes. The trajectory boundary determined by the wheel speed integral estimates the vehicle's travel distance based on wheel rotation information, providing a basic mileage reference for position prediction. Applying these three types of constraints together to the trajectory prediction process ensures that the generated feasible location domain is strictly limited to the three-dimensional space that the vehicle can physically reach under the current operating conditions, effectively eliminating a large number of unreasonable or infeasible candidate locations, and significantly improving the search efficiency of the subsequent Gray Wolf optimization algorithm and the physical rationality of the positioning results.

[0047] In some embodiments of this application, environmental perception data includes visual images or LiDAR point cloud data, and high-precision maps contain geometric and semantic information of lane lines, curbs, traffic signs, and parking areas. By clarifying the source types of environmental perception data and the information content carried by high-precision maps, a clear data foundation and semantic support are provided for location matching based on environmental features.

[0048] Visual images can capture rich texture, color, and structural information, making them suitable for recognizing road elements with distinct visual features, such as lane lines and traffic signs. LiDAR point cloud data provides high-precision three-dimensional spatial geometry, offering superior characterization of curbs, obstacle boundaries, and road contours. Both can serve as input for environmental perception, used for comparison with high-precision maps. High-precision maps not only contain the precise geometric shapes and spatial locations of elements such as lane lines, curbs, traffic signs, and parking areas, but also integrate their corresponding semantic attributes, such as lane type, sign meaning, or area function. This map structure, which emphasizes both geometry and semantics, allows the system to not only determine positional alignment during the matching process but also verify the consistency of scene semantics, thereby significantly improving the accuracy and environmental adaptability of the positioning results, especially providing reliable external references for vehicles when GPS signals are lacking.

[0049] In some embodiments of this application, road semantic features are used to verify the reliability of candidate locations, including lane line continuity, curb geometry, traffic sign positions, or parking area boundaries; the matching degree is obtained by calculating the spatial overlap or structural similarity between the multi-sensor features at the candidate location and the corresponding semantic elements in the high-precision map, providing a criterion with physical meaning and environmental consistency for the positioning results.

[0050] Lane continuity reflects whether a vehicle is on a reasonable lane trajectory, preventing the positioning point from deviating from the normal driving path; curb geometry provides spatial constraints on road boundaries, helping to eliminate invalid solutions located in sidewalks, green belts, or obstacle areas; traffic sign locations, as highly recognizable static landmarks, can serve as key anchor points for precise positioning; parking area boundaries provide structured spatial references for vehicles in specific scenarios. These semantic features collectively constitute the structural skeleton of the vehicle's environment. The matching degree is calculated by comparing the spatial overlap or structural similarity between the actual observed features extracted from multiple sensors at candidate locations and the corresponding semantic elements in the high-precision map. For example, it assesses whether the detected lane line direction aligns with the ground... Figure 1 The system assesses whether the road edge contour reconstructed from the laser point cloud matches the map model. This semantic consistency-based quantitative evaluation mechanism enables the system to identify the location that best matches the real road environment from multiple candidate solutions, effectively improving the accuracy, rationality, and robustness of the localization results, and providing a reliable basis for early termination of the search iteration.

[0051] In some embodiments of this application, during the frame-by-frame localization of the vehicle, the optimal gray wolf position obtained from the previous frame is retained, and a trajectory smoothness energy function is constructed based on the candidate positions of the current frame and the previous frame. This trajectory smoothness energy function is used as a regularization term of the fitness function to penalize candidate trajectories with abrupt acceleration changes or discontinuous path curvature. By introducing trajectory smoothness constraints, the consistency and physical plausibility of continuous localization results in the time dimension are improved.

[0052] During continuous localization, the system not only focuses on the optimal position estimate at the current moment but also retains the historical best gray wolf position sequence to construct the vehicle's historical trajectory. Based on this, a trajectory smoothness energy function is designed and used as a regularization term in the fitness function of the gray wolf optimization algorithm for additional evaluation of candidate trajectories. The core objective of this energy function is to penalize kinematically unreasonable trajectory features, such as abrupt changes in acceleration or sudden changes in path curvature. These phenomena are typically caused by sensor noise, matching errors, or local optimization trapping in outliers. By explicitly suppressing such non-smooth behaviors during optimization, the system can guide the search direction towards a continuous and natural trajectory that conforms to the actual dynamic characteristics of the vehicle. This effectively reduces localization jitter, jumps, or drift, enhancing the stability and reliability of the localization output. Especially in situations with prolonged absence of GPS signals or fluctuations in environmental perception quality, it significantly improves the overall consistency of the trajectory and driving safety.

