Dynamic traffic flow pressure prediction method and system for parking lot

Abstract

The invention provides a dynamic traffic flow pressure prediction method and system for a parking lot, and is applied to the field of Internet data processing. The dynamic traffic flow network diagram in the parking lot is gradually constructed from real-time acquisition of entrance vehicle information, a pressure sensing mechanism is started at a channel level, real-time monitoring and fine sensing of local area flow are realized, and compared with a traditional mode of only depending on entrance and exit counting, the method has the advantages that the real-time monitoring and fine sensing are realized; according to the scheme, microcosmic flow characteristics such as local density, passing rate and queuing increasing speed of specific nodes can be captured in the parking lot, potential congestion nodes are judged and identified in time in combination with real-time pressure increasing rate, and the traffic flow distribution accuracy is improved by identifying adjacent areas and dynamically implementing a flow distribution strategy. Local diversion and active intervention before congestion formation are realized, and congestion diffusion and secondary congestion are effectively avoided.

Description

Technical Field

[0001] This invention relates to the field of Internet data processing, and in particular to a method and system for predicting dynamic traffic flow pressure in parking lots. Background Technology

[0002] Current parking lot traffic flow prediction mainly relies on entrance and exit statistics and uses overall parking space occupancy rate to estimate trends. However, it ignores the complex road topology inside the parking garage, such as the impact of different lane widths, turning radii, ramp locations, and pillar obstructions on local traffic capacity. This modeling method essentially treats the entire parking garage as a single container, which cannot describe the spatial distribution characteristics of traffic flow inside the garage, nor can it perceive changes in vehicle speed, queue length, and local density in different areas. Therefore, it cannot predict the congestion that will occur at a specific node. Summary of the Invention

[0003] This invention aims to address the current lack of real-time perception and traffic flow management capabilities for localized traffic flow within parking garages, and provides a method and system for predicting dynamic traffic flow pressure in parking lots.

[0004] The present invention employs the following technical means to solve the technical problem: This invention provides a method for predicting dynamic traffic flow pressure in parking lots, comprising: Based on the vehicle information pre-entered at the parking lot entrance and exit, traffic flow data corresponding to the vehicle information is generated in real time. Determine whether the traffic flow data has reached a preset traffic flow threshold; If so, then based on the internal traffic structure preset by the parking terminal, a traffic flow network map inside the parking lot is constructed. Based on the basic channel attributes preset by the parking terminal for the garage passage, the channel sensing device of the garage passage is dynamically activated. Through the channel sensing device, the pressure index of the garage passage is collected. The internal traffic structure specifically includes turning nodes, pillar positions and ramp entrances. The basic channel attributes specifically include width, speed limit and capacity. The pressure index specifically includes local density, congestion index, throughput and queue growth rate. Determine whether the growth rate of the pressure index exceeds a preset rate threshold; If the growth rate is exceeded, the congested nodes inside the parking lot are obtained based on the growth rate, the adjacent areas of the congested nodes are identified, and the diversion strategy of the garage passage is dynamically adjusted according to the adjacent areas. Based on the diversion strategy, the parking rights of the congested nodes within a preset range are restricted. The adjacent areas specifically include entrance ramps, exit convergence points and main roads, and the diversion strategy specifically includes internal diversion and external flow restriction.

[0005] Furthermore, after the step of constructing a traffic flow network map inside the parking lot based on the preset internal traffic structure of the parking lot terminal, the method further includes: Based on the pre-divided vehicle traffic lanes of the internal traffic structure, the access area after a vehicle enters the parking lot through the parking lot entrance / exit is obtained. Determine whether the access area intersects with the vehicle traffic lane; If so, the frequency of the vehicles meeting in the traffic lanes is collected, and the traffic patterns of the traffic lanes are dynamically adjusted according to the meeting frequency. Based on the traffic patterns, path guidance information for the vehicles inside the parking lot is generated, wherein the traffic patterns specifically include one-way traffic, temporary closure, and detour.

