Intelligent Safety Management System and Working Method for Bus Depots Based on BDS / UWB Joint Positioning

Abstract

This invention relates to an intelligent safety management system and operating method for bus depots based on BDS / UWB joint positioning, belonging to the field of intelligent transportation technology. It includes a joint positioning device: achieving positioning in different spaces and environments through BDS / UWB joint positioning; a manual patrol device: security personnel receive patrol plans and routes through the manual patrol device, reporting abnormal situations and location information; an unmanned automatic patrol device: automatically receiving patrol plans and conducting safety checks on vehicles within the depot or other loosely located areas according to the plan and patrol routes; and a network service platform: regulating vehicle operation and controlling the joint positioning device, manual patrol device, and unmanned automatic patrol device. This invention reduces safety hazards, improves vehicle safety management, enhances the intelligence level of bus company management, reduces the workload of patrol personnel, and improves the economic benefits of bus operating companies.

Description

Technical Field

[0001] This invention relates to an intelligent safety management system and operating method for bus depots based on BDS / UWB joint positioning, belonging to the field of intelligent transportation technology. Background Technology

[0002] With the acceleration of urbanization, public transportation systems face multiple challenges, especially in the intelligent management of bus depots. Traditional bus depots rely on manual patrols and fixed parking space management, resulting in a high dependence on manual vehicle dispatching, delayed response, and safety hazards such as missed inspections and blind spots in complex nighttime environments. Furthermore, the large-scale application of new energy buses places higher demands on the thermal safety management of power batteries, especially since abnormal battery pack temperatures at night can easily lead to thermal runaway. Traditional manual infrared inspection methods suffer from low monitoring frequency and high missed inspection rates.

[0003] Meanwhile, the scarcity of land resources has spurred the widespread adoption of multi-level, multi-level bus depots. However, the complex indoor structures cause BeiDou signals to fail, necessitating the development of high-precision joint positioning technology adapted to multi-level parking lots to achieve functions such as automatic parking location planning and guidance for buses. Existing depot management systems have significant shortcomings: patrol routes are static and fixed, making it impossible to dynamically adjust inspection priorities based on real-time data such as charging status and battery temperature; the gate system and dispatch platform are disconnected, resulting in low entry and exit efficiency, and vehicle placement cannot be reasonably arranged according to departure time.

[0004] For some areas lacking parking facilities or where vehicles are parked both inside and outside the facility simultaneously, there are a series of problems, such as missed inspections during patrols, low inspection efficiency due to dispersed vehicles, and high workload for inspection personnel. The level of unmanned automation and intelligence is very low. The inspection process does not use computer vision to achieve non-contact monitoring of battery temperature, and reinforcement learning algorithms are not used to dynamically optimize the drone inspection path and gate control strategy to form a closed loop of perception-decision-execution. As a result, the systemic safety and efficiency of facility operation are not high. Summary of the Invention

[0005] To address the shortcomings of existing technologies, this invention provides an intelligent safety management system and operating method for bus depots based on BDS / UWB joint positioning. By integrating and analyzing multiple elements such as people, vehicles, depots, and cloud, it better applies BDS / UWB joint positioning and the Internet of Things to the management and scheduling of new energy bus depots. This enables the rational planning of patrol routes for new energy bus depots, reduces safety hazards, improves vehicle safety management, enhances the intelligence level of bus company management, reduces the workload of patrol staff, and improves the economic benefits of bus operating companies.

[0006] Terminology Explanation

[0007] 1. UWB (Ultra Wide Band): Ultra Wide Band technology is a wireless carrier communication technology that does not use sinusoidal carriers but instead uses nanosecond-level non-sinusoidal narrow pulses to transmit data. Therefore, it occupies a wide spectrum range. UWB technology has advantages such as low system complexity, low transmitted signal power spectral density, insensitivity to channel fading, low interception capability, and high positioning accuracy. It is especially suitable for high-speed wireless access in dense multipath environments such as indoor spaces.

[0008] 2. BeiDou Navigation Satellite System (BDS): The BeiDou Navigation Satellite System is a global satellite navigation system independently developed by China, and is the third mature satellite navigation system after GPS and GLONASS. The BeiDou Navigation Satellite System consists of three segments: space segment, ground segment, and user segment. It provides high-precision, high-reliability positioning, navigation, and timing services to various users worldwide, 24 / 7, and also has short message communication capabilities. After years of development, the BeiDou system has become an important new infrastructure providing all-weather, all-time, high-precision positioning, navigation, and timing services to global users.

[0009] 3. Multi-technology Positioning: This is a comprehensive solution that overcomes the limitations of single technologies by integrating multiple sensors, signal sources, and positioning algorithms into a composite system. BeiDou / UWB (Ultra-Wideband): Applicable to enclosed scenarios requiring centimeter-level accuracy, with positioning accuracy of 0.1-0.3 meters, depending on UWB base station density (0.1-meter accuracy can be achieved with anchor points deployed every 20 meters); BeiDou / 5G: Applicable to city-level dynamic positioning needs, with positioning accuracy of 0.3-1 meter; BeiDou / Laser Inertial Navigation (INS): Applicable to highly reliable, strong anti-interference environments, with positioning accuracy of 0.01-0.1 meters (depending on the zero-bias stability of laser gyroscopes); BeiDou / Visual SLAM: Applicable to complex terrain areas without base station coverage, with positioning accuracy of 1-3 meters.