[0053] In some embodiments of this application, when the confidence level of the environmental perception data is lower than a preset perception threshold, the weight of the α wolf is dynamically increased and the weight of the β wolf is decreased, so as to realize the dynamic adaptive adjustment of the sensor fusion strategy during the positioning process, in order to cope with the uncertainty caused by the fluctuation of the quality of the environmental perception data.

[0054] In actual driving, environmental perception data such as visual images or LiDAR may become unreliable due to factors such as drastic changes in lighting, rain and fog, interference from dynamic obstacles, or missing maps, leading to a decrease in their confidence level. In such cases, assigning a high weight to the beta wolf (a sensor based on environmental perception) could introduce incorrect matching results and mislead the overall positioning direction. Therefore, the system sets a preset perception threshold to evaluate the reliability of environmental perception data in real time. Once the confidence level is detected to be below this threshold, the weight of the alpha wolf is automatically increased, strengthening its influence on the dead reckoning generated by the inertial measurement unit and wheel speed sensors, while the weight of the beta wolf is correspondingly decreased, weakening the role of environmental matching results in the fusion decision. This online reallocation mechanism allows the system to flexibly adjust the dependence on information sources according to the current sensor status. Under poor perception conditions, it relies more on the vehicle's own motion model, while under good perception conditions, it fully utilizes environmental semantic information, thus maintaining robustness and accuracy in positioning under various complex scenarios.

[0055] Secondly, refer to Figure 2 This application provides a vehicle positioning system based on the mapping between the gray wolf role and sensors. The system includes: The data acquisition module is used to acquire data from the vehicle's inertial measurement unit, wheel speed sensors, and environmental perception. This data forms the foundational input for the entire positioning system, providing essential observational data for subsequent autonomous positioning and environmental matching when GPS signals are abnormal.

[0056] The signal detection module is used to determine whether the GPS signal is lost or its quality is below a preset threshold. As a key component of the system's triggering mechanism, this module can promptly initiate an auxiliary positioning process based on the Grey Wolf optimization algorithm when positioning conditions deteriorate. This ensures that the system can seamlessly switch to a backup positioning mode even when external signals are unreliable, guaranteeing the continuity of positioning services.

[0057] The feasible region generation module is used to predict the feasible motion trajectory of the vehicle within a preset time window based on the vehicle's kinematics model when the GPS signal is abnormal, generating a physically constrained feasible region. This feasible region comprehensively considers the vehicle's maximum steering angle, longitudinal acceleration limit, and motion boundary determined by wheel speed integral, effectively limiting the spatial range of subsequent searches, avoiding blind exploration of invalid areas, and improving positioning efficiency and rationality.

[0058] The gray wolf initialization module is used to initialize the gray wolf population within the feasible location domain and divide the population into α wolves, β wolves, δ wolves, and ω wolves according to their social hierarchy. This structured initialization not only assigns specific behavioral responsibilities to each individual but also lays the organizational foundation for subsequent multi-role collaborative search and hierarchical optimization, giving the entire localization process characteristics of swarm intelligence.

[0059] The role mapping module is used to: configure the α wolf to perform dead reckoning by fusing inertial measurement unit data and wheel speed sensor data to generate the first candidate position; configure the β wolf to match environmental perception data with high-precision maps to generate the second candidate position; and configure the δ wolf to evaluate the consistency between the first and second candidate positions and dynamically adjust their respective confidence weights based on the consistency results, thereby realizing intelligent fusion of multi-source information and credibility guidance.

[0060] The collaborative search module updates the positions of each wolf in the gray wolf population based on confidence weights and drives the ω wolf to perform a local intensive search in the weighted high-confidence region. This module achieves a balance between global exploration and local refinement through a group cooperation mechanism, enabling the search process to quickly approximate the true location while avoiding getting trapped in local optima, thereby improving positioning accuracy and convergence speed.