[0006] Furthermore, the step of dynamically activating the channel sensing device of the garage passage based on the preset basic attributes of the garage passage by the parking terminal also includes: Based on the pre-set congestion contagion radius of the parking lot terminal, the sensing spread efficiency of the garage passage from one node to another is collected. Determine whether the sensing spread efficiency reaches a preset efficiency threshold; If so, then based on the target content pre-collected by the parking terminal for vehicles, the node traffic flow offset of the garage passage within a preset time period is constructed. Based on the node traffic flow offset and the passage geometry of the garage passage, the entry and exit weight of the garage passage at the parking terminal is dynamically adjusted. The target content specifically includes the exit direction, parking blocks and charging pile distribution, and the passage geometry specifically includes narrow passages, blind corners, sharp bends and ramps.

[0007] Furthermore, the step of identifying the adjacent areas of the congested node and dynamically adjusting the traffic diversion strategy of the garage passage based on the adjacent areas also includes: Based on the vehicle passage clearance pre-detected by the parking terminal for the garage passage, the vehicle flow rate reduction rate of the congestion node is identified. Determine whether the vehicle flow rate decreases continuously; If not, then obtain the vehicle congestion type of the congestion node, guide the vehicle queue length of the congestion node according to the vehicle congestion type, and dynamically introduce the internal diversion mechanism preset by the parking terminal according to the vehicle queue length. The vehicle congestion type specifically includes flow-type congestion, behavior-type congestion and structure-type congestion, and the internal diversion mechanism specifically includes volume restriction, speed restriction and target restriction.

[0008] Furthermore, the step of determining whether the traffic flow data has reached a preset traffic flow threshold also includes: Based on the number of vehicles pre-identified in the garage passage, the spacing information between a single vehicle and another vehicle is identified; Determine whether the spacing information reaches a preset spacing threshold; If not, then based on the traffic flow data, the average speed of vehicles in the garage passage is detected, and based on the average speed, abnormal vehicle behavior in the garage passage is identified, wherein the abnormal vehicle behavior specifically includes abnormal driving, abnormal path and abnormal vehicle characteristics.

[0009] Furthermore, the step of determining whether the growth rate of the pressure index exceeds a preset rate threshold also includes: Based on the pre-detected vehicle types in the garage passage, the pressure contribution values ​​of different vehicles to the garage passage are collected, wherein the vehicle types specifically include standard vehicles, large vehicles and special vehicles; Determine whether the pressure contribution values ​​are balanced; If not, then based on the vehicle type, obtain the pressure boosting information of different vehicles on the garage passage, and based on the pressure boosting information, generate the time required for the garage passage to go from single-point congestion to patchy congestion. Specifically, the pressure boosting information refers to the decrease in the passage capacity of the garage passage caused by differences in the size of different vehicles occupying the lane, their speed, and their operating behavior.

[0010] Furthermore, the step of generating traffic flow data corresponding to the vehicle information in real time based on the pre-entered vehicle information at the parking lot entrance and exit also includes: Based on the parking terminal's preset entrance and exit points for the parking lot entrance and exit, the corresponding area parking space is identified when a vehicle passes through the entrance and exit point; Determine whether the number of vehicles parked in the parking spaces of the area has reached a preset parking saturation value; If so, the parking terminal guides the vehicles for parking and diversion. Based on the parking diversion, the total number of vehicles inside the parking lot is dynamically updated. Specifically, the parking diversion includes guiding vehicles to turn around and re-enter the parking lot from other entrances and exits, guiding vehicles from the current parking space to other parking spaces, and guiding vehicles to leave the parking lot.

[0011] The present invention also provides a dynamic traffic flow pressure prediction system for parking lots, comprising: The generation module is used to generate traffic flow data corresponding to the vehicle information pre-entered at the parking lot entrance and exit in real time. The judgment module is used to determine whether the traffic flow data has reached a preset traffic flow threshold; The execution module is used to, if so, construct a traffic flow network map inside the parking lot based on the internal traffic structure preset by the parking lot terminal, dynamically activate the channel sensing device of the parking lot channel according to the channel basic attributes preset by the parking lot terminal for the parking lot channel, and collect the pressure index of the parking lot channel through the channel sensing device. The internal traffic structure specifically includes turning nodes, pillar positions and ramp entrances, the channel basic attributes specifically include width, speed limit and capacity, and the pressure index specifically includes local density, congestion index, throughput and queue growth rate. The second judgment module is used to determine whether the growth rate of the pressure index exceeds a preset rate threshold. The second execution module is used to, if the growth rate is exceeded, obtain the congested nodes inside the parking lot, identify the adjacent areas of the congested nodes, dynamically adjust the diversion strategy of the garage passage according to the adjacent areas, and restrict the vehicle parking rights of the congested nodes within a preset range according to the diversion strategy. The adjacent areas specifically include entrance ramps, exit convergence points and main roads, and the diversion strategy specifically includes internal diversion and external flow restriction.