[0010] 4. Adaptive Patrolling Planning refers to an intelligent decision-making system that dynamically optimizes patrol paths by sensing environmental changes and mission requirements in real time. Its core features include: environmental responsiveness: nonlinear adaptation to dynamic elements such as lighting, weather, and obstacles; mission priority scheduling: multi-objective optimization based on threat level, energy efficiency, data collection requirements, etc.; and autonomous disturbance recovery: generating a new optimized path within 10 seconds after encountering a sudden event.

[0011] 5. The vehicle dispatching system utilizes GPS technology, wireless data transmission, and computer software (MIS) to manage various static and dynamic information of public transport vehicles. Its main functions include automatic voice announcements of bus stops, route management, route statistics, vehicle dispatching management, speeding alarms, designated route operation, shift scheduling, driver management, and text information display.

[0012] 6. The Intelligent Terminal Safety Management System (ITSMS) is an integrated platform that combines IoT sensing, AI analysis, and digital twin technologies. It aims to achieve comprehensive safety management across transportation hubs, logistics parks, energy stations, and other similar scenarios. This includes risk prevention and control: through multi-source data analysis, it transforms incident response from reactive handling to proactive prevention; and efficiency collaboration: deep integration of safety supervision and operational processes avoids disconnect from human management.

[0013] The technical solution of the present invention is as follows:

[0014] The BDS / UWB-based intelligent safety management system for bus depots includes a joint positioning device, a manual patrol and scanning device, an unmanned automatic patrol device, and a network service platform, among which:

[0015] Joint positioning device: It achieves accurate positioning in different spaces and environments through joint positioning via BDS / UWB, realizes vehicle positioning through vehicle-mounted positioning device, realizes spatial positioning within the station through station positioning device, and builds internal station maps. The positioning device for inspection equipment is used to realize the positioning of inspection equipment.

[0016] Manual patrol device: Security personnel receive patrol plans and routes, report abnormal situations and location information, and conduct relevant safety tests on electric vehicles through the manual patrol device;

[0017] Unmanned automatic patrol device: Automatically accepts patrol plans and conducts safety inspections of vehicles in the station or other loose areas according to the plan and patrol route. The unmanned automatic patrol device has visual perception function, and reports to the network service platform after detecting abnormalities for manual verification and remote control operation.

[0018] Network service platform: regulates vehicle operation and controls joint positioning devices, manual patrol devices, and unmanned automatic patrol devices.

[0019] When vehicles return to the depot, the system automatically matches parking locations based on departure times and charging schedules, rationally allocating charging stations and avoiding obstruction between vehicles to improve depot operational efficiency. The system constructs a vehicle location map for the multi-level parking garage. Based on pre-set patrol rules, it automatically generates patrol plans and routes, records any anomalies during patrols, analyzes these anomalies, and promptly notifies relevant personnel for handling. The system can also adjust patrol plans and routes in a timely manner based on anomalies within the depot and deploy unmanned equipment for multi-faceted environmental inspection.

[0020] According to a preferred embodiment of the present invention, the joint positioning device comprises:

[0021] Receiver unit: includes a BDS module and a UWB module, used to receive BeiDou positioning data and UWB communication distance data;

[0022] The computing unit includes time and space synchronization of BeiDou positioning data and UWB data, UWB data conversion, data processing, filtering and resolution, implementation of filtering iteration and anomaly observation and detection, and realization of BDS / UWB joint positioning and generation of joint positioning data according to different application scenarios.

[0023] Output unit: Different interface methods are designed, and corresponding interfaces can be customized to output data according to the interface requirements of different devices.

[0024] The working method of the above-mentioned intelligent safety management system for bus depots based on BDS / UWB joint positioning is as follows:

[0025] (1) The joint positioning device locates the vehicle's position in real time. After the vehicle arrives at the station, it obtains the vehicle's charging plan and operation plan. Based on the charging amount and charging time in the charging plan, it allocates charging piles and displays and guides the vehicle to the designated location at the gate when the vehicle enters the station. If the vehicle is not charging in the designated location or has completed charging, it reminds the driver to end charging in time and park the vehicle in the designated location through information push.

[0026] If there is no charging plan for the vehicle, the parking location will be planned according to the vehicle's operation plan, taking into account factors such as the next departure time and the utilization efficiency of the parking location. The location will be displayed at the gate, and the driver will park the vehicle in the designated location according to the prompts. If the vehicle is not parked in the designated parking space, the station management personnel and the driver will be notified by pushing the message.

