CN-121982926-A - Collaborative parking system and method based on distributed wireless sensor network

CN121982926ACN 121982926 ACN121982926 ACN 121982926ACN-121982926-A

Abstract

The invention provides a cooperative parking system based on a distributed wireless sensor network, which comprises a distributed sensing module, a visual positioning module, a berth distribution module, a path planning module and a joint simulation module. All modules are connected in sequence to form a complete technical closed loop of perception-positioning-decision-planning-verification. The invention creatively integrates the visual SLAM based on natural characteristics and the correction technology based on the manual label APRILTAGS, solves the problem of accurate positioning of vehicles in the GPS-free environment of the underground parking lot on the premise of hardly increasing the cost of infrastructure, achieves the practical requirement of positioning precision, avoids high deployment cost of a laser or UWB scheme, and provides a berth allocation optimization model based on dynamic adjustment of saturation and an efficient integer code particle swarm solving algorithm, so that global coordination of berth requirements of multiple vehicles can be completed in second-level time.

Inventors

CHU CHUANCHUAN
ZHU ZHONGPAN
XIANG JINGJING

Assignees

上海凡纳比里智能科技有限公司

Dates

Publication Date: 20260505
Application Date: 20260206

Claims (7)

1. A cooperative parking system based on a distributed wireless sensor network is characterized by comprising a distributed sensing module, a visual positioning module, a berth allocation module, a path planning module and a joint simulation module, The distributed sensing module comprises a camera, a physiological signal acquisition instrument, an eye movement instrument, an audio acquisition device and a vehicle operation sensor, and is used for acquiring visual data of the environment of a parking lot, physiological data of a driver, eye movement data of the driver, audio data in a vehicle and vehicle operation data, The visual positioning module is used for extracting and matching the characteristics of the images acquired by the cameras, combining an instant positioning and map construction algorithm and APRILTAGS visual reference labels to realize the accurate positioning of the vehicle in the parking lot, The berth allocation module is used for constructing an optimization model with the minimum total berthing time consumption of the system as a target according to the saturation of the parking lot and the preference information of the user, adopting an integer-coded particle swarm optimization algorithm to solve, outputting a globally optimal berth allocation scheme, The path planning module is used for planning an optimal running path from the current position to the allocated berth for each vehicle by adopting an improved A search algorithm combining a weighted heuristic function and a multi-father node strategy based on a grid map containing multi-layer information, The joint simulation module is used for integrating and verifying functions and performances of the visual positioning module, the berth allocation module and the path planning module in a joint simulation environment of a robot operating system and MATLAB.
2. The co-located parking system based on a distributed wireless sensor network of claim 1, wherein: The system comprises a camera, a physiological signal acquisition instrument, an eye movement instrument, an audio acquisition device, a vehicle operation sensor and a steering wheel/pedal force sensing device, wherein the camera is an RGB camera with 1080p resolution and 30fps frame rate, the physiological signal acquisition instrument is a wearable electrocardiograph with 250Hz sampling rate, the eye movement instrument is a remote measuring eye movement instrument with 120Hz sampling rate and a gaze point tracking error of less than or equal to 1.2 degrees, the sampling rate of the audio acquisition device is 16000Hz, and the vehicle operation sensor comprises a steering wheel/pedal force sensing device with 500Hz sampling rate.
3. The co-located parking system based on a distributed wireless sensor network of claim 1, wherein: The feature extraction executed by the visual positioning module comprises the steps of constructing a multi-layer Gaussian pyramid for an input image, dividing grids at a pixel layer to extract ORB feature points, screening the feature points by adopting a quadtree algorithm to realize uniform distribution, and performing rough matching by adopting a rapid nearest neighbor search algorithm and fine matching by adopting a random sampling consistency algorithm to remove mismatching points.
4. The co-located parking system based on a distributed wireless sensor network of claim 1, wherein: The optimization model is a 0-1 integer programming model, the objective function of the optimization model is to minimize the total weighted parking time consumption of all vehicles, the constraint condition ensures that each vehicle is only allocated with one berth and each berth is allocated to one vehicle at most, and the position of each particle in the integer-coded particle swarm optimization algorithm represents a complete berth allocation sequence.
5. The co-located parking system based on a distributed wireless sensor network of claim 1, wherein: in the path planning module, the cost function of the improved search algorithm is f (n) =g (n) +w.h (n), wherein g (n) is the actual cost from a starting point to a node n, h (n) is a Manhattan distance heuristic value from the node n to an end point, and w is an adjustable weight coefficient, and the multi-father node strategy means that when the node is expanded, if a plurality of father nodes exist to make g (n) equal, the father node which can make the path trend more straight is preferentially selected.
6. The co-located parking system based on a distributed wireless sensor network of claim 1, wherein: The combined simulation module is built based on a Ubuntu operation system, the robot operation system is responsible for running sensor driving, immediately positioning and mapping nodes and issuing vehicle pose topics, and the MATLAB environment is responsible for subscribing the pose topics and executing the berth allocation and path planning algorithm.
7. The co-located parking system based on a distributed wireless sensor network of claim 1, wherein: In the visual positioning module, the detection and identification process of the APRILTAGS visual reference label comprises image graying and binarization, connected domain clustering and quadrilateral fitting, encoding and decoding based on hamming distance and sub-pixel level angular point positioning.