[0061] The semantic verification and output module is used to terminate the iteration in advance when the matching degree between multiple candidate positions of ω wolves and road semantic features exceeds a preset matching threshold during the search process, and output the candidate position with the highest matching degree as the final positioning result of the vehicle, ensuring that the output is both accurate and consistent with the actual road environment.

[0062] In some embodiments of this application, the role mapping module is further configured to dynamically increase the weight of the alpha wolf and decrease the weight of the beta wolf when the confidence level of the environmental perception data is lower than a preset perception threshold. This module automatically increases the weight of the alpha wolf and decreases the weight of the beta wolf, thereby reducing reliance on environmental matching results and enhancing trust in dead reckoning information when perception conditions are poor. This mechanism enables online adaptive optimization of the sensor fusion strategy, significantly improving the robustness and reliability of the system in complex and changing environments.

[0063] Furthermore, embodiments of this application provide an electronic device, which includes a memory and a processor. The memory stores a computer program, and the processor executes the computer program to implement the aforementioned vehicle positioning method based on the mapping between the gray wolf role and sensors.

[0064] Furthermore, embodiments of this application provide a computer-readable storage medium storing a computer program that, when executed by a processor, implements the aforementioned vehicle localization method based on the mapping between the gray wolf character and sensors.

[0065] In summary, the vehicle localization method and related equipment based on the mapping between the gray wolf role and sensors provided in this application have the following technical effects.

[0066] This application's embodiments effectively integrate inertial measurement unit (IMU), wheel speed sensor, and environmental perception data when GPS signals are lost or degraded. By introducing a hierarchical swarm intelligence mechanism using the Grey Wolf optimization algorithm, it achieves role-based division of labor and collaborative optimization of multi-source information. The system first constructs a physically constrained feasible region based on the vehicle's kinematic model, significantly reducing the search space and improving positioning rationality. Then, α-wolf, β-wolf, δ-wolf, and ω-wolf algorithms are mapped to dead reckoning, map matching, consistency assessment, and local search tasks, respectively, forming a hierarchical and responsive positioning architecture. By dynamically adjusting sensor confidence weights, the system can adaptively respond to fluctuations in perception quality. When environmental features are unreliable, it prioritizes the vehicle's own motion model; when features are clear, it fully utilizes semantic information, thus balancing accuracy and robustness. Simultaneously, a trajectory smoothness energy function is introduced as a regularization term to effectively suppress abrupt acceleration changes and path curvature discontinuities, ensuring the temporal consistency of continuous positioning results. Ultimately, by leveraging the matching and verification mechanism of road semantic features, the system can converge early in high-confidence regions and output the optimal positioning results that conform to the actual road structure, significantly improving the accuracy, stability, real-time performance, and environmental adaptability of vehicle positioning in weak GNSS scenarios such as complex urban environments, tunnels, and underground parking lots.

[0067] It should be noted that in all specific embodiments of this application, all data processing activities related to user identity or personal characteristics, such as user information, user behavior data, historical data, and location information, will be conducted in accordance with the principles of legality, legitimacy, and necessity. All data collection, use, storage, and processing will be subject to compliance with applicable national and regional laws, regulations, and industry standards, and informed consent from users will be obtained in a clear and explicit manner before processing. For the processing of sensitive personal information, separate consent from users will be obtained through prominent means such as pop-up prompts and independent confirmation pages. If any processing conflicts with laws and regulations, the laws and regulations will prevail, and necessary data processing will only be carried out within the scope permitted by laws and regulations, ensuring that all data-based applications, analyses, and technical implementations are conducted within the scope permitted by laws and regulations.

[0068] In some alternative embodiments, the functions / operations mentioned in the block diagrams may not occur in the order shown in the operation diagrams. For example, depending on the functions / operations involved, two consecutively shown blocks may actually be executed substantially simultaneously, or the blocks may sometimes be executed in reverse order. Furthermore, the embodiments presented and described in the flowcharts of this application are provided by way of example to provide a more comprehensive understanding of the technology. The disclosed methods are not limited to the operations and logic flows presented herein. Alternative embodiments are contemplated in which the order of various operations is changed and sub-operations described as part of a larger operation are executed independently.