[0027] (2) Security patrol: The station management personnel shall set patrol requirements according to the station management requirements, including the number of patrols, patrol locations, number of patrol vehicles, patrol methods, etc., and formulate patrol plans based on the set patrol requirements and the parking status of station vehicles, and issue the patrol plans to manual patrol devices or unmanned automatic patrol devices.

[0028] Manual or unmanned automatic patrol devices conduct patrols according to the patrol plan, at the prescribed time, along the prescribed route, and for the prescribed items. The patrol results are uploaded in real time. When an anomaly is detected, the patrol plan is changed, the patrol personnel are notified, and unmanned equipment is mobilized to conduct a joint investigation of the anomaly.

[0029] If the patrol anomaly is resolved, patrols will continue according to the patrol plan. If the anomaly requires escalation, personnel will be notified in accordance with the pre-set anomaly management requirements to handle the situation until the patrol anomaly is resolved.

[0030] After the anomaly is resolved, the patrol plan is modified according to the fault location. The areas or vehicle parts where the anomaly occurred are inspected in a focused manner. The patrol is carried out from multiple directions and angles using both manual and unmanned equipment until the entire patrol plan is completed and a patrol report is generated.

[0031] According to a preferred embodiment of the present invention, in step (1), the positioning method of the combined positioning device is as follows:

[0032] (11) Coordinate transformation: BeiDou coordinate transformation (UTM plane coordinates):

[0033] (x UTM ,y UTM = wgs2utm(lat,lon)

[0034] Where lat,lon represents the latitude and longitude coordinates output by BeiDou; x UTM ,y UTM The coordinates are the transformed UTM plane coordinates (unit: meters); wgs2utm is the coordinate transformation function;

[0035] (12) Time alignment:

[0036]

[0037] Where: t is the timestamp for target alignment; z k ,z k+1 They are respectively at t l ,t k+1 The original observation value at time z; synced The interpolation result is aligned to t;

[0038] (13) Dynamic accuracy weight allocation: weight allocation based on inherent sensor indicators;

[0039] BDS weights:

[0040]

[0041] Where, N sat N represents the current number of satellites.max The maximum number of visible satellites; HDOP is the horizontal precision factor; ∈ is a small constant (e.g., 0.1) to prevent division by zero.

[0042] UWB weights:

[0043]

[0044] Where: SNR is the signal-to-noise ratio α, and β are parameters calibrated experimentally;

[0045] Weighted fusion position P fused :

[0046]

[0047] Where: P BDS For BeiDou positioning, P UWB This is the UWB location.

[0048] According to a preferred embodiment of the present invention, in step (1), the step of planning the parking location based on the vehicle's operation plan specifically includes:

[0049] Determine if the vehicle is no longer in operation. If it is, obtain the departure plan for the next day and allocate the final parking space. If it is in normal operation, allocate a temporary parking space.

[0050] According to a further preferred embodiment of the present invention, the dynamic allocation of parking spaces is implemented using a greedy optimization algorithm, with the following steps:

[0051] a. Design input parameters:

[0052] Vehicle set V = {v1, v2, ..., v n}, each car v i Includes attribute: arrival time Parking duration d (i) Vehicle type (Gasoline-powered vehicles, pure electric vehicles, hybrid vehicles), vehicle dimensions s (i) ;

[0053] Parking space set P = {p1, p2, ..., p m}, each parking space p j Includes attribute: Parking space type (Charging station parking spaces, obstructed parking spaces, temporary parking spaces), parking space dimensions Available time (Initially set to 0, indicating that it is always available);

[0054] Type compatibility rules: Define the allowed matching relationship between vehicle type and parking space type. Pure electric vehicles can park in charging parking spaces, temporary parking spaces, and obstructed parking spaces. Fuel vehicles can park in obstructed parking spaces or temporary parking spaces. Hybrid vehicles can only park in charging parking spaces and temporary parking spaces.

[0055] Objective: To assign each vehicle to a suitable parking space, minimizing total waiting time or maximizing parking space utilization;

[0056] b. Use a greedy algorithm to allocate parking spaces:

[0057] First, the buses entering the site are prioritized:

[0058] Vehicles are processed in ascending order of arrival time:

[0059]

[0060] in, Let V be the arrival time, and V be the vehicle group. sorted The set of vehicles that have been prioritized;

[0061] Then, based on the vehicle's attribute requirements, it is matched with the parking space conditions;

[0062] For vehicle v i A legal parking space must meet the following requirements:

[0063]

[0064] Where, p j For the required parking spaces, To meet the requirements for the set of parking space attributes, Parking space type Parking space dimensions Available time, For vehicle type, s (i) For vehicle dimensions, This refers to the arrival time.