Description

Collaborative parking system and method based on distributed wireless sensor network Technical Field The invention relates to the technical field of intelligent network-connected automobiles and automatic parking, in particular to a cooperative parking system and method based on a distributed wireless sensor network, which are suitable for closed or semi-closed scenes with limited satellite navigation signals and dense traffic flow such as large underground parking lots, market garages and the like, and aim to realize high-precision positioning, intelligent berth allocation and efficient path planning of vehicles through multi-mode sensing, cooperative calculation and global optimization. Background With the acceleration of the urban process and the continuous increase of the quantity of vehicles kept, the "parking difficulty" has become a common problem which plagues urban traffic. According to statistics, the phenomenon that a large commercial complex and an underground parking garage in an office area are difficult to find in one place and are jammed and coexist in the interior in a peak period is frequently caused, and meanwhile, part of parking spaces are idle due to opaque information. The existing solutions have mainly the following limitations: 1. The intelligent parking (AVP) of the single vehicle has high cost and limited visual field, depends on an ultrasonic radar, an all-round camera and even a laser radar carried by the vehicle, not only remarkably increases the cost of the single vehicle, but also has limited perception range, can not acquire the global idle parking space distribution and real-time traffic situation of the parking lot, easily causes blind cruising of multiple vehicles and aggravates the congestion in the field. 2. The underground environment positioning technology has the bottleneck that the global satellite navigation system (GNSS) signal is completely disabled in the underground parking lot. Alternatives such as Ultra Wideband (UWB) or lidar SLAM can provide high accuracy positioning, but have problems of high infrastructure deployment cost, complex maintenance, etc., which are difficult to popularize on a large scale. 3. The parking space allocation strategy is simplified, the existing parking space guiding system or primary car networking application mostly adopts simple strategies such as 'nearest parking space priority' or 'partition guiding', and the like, and the consideration of personalized preferences (such as approaching an elevator port, needing charging piles and the like) of users is lacking from the aspect of the overall efficiency optimization of the system. 4. The path planning does not consider the on-site driving experience, namely, the paths generated on the parking lot grid map by the traditional path planning algorithm (such as Dijkstra and basic A-x algorithm) often have too many unnecessary right angle turns or bypasses, so that the driving time and energy consumption are increased, and the driving smoothness and comfort are also influenced. In summary, in the prior art, it is difficult to realize high-precision and low-cost positioning of vehicles in an underground parking scene on the premise of controllable cost, and on the basis, collaborative parking decision and planning considering multi-objective optimization are completed. Therefore, a collaborative parking system integrating advanced sensing, positioning, decision-making and planning technologies is needed to improve the overall operation efficiency of the parking lot and the parking experience of the user. Disclosure of Invention The invention is made to solve the above problems, and an object of the invention is to provide a cooperative parking system and method based on a distributed wireless sensor network. The invention provides a cooperative parking system based on a distributed wireless sensor network, which is characterized by comprising a distributed sensing module, a visual positioning module, a berth allocation module, a path planning module and a joint simulation module, The distributed sensing module comprises a camera, a physiological signal acquisition instrument, an eye movement instrument, an audio acquisition device and a vehicle operation sensor, and is used for acquiring visual data of the environment of a parking lot, physiological data of a driver, eye movement data of the driver, audio data in a vehicle and vehicle operation data, The visual positioning module is used for extracting and matching the characteristics of the images acquired by the cameras, combining an instant positioning and map construction algorithm and APRILTAGS visual reference labels to realize the accurate positioning of the vehicle in the parking lot, The berth allocation module is used for constructing an optimization model with the minimum total berthing time consumption of the system as a target according to the saturation of the parking lot and the preference information of the user, adopting