[0069] Furthermore, although this application is described in the context of functional modules, it should be understood that, unless otherwise stated, one or more of the functions and / or features may be integrated into a single physical device and / or software module, or one or more functions and / or features may be implemented in a separate physical device or software module. It is also understood that a detailed discussion of the actual implementation of each module is unnecessary for understanding this application. Rather, given the properties, functions, and internal relationships of the various functional modules in the apparatus disclosed herein, the actual implementation of the module will be understood within the scope of ordinary skill of an engineer. Therefore, those skilled in the art can implement the application set forth in the claims using ordinary skill. It is also understood that the specific concepts disclosed are merely illustrative and are not intended to limit the scope of this application, which is determined by the full scope of the appended claims and their equivalents.

[0070] If a function is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this invention, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several programs to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods of the various embodiments of this invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0071] The logic and / or steps represented in the flowchart or otherwise described herein, for example, can be considered as a sequential list of executable programs for implementing logical functions, and can be embodied in any computer-readable medium for use by, or in conjunction with, a program execution system, apparatus, or device (such as a computer-based system, a processor-included system, or other system that can retrieve and execute a program from or in conjunction with such a program execution system, apparatus, or device). For the purposes of this specification, "computer-readable medium" can mean any means that can contain, store, communicate, propagate, or transmit a program for use by or in conjunction with a program execution system, apparatus, or device.

[0072] More specific examples (a non-exhaustive list) of computer-readable media include: electrical connections (electronic devices) having one or more wires, portable computer disk drives (magnetic devices), random access memory (RAM), read-only memory (ROM), erasable and editable read-only memory (EPROM or flash memory), fiber optic devices, and portable optical disc read-only memory (CDROM). Additionally, computer-readable media can even be paper or other suitable media on which programs can be printed, for example, by optically scanning the paper or other media, then editing, interpreting, or, if necessary, processing it in a suitable manner to obtain the program electronically, and then storing it in computer memory.

[0073] It should be understood that various parts of the present invention can be implemented in hardware, software, firmware, or a combination thereof. In the above embodiments, multiple steps or methods can be implemented in software or firmware stored in memory and executed by a suitable program execution system. For example, if implemented in hardware, as in another embodiment, it can be implemented using any one or a combination of the following techniques known in the art: discrete logic circuits having logic gates for implementing logical functions on data signals, application-specific integrated circuits (ASICs) having suitable combinational logic gates, programmable gate arrays (PGAs), field-programmable gate arrays (FPGAs), etc.

[0074] In the foregoing description of this specification, the reference to terms such as "one embodiment / implementation," "another embodiment / implementation," or "certain embodiments / implementations," etc., indicates that a specific feature, structure, material, or characteristic described in connection with an embodiment or example is included in an embodiment or example of the present invention. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples.

[0075] Although embodiments of the invention have been shown and described, those skilled in the art will understand that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the claims and their equivalents.

[0076] The above is a detailed description of the preferred embodiments of the present invention. However, the present invention is not limited to the embodiments. Those skilled in the art can make various equivalent modifications or substitutions without departing from the spirit of the present invention. All such equivalent modifications or substitutions are included within the scope defined by the claims of the present invention.

Claims

1. A vehicle positioning method based on grey wolf role and sensor mapping, characterized by, The method comprises the following steps: obtaining inertial measurement unit data, wheel speed sensor data and environment perception data of a vehicle at a current time point; when it is detected that a GPS signal is lost or the signal quality is lower than a preset threshold, predicting feasible motion trajectories of the vehicle within a preset time window based on a vehicle kinematic model to generate a physically constrained position feasible region; initializing a grey wolf population in the position feasible region, and dividing the population into alpha wolves, beta wolves, delta wolves and omega wolves according to the social ranks of the grey wolves; configuring the alpha wolves to perform dead reckoning by fusing the inertial measurement unit data and the wheel speed sensor data to generate first candidate positions; configuring the beta wolves to match the environment perception data with a high-definition map to generate second candidate positions; configuring the delta wolves to evaluate the consistency between the first candidate positions and the second candidate positions, and dynamically adjusting the confidence weights of the respective wolves according to the consistency results; updating the positions of the wolves in the grey wolf population based on the confidence weights, and performing local dense search in the high-confidence region after weighting by the omega wolves; in the search process, if the matching degrees of the candidate positions of multiple omega wolves with road semantic features exceed a preset matching threshold, the iteration is terminated in advance, and the candidate position with the highest matching degree is output as the final positioning result of the vehicle.