[0065] c. Use a greedy algorithm to select the parking space that best meets all the criteria:

[0066] Select the earliest available parking space from the legal parking spaces to minimize subsequent waiting times:

[0067]

[0068] in, To meet the requirements for the set of parking space attributes, p j For the required parking spaces, To find the most suitable parking space Available time;

[0069] d. Assuming all conditions are met, introduce priority weights to select the most suitable parking space:

[0070] Define a comprehensive scoring function that combines vehicle type priority, size matching degree, and availability time to select the earliest available parking space from the legal parking spaces, thereby minimizing subsequent waiting times:

[0071]

[0072] Where α is the vehicle type weight matching, β is the vehicle size matching weight, and γ is the available time matching weight;

[0073] e. After parking spaces are allocated, update the parking space status promptly:

[0074] The availability time of parking spaces is updated after allocation:

[0075]

[0076] in, Available time, For arrival time, d (i) This refers to the parking duration.

[0077] According to a preferred embodiment of the present invention, in step (2), the patrol route planning step is as follows:

[0078] (21) Input parameters:

[0079] Patrol area topology map G=(V,E):

[0080] Where the node set V = {v1, v2, ..., v} n} indicates key patrol points (such as entrances and exits, charging areas, vehicle parking areas, and blind spots in surveillance).

[0081] Let E be the set of edges, representing the paths between nodes, and each edge e ij Includes: distance d ij ;

[0082] Security risk weight w ij ∈[0,1], the larger the value, the higher the risk, and the more priority should be given to patrolling;

[0083] Patrol time cost t ij =d ij / v, where v is the speed of security personnel or unmanned equipment;

[0084] Patrol node attributes include:

[0085] Importance weight ρ i ∈[1,5], a larger value indicates that a higher patrol frequency is required.

[0086] Minimum patrol interval Δti (Unit: minutes);

[0087] Patrol path algorithm parameters include:

[0088] Ant count (number of patrol personnel / number of unmanned devices) m;

[0089] Pheromone importance factor α;

[0090] Heuristic importance factor β;

[0091] The pheromone evaporation rate ρ∈(0,1);

[0092] The pheromone increment intensity Q;

[0093] Maximum number of iterations T max ;

[0094] (22) Ant colony algorithm is used for path selection:

[0095] (221) The probability that patrol device k selects the next node j at node i:

[0096]

[0097] Where, τ ij For edge e ii The concentration of pheromones on the surface;

[0098] eta ij Heuristic functions are defined as follows:

[0099]

[0100] (222) Patrol environment environmental pheromone update:

[0101] Global update (after all patrol devices have completed their routes):

[0102]

[0103] The increment of the k-th patrol device is:

[0104]

[0105] Cost k The total cost of the route (time + risk);

[0106] Time k Total path time;

[0107] ρ p The sum of importance weights of nodes in the path;

[0108] Partial update (real-time update as patrol equipment moves):

[0109] τ ij ←(1-ρ)τ ij +ρ·τ0 (τ0 is the initial pheromone concentration, to avoid excessive path monopolization);

[0110] (223) Handling of patrol dynamic constraints:

[0111] Patrol node coverage constraints:

[0112] Introducing a penalty term, if node v i The patrol interval exceeds Δt i :

[0113]

[0114] Add penalty items to path cost calculation;

[0115] Multi-patrol group coordination constraints:

[0116] The patrol equipment is divided into K groups, each searching its own path. The objective function is then modified as follows:

[0117]

[0118] mu represents the load balancing weight. Through the path planning of the ant colony algorithm mentioned above, the design and formulation of each patrol route within the bus station are realized. Patrol constraints are set according to work needs to control the patrol range and realize multi-group collaborative patrol and human-machine collaborative patrol mechanism.

[0119] The beneficial effects of this invention are as follows:

[0120] 1. Heterogeneous positioning fusion technology: This invention designs an adaptive weight allocation strategy to prioritize the use of BeiDou signals in open areas and switch to UWB positioning in obstructed areas, resulting in accurate positioning.

[0121] 2. Human-machine collaborative patrol mechanism, supporting joint operation mode of drones and security personnel. When the vision system detects an anomaly, it automatically dispatches the nearest drone to conduct a close inspection.

[0122] 3. Parking space-gate linkage control: Parking spaces are allocated according to vehicle scheduling and charging plans. After allocation, the internal route plan is automatically generated and linked with the gate to realize automatic gate opening and closing, allowing vehicles to exit without stopping and avoiding vehicle accumulation at the exit.

[0123] This invention constructs a complete closed loop for site safety management through the deep integration of multiple technologies. Attached Figure Description

[0124] Figure 1 This is a schematic diagram of the structure of the present invention;

[0125] Figure 2 This is a flowchart illustrating the vehicle parking space allocation process of the present invention.

[0126] Figure 3 This is a flowchart illustrating the security patrol workflow of the present invention.

[0127] Figure 4 This is a schematic diagram of the combined positioning device of the present invention. Detailed Implementation

[0128] The present invention will be further described below with reference to the embodiments and accompanying drawings, but is not limited thereto.

[0129] Example 1:

[0130] like Figure 1-4 As shown, this embodiment provides an intelligent safety management system for bus depots based on BDS / UWB joint positioning, including a joint positioning device, a manual patrol and scanning device, an unmanned automatic patrol device, and a network service platform, wherein:

[0131] Joint positioning device: It achieves accurate positioning in different spaces and environments through joint positioning via BDS / UWB, realizes vehicle positioning through vehicle-mounted positioning device, realizes spatial positioning within the station through station positioning device, and builds internal station maps. The positioning device for inspection equipment is used to realize the positioning of inspection equipment.

[0132] Manual patrol device: Security personnel receive patrol plans and routes, report abnormal situations and location information, and conduct relevant safety tests on electric vehicles through the manual patrol device;

[0133] Unmanned automatic patrol device: Automatically accepts patrol plans and conducts safety inspections of vehicles in the station or other loose areas according to the plan and patrol route. The unmanned automatic patrol device has visual perception function, and reports to the network service platform after detecting abnormalities for manual verification and remote control operation.

[0134] Network service platform: regulates vehicle operation and controls joint positioning devices, manual patrol devices, and unmanned automatic patrol devices.

[0135] When vehicles return to the depot, the system automatically matches parking locations based on departure times and charging schedules, rationally allocating charging stations and avoiding obstruction between vehicles to improve depot operational efficiency. The system constructs a vehicle location map for the multi-level parking garage. Based on pre-set patrol rules, it automatically generates patrol plans and routes, records any anomalies during patrols, analyzes these anomalies, and promptly notifies relevant personnel for handling. The system can also adjust patrol plans and routes in a timely manner based on anomalies within the depot and deploy unmanned equipment for multi-faceted environmental inspection.

[0136] According to a preferred embodiment of the present invention, the joint positioning device comprises:

[0137] Receiver unit: includes a BDS module and a UWB module, used to receive BeiDou positioning data and UWB communication distance data;

[0138] The computing unit includes time and space synchronization of BeiDou positioning data and UWB data, UWB data conversion, data processing, filtering and resolution, implementation of filtering iteration and anomaly observation and detection, and realization of BDS / UWB joint positioning and generation of joint positioning data according to different application scenarios.

[0139] Output unit: Different interface methods are designed, and corresponding interfaces can be customized to output data according to the interface requirements of different devices.

[0140] The working method of the above-mentioned intelligent safety management system for bus depots based on BDS / UWB joint positioning is as follows:

[0141] (1) The joint positioning device locates the vehicle's position in real time. After the vehicle arrives at the station, it obtains the vehicle's charging plan and operation plan. Based on the charging amount and charging time in the charging plan, it allocates charging piles and displays and guides the vehicle to the designated location at the gate when the vehicle enters the station. If the vehicle is not charging in the designated location or has completed charging, it reminds the driver to end charging in time and park the vehicle in the designated location through information push.

[0142] If there is no charging plan for the vehicle, the parking location will be planned according to the vehicle's operation plan, taking into account factors such as the next departure time and the utilization efficiency of the parking location. The location will be displayed at the gate, and the driver will park the vehicle in the designated location according to the prompts. If the vehicle is not parked in the designated parking space, the station management personnel and the driver will be notified by pushing the message.

[0143] The positioning method of the joint positioning device is as follows:

[0144] (11) Coordinate transformation: BeiDou coordinate transformation (UTM plane coordinates):

[0145] (x UTM ,y UTM = wgs2utm(lat,lon)

[0146] Where lat,lon represents the latitude and longitude coordinates output by BeiDou; x UTM ,y UTM The coordinates are the transformed UTM plane coordinates (unit: meters); wgs2utm is the coordinate transformation function;

[0147] (12) Time alignment:

[0148]

[0149] Where: t is the timestamp for target alignment; z k ,z k+1 They are respectively at t k ,t k+1 The original observation value at time z; synced The interpolation result is aligned to t;

[0150] (13) Dynamic accuracy weight allocation: weight allocation based on inherent sensor indicators;

[0151] BDS weights:

[0152]

[0153] Where, N sat N represents the current number of satellites. max The maximum number of visible satellites; HDOP is the horizontal precision factor; ∈ is a small constant (e.g., 0.1) to prevent division by zero.

[0154] UWB weights:

[0155]

[0156] Where: SNR is the signal-to-noise ratio α, and β are parameters calibrated experimentally;

[0157] Weighted fusion position P fused :

[0158]

[0159] Where: P BDS For BeiDou positioning, P UWB This is the UWB location.

[0160] According to the vehicle operation plan, the specific steps for planning parking locations are as follows:

[0161] Determine if the vehicle is no longer in operation. If it is, obtain the departure plan for the next day and allocate the final parking space. If it is in normal operation, allocate a temporary parking space.