2. The gray wolf role and sensor mapping based vehicle positioning method according to claim 1, wherein, The vehicle kinematic model comprises a maximum steering angle constraint, a longitudinal acceleration limit and a trajectory boundary determined by wheel speed integration, and the position feasible region is a three-dimensional space region that satisfies the constraint conditions at the same time. 3.The gray wolf role and sensor mapping based vehicle positioning method of claim 1, wherein, The environment perception data comprises visual images or laser radar point cloud data, and the high-definition map contains geometric and semantic information of lane lines, road edges, traffic signs and parking areas. 4.The vehicle positioning method based on gray wolf role and sensor mapping according to claim 1, wherein, The road semantic features comprise lane line continuity, road edge geometry, traffic sign position or parking area boundary, and the matching degree is obtained by calculating the spatial coincidence or structural similarity between the multi-sensor features at the candidate position and the corresponding semantic elements in the high-definition map. 5.The gray wolf role and sensor mapping based vehicle positioning method of claim 1, wherein, In the process of frame-by-frame positioning of the vehicle, the optimal grey wolf position obtained in the previous frame search is retained, and a trajectory smoothing performance energy function is constructed based on the candidate positions of the current frame and the previous frame, the trajectory smoothing performance energy function is used as a regularization term of the fitness function to punish candidate trajectories with sudden acceleration or discontinuous path curvature. 6.The gray wolf role and sensor mapping based vehicle positioning method of claim 1, wherein, When the confidence of the environment perception data is lower than a preset perception threshold, the weight of the alpha wolves is dynamically increased and the weight of the beta wolves is decreased.

7. A vehicle positioning system based on grey wolf role and sensor mapping, characterized by, The system comprises: a data acquisition module for obtaining inertial measurement unit data, wheel speed sensor data and environment perception data of a vehicle; a signal detection module for determining whether a GPS signal is lost or the signal quality is lower than a preset threshold; a feasible region generation module for predicting feasible motion trajectories of the vehicle within a preset time window based on a vehicle kinematic model when the GPS signal is abnormal to generate a physically constrained position feasible region; a grey wolf initialization module for initializing a grey wolf population in the position feasible region, and dividing the population into alpha wolves, beta wolves, delta wolves and omega wolves according to the social ranks of the grey wolves; a role mapping module for: The alpha wolf is configured to fuse the inertial measurement unit data and wheel speed sensor data for dead reckoning to generate a first candidate position; The beta wolf is configured to match the environment perception data with a high-definition map to generate a second candidate position; The delta wolf is configured to evaluate consistency between the first candidate position and the second candidate position, and dynamically adjust respective confidence weights according to consistency results; The collaborative search module is configured to update positions of wolves in the gray wolf population based on the confidence weights, and drive the omega wolf to perform local intensive search in a high-confidence region after weighting. The semantic verification and output module is configured to terminate iteration in advance and output a candidate position with the highest matching degree as a final positioning result of the vehicle when the matching degree of candidate positions of multiple omega wolves and road semantic features exceeds a preset matching threshold during the search process.

8. The grey wolf role and sensor mapping based vehicle positioning system of claim 7, wherein, The role mapping module is further configured to dynamically increase a weight of the alpha wolf and decrease a weight of the beta wolf when a confidence degree of the environment perception data is lower than a preset perception threshold.

9. An electronic device, comprising: The electronic device includes a memory and a processor, the memory stores a computer program, and the processor implements the vehicle positioning method based on the gray wolf role and sensor mapping according to any one of claims 1 to 6 when executing the computer program.

10. A computer-readable storage medium storing a computer program, the computer program comprising instructions that, when executed by a computer, cause the computer to perform the method of any one of claims 1 to 9. The computer program is executed by the processor to implement the vehicle positioning method based on the gray wolf role and sensor mapping according to any one of claims 1 to 6.

Images