[0162] The dynamic allocation of parking spaces is implemented using a greedy optimization algorithm, with the following steps:

[0163] a. Design input parameters:

[0164] Vehicle set V = {v1, v2, ..., v n}, each car v i Includes attribute: arrival time Parking duration d (i) Vehicle type (Gasoline-powered vehicles, pure electric vehicles, hybrid vehicles), vehicle dimensions s (i) ;

[0165] Parking space set P = {p1, p2, ..., p m}, each parking space p j Includes attribute: Parking space type (Charging station parking spaces, obstructed parking spaces, temporary parking spaces), parking space dimensions Available time (Initially set to 0, indicating that it is always available);

[0166] Type compatibility rules: Define the allowed matching relationship between vehicle type and parking space type. Pure electric vehicles can park in charging parking spaces, temporary parking spaces, and obstructed parking spaces. Fuel vehicles can park in obstructed parking spaces or temporary parking spaces. Hybrid vehicles can only park in charging parking spaces and temporary parking spaces.

[0167] Objective: To assign each vehicle to a suitable parking space, minimizing total waiting time or maximizing parking space utilization;

[0168] b. Use a greedy algorithm to allocate parking spaces:

[0169] First, the buses entering the site are prioritized:

[0170] Vehicles are processed in ascending order of arrival time:

[0171]

[0172] in, Let V be the arrival time, and V be the vehicle group. sorted The set of vehicles that have been prioritized;

[0173] Then, based on the vehicle's attribute requirements, it is matched with the parking space conditions;

[0174] For vehicle v i A legal parking space must meet the following requirements:

[0175]

[0176] Where, p j For the required parking spaces, To meet the requirements for the set of parking space attributes, Parking space type Parking space dimensions Available time, For vehicle type, s (i) For vehicle dimensions, This refers to the arrival time.

[0177] c. Use a greedy algorithm to select the parking space that best meets all the criteria:

[0178] Select the earliest available parking space from the legal parking spaces to minimize subsequent waiting times:

[0179]

[0180] in, To meet the requirements for the set of parking space attributes, p j For the required parking spaces, To find the most suitable parking space Available time;

[0181] d. Assuming all conditions are met, introduce priority weights to select the most suitable parking space:

[0182] Define a comprehensive scoring function that combines vehicle type priority, size matching degree, and availability time to select the earliest available parking space from the legal parking spaces, thereby minimizing subsequent waiting times:

[0183]

[0184] Where α is the vehicle type weight matching, β is the vehicle size matching weight, and γ is the available time matching weight;

[0185] e. After parking spaces are allocated, update the parking space status promptly:

[0186] The availability time of parking spaces is updated after allocation:

[0187]

[0188] in, Available time, For arrival time, d (i) This refers to the parking duration.

[0189] (2) Security patrol: The station management personnel shall set patrol requirements according to the station management requirements, including the number of patrols, patrol locations, number of patrol vehicles, patrol methods, etc., and formulate patrol plans based on the set patrol requirements and the parking status of station vehicles, and issue the patrol plans to manual patrol devices or unmanned automatic patrol devices.

[0190] Manual or unmanned automatic patrol devices conduct patrols according to the patrol plan, at the prescribed time, along the prescribed route, and for the prescribed items. The patrol results are uploaded in real time. When an anomaly is detected, the patrol plan is changed, the patrol personnel are notified, and unmanned equipment is mobilized to conduct a joint investigation of the anomaly.

[0191] If the patrol anomaly is resolved, patrols will continue according to the patrol plan. If the anomaly requires escalation, personnel will be notified in accordance with the pre-set anomaly management requirements to handle the situation until the patrol anomaly is resolved.

[0192] After the anomaly is resolved, the patrol plan is modified according to the fault location. The areas or vehicle parts where the anomaly occurred are inspected in a focused manner. The patrol is carried out from multiple directions and angles using both manual and unmanned equipment until the entire patrol plan is completed and a patrol report is generated.

[0193] The patrol route planning steps are as follows:

[0194] (21) Input parameters:

[0195] Patrol area topology map G=(V,E):

[0196] Where the node set V = {v1, v2, ..., v} n} indicates key patrol points (such as entrances and exits, charging areas, vehicle parking areas, and blind spots in surveillance).

[0197] Let E be the set of edges, representing the paths between nodes, and each edge e ij Includes: distance d ij ;

[0198] Security risk weight w ij ∈[0,1], the larger the value, the higher the risk, and the more priority should be given to patrolling;

[0199] Patrol time cost t ij =d ij / v, where v is the speed of security personnel or unmanned equipment;

[0200] Patrol node attributes include:

[0201] Importance weight ρ i ∈[1,5], a larger value indicates that a higher patrol frequency is required.

[0202] Minimum patrol interval Δt i (Unit: minutes);

[0203] Patrol path algorithm parameters include:

[0204] Ant count (number of patrol personnel / number of unmanned devices) m;

[0205] Pheromone importance factor α;

[0206] Heuristic importance factor β;

[0207] The pheromone evaporation rate ρ∈(0,1);

[0208] The pheromone increment intensity Q;

[0209] Maximum number of iterations T max ;

[0210] (22) Ant colony algorithm is used for path selection:

[0211] (221) The probability that patrol device k selects the next node j at node i:

[0212]

[0213] Where, τ ij For edge e ij The concentration of pheromones on the surface;

[0214] eta ij Heuristic functions are defined as follows:

[0215]

[0216] (222) Patrol environment environmental pheromone update:

[0217] Global update (after all patrol devices have completed their routes):

[0218]

[0219] The increment of the k-th patrol device is:

[0220]

[0221] Cost k The total cost of the route (time + risk);

[0222] Time k Total path time;

[0223] ρ p The sum of importance weights of nodes in the path;

[0224] Partial update (real-time update as patrol equipment moves):

[0225] τ ij ←(1-ρ)τ ij +ρ·τ0 (τ0 is the initial pheromone concentration, to avoid excessive path monopolization);

[0226] (223) Handling of patrol dynamic constraints:

[0227] Patrol node coverage constraints:

[0228] Introducing a penalty term, if node v i The patrol interval exceeds Δt i :

[0229]

[0230] Add penalty items to path cost calculation;

[0231] Multi-patrol group coordination constraints:

[0232] The patrol equipment is divided into K groups, each searching its own path. The objective function is then modified as follows:

[0233]

[0234] mu represents the load balancing weight. Through the path planning of the ant colony algorithm mentioned above, the design and formulation of each patrol route within the bus station are realized. Patrol constraints are set according to work needs to control the patrol range and realize multi-group collaborative patrol and human-machine collaborative patrol mechanism.

[0235] The above description is merely a preferred embodiment of the present invention and is not intended to limit the invention. Various modifications and variations can be made to the present invention by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A working method of a bus station intelligent safety management system based on BDS / UWB combined positioning, characterized in that, The steps are as follows: (1) The combined positioning device locates the vehicle position in real time. After the vehicle arrives at the station, the charging plan and operation plan of the vehicle are obtained. According to the charging capacity and charging time in the charging plan, the charging pile is allocated, and the vehicle is displayed and guided to enter the specified position at the gate when the vehicle enters. If the vehicle does not charge at the specified position or has completed charging, the driver is reminded in time to end the charging through information pushing, and the vehicle is parked at the specified location; If the vehicle has no charging plan, it is judged whether the vehicle stops operation. If it stops operation, the next day's departure plan is obtained, and the final parking space is allocated. If it operates normally, a temporary parking space is allocated, and the position is displayed at the gate. The driver parks the vehicle at the specified position according to the prompt. If the vehicle is not parked in the specified parking space, the station manager and the driver are notified through the push information; The dynamic allocation of parking spaces is realized by using the greedy optimization algorithm, and the steps are as follows: a. Design input parameters: Vehicle set V = {v1, v2, ..., v n }, each car v i Includes attribute: arrival time Parking duration d (i) Vehicle type Vehicle size s (i) ; A set of parking spaces P = {p1, p2,..., p m} each parking space p j contains attributes: parking space type parking space size available time Type compatibility rules: define the allowed matching relationship between vehicle type and parking space type; Objective: allocate each vehicle to a parking space that meets the requirements, minimize the total waiting time or maximize the utilization rate of parking spaces; b. Use the greedy algorithm to allocate parking spaces: First, prioritize the incoming public transport vehicles: Process the vehicles in ascending order of arrival time: wherein, is the time of arrival, V is the set of vehicles, V sorted is the set of vehicles that are prioritized; Then, match the vehicle attributes with the parking space conditions; For a vehicle v i , a legal parking space needs to satisfy: wherein p j is a desired parking space, is a set of required parking space attributes, is a parking space type, is a parking space size, is an available time, is a vehicle type, s (i) is a vehicle size, is an arrival time; c. Select the most suitable parking space for each factor using the greedy rule: Select the earliest available parking space from the legal parking spaces to minimize subsequent waiting time: wherein, p is a set of required parking attributes, j p is a required parking spot, p is a most suitable parking spot, t is an available time; d. Under the premise that the conditions are met, introduce priority weights to select the most suitable parking space: Define a comprehensive scoring function, combine vehicle type priority, size matching degree, and available time, and select the earliest available parking space from the legal parking spaces to minimize subsequent waiting time: Where, α is the vehicle type weight matching, β is the vehicle size matching weight, and γ is the available time matching weight; e. After allocating the parking space, update the parking space status in time: Update the available time of the parking space after allocation: wherein, is the available time, is the arrival time, d (i) is the parking duration; (2) Safety patrol, the station manager sets the patrol requirements according to the station management requirements, including the number of patrols, patrol positions, number of vehicles to be patrolled, and patrol methods. According to the set patrol requirements, the station vehicle parking situation is formulated, and the patrol plan is issued to the artificial patrol device or unmanned automatic patrol device; The artificial patrol device or unmanned automatic patrol device carries out patrol work according to the patrol plan, the specified time, the specified route, and the specified project, and uploads the patrol results in real time. When an abnormal situation is detected, the patrol plan is changed, the patrol personnel are notified, and the unmanned equipment is mobilized to jointly investigate the abnormal point; If the patrol abnormality is resolved, continue to patrol according to the patrol plan. If the abnormal situation needs to be upgraded, notify the personnel for processing according to the pre-set abnormal management requirements, until the patrol abnormality is resolved. After the abnormality is removed, the patrol plan is modified according to the fault point, the abnormal area or vehicle part is highlighted, and the patrol is carried out from multiple directions and angles until the patrol plan is completed, and a patrol report is generated. 2.The working method of the bus station intelligent safety management system based on BDS / UWB combined positioning according to claim 1, wherein, In step (1), the joint positioning device positioning method is as follows: (11) Coordinate conversion: Beidou coordinate conversion: (x ITM ,y UTM ) = wgs2utm(lat,lon) Wherein, lat,lon are longitude and latitude coordinates output by Beidou; x UTM ,y UTM are converted UTM plane coordinates; wgs2utm is a coordinate conversion function; (12) Time alignment: where: t is the time stamp that the target is aligned to; z k k+1 are the original observations at time t k k+1 ; z synced is the interpolated result aligned to t.​​ (13) Dynamic precision weight distribution: weight distribution based on sensor intrinsic indicators; BDS weight: where N sat is the current number of satellites, N max is the maximum number of visible satellites, HDOP is the horizontal dilution of precision, and ε is a small constant to prevent division by zero. UWB weight: Wherein: SNR is the signal-to-noise ratio, and a and β are parameters calibrated through experiments; Weighted fused position P fused : where: P BDS is the Beidou position, P UWB is the UWB position. 3.The working method of the bus station intelligent safety management system based on BDS / UWB combined positioning according to claim 2, wherein, In step (2), the patrol path planning step is: (21) Input parameters: Patrol area topology G=(V,E): wherein the node set V = {v1, v2,..., v n} represents a patrol key point; An edge set E, representing paths between nodes, each edge e ij Comprising: a distance d ij ; Security risk weight w ij ∈ [0,1], the greater the value, the higher the risk, and the need for priority patrol; patrol time cost t ij = d ij / v, v is the speed of security personnel or unmanned equipment; Patrol node attributes, including: Importance weightings p i ∈ [1, 5], the greater the value indicates a higher patrol frequency minimum patrol interval At i ; Patrol path algorithm parameters, including: The number of ants m, indicating the number of patrol personnel or unmanned equipment; Information pheromone importance factor a; Heuristic importance factor β; Information pheromone evaporation rate p∈(0,1); Information pheromone increment strength Q; Maximum number of iterations T max ; (22) Path selection by using the ant colony algorithm: (221) The probability of the patrol device k selecting the next node j at the node i: where τ ij is the pheromone concentration on edge e ij . eta ij : Heuristic function, defined as: (222) Patrol environment information pheromone update: Global update: Wherein, the increment of the kth patrol device: wherein Cost k is the total cost of the path; Time k is the total time for the path; ρ p importance weight of a node in the path; Local update: τ ij ←(1-ρ)τ ij +ρ·τ0; (223) Patrol dynamic constraint condition processing: Patrol node coverage constraint: A penalty term is introduced if the patrol interval of a node v i exceeds Δt i : Add a penalty term to the path cost calculation; Multi-patrol group cooperation constraint: Divide the patrol devices into K groups, each group independently searches the path, and the objective function is changed to: Through the above path planning of the ant colony algorithm, the design and formulation of each patrol path in the bus station are realized, the patrol constraints are set according to the work needs, the patrol range is controlled, the multi-group cooperative patrol is realized, and the man-machine cooperative patrol mechanism is realized.

4. A bus station intelligent safety management system based on BDS / UWB combined positioning, applied to the working method of the bus station intelligent safety management system based on BDS / UWB combined positioning in claim 1, characterized in that, It includes a joint positioning device, an artificial patrol scanning device, an unmanned automatic patrol device, and a network service platform, wherein: The joint positioning device: through BDS / UWB joint positioning, positioning in different spaces and environments is realized; The artificial patrol device: the security personnel receive the patrol plan and patrol route through the artificial patrol device, report abnormal conditions and position information, and detect the safety indicators related to the electric vehicle; The unmanned automatic patrol device: automatically receives the patrol plan, and according to the plan and the patrol route, the safety of the vehicle in the station or other loose areas is checked. The unmanned automatic patrol device has visual perception function and reports the network service platform after detecting the abnormality; The network service platform: controls the vehicle operation, controls the joint positioning device, the artificial patrol device, and the unmanned automatic patrol device. 5.The intelligent safety management system for bus station based on BDS / UWB combined positioning according to claim 2, wherein, The joint positioning device includes: The receiving unit includes a BDS module and a UWB module, which is used for receiving Beidou positioning data and UWB communication distance data; The computing unit includes time and space synchronization of Beidou positioning data and UWB data, UWB data conversion, data processing, filtering resolution, filtering iteration and abnormal observation detection, BDS / UWB joint positioning is realized according to different scene applications, and joint positioning data is generated; The output unit outputs the data.

Images