[0012] Furthermore, it also includes: The acquisition module is used to acquire the access area of ​​a vehicle after it enters the parking lot through the parking lot entrance / exit, based on the vehicle traffic lanes pre-divided by the internal traffic structure. The third judgment module is used to determine whether the access area intersects with the vehicle traffic lane; The third execution module is used to collect the frequency of the vehicles' intersections in the vehicle traffic lanes if the condition is met, dynamically adjust the traffic patterns of the vehicle traffic lanes based on the intersection frequencies, and generate path guidance information for the vehicles within the parking lot based on the traffic patterns. The traffic patterns specifically include one-way traffic, temporary closures, and detours.

[0013] Furthermore, the execution module also includes: The data acquisition unit is used to collect the sensing spread efficiency of the garage passage from one node to another based on the congestion spread radius preset by the parking lot terminal. The judgment unit is used to determine whether the sensing spread efficiency reaches a preset efficiency threshold. The execution unit is configured to, if so, construct the node traffic flow offset of the garage passage within a preset time period based on the target content pre-collected by the parking terminal, and dynamically adjust the entry and exit weight of the garage passage at the parking terminal based on the node traffic flow offset and the passage geometry of the garage passage. The target content specifically includes the exit direction, parking blocks and charging pile distribution, and the passage geometry specifically includes narrow passages, blind corners, sharp bends and ramps.

[0014] This invention provides a method and system for predicting dynamic traffic flow pressure in parking lots, which has the following beneficial effects: This invention begins by acquiring real-time vehicle information at the entrance, gradually constructing a dynamic traffic flow network map within the parking lot. By activating a pressure sensing mechanism at the channel level, it achieves real-time monitoring and precise perception of traffic flow in local areas. Compared to traditional methods that rely solely on entrance and exit counts, this solution can capture micro-level traffic flow characteristics such as local density, throughput, and queue growth rate at specific nodes within the parking lot. Combined with real-time pressure growth rate judgment, it promptly identifies potential congestion nodes. Through the identification of adjacent areas and the dynamic implementation of diversion strategies, it achieves localized diversion and proactive intervention before congestion occurs, effectively preventing congestion spread and secondary blockages. This significantly improves the parking lot's visibility and responsiveness to internal traffic flow, achieving a technological leap from "entrance traffic perception" to "real-time perception of internal local traffic flow," thereby improving overall traffic efficiency and management intelligence. Attached Figure Description

[0015] Figure 1 This is a flowchart illustrating an embodiment of the dynamic traffic flow pressure prediction method for parking lots according to the present invention. Figure 2 This is a structural block diagram of an embodiment of the dynamic traffic flow pressure prediction system for parking lots according to the present invention. Detailed Implementation

[0016] It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention. The realization of the purpose, functional features and advantages of the invention will be further explained in conjunction with the embodiments and with reference to the accompanying drawings.

[0017] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present invention, and not all of them. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0018] Reference Appendix Figure 1 A dynamic traffic flow pressure prediction method for parking lots according to one embodiment of the present invention includes: S1: Based on the vehicle information pre-entered at the parking lot entrance and exit, generate traffic flow data corresponding to the vehicle information in real time; S2: Determine whether the traffic flow data has reached a preset traffic flow threshold; S3: If so, then according to the internal traffic structure preset by the parking terminal, construct the traffic flow network map inside the parking lot, and dynamically activate the channel sensing device of the parking garage channel according to the channel basic attributes preset by the parking terminal for the garage channel. Through the channel sensing device, collect the pressure index of the garage channel. The internal traffic structure specifically includes turning nodes, pillar positions and ramp entrances. The channel basic attributes specifically include width, speed limit and capacity. The pressure index specifically includes local density, congestion index, throughput and queue growth rate. S4: Determine whether the growth rate of the pressure index exceeds a preset rate threshold; S5: If the growth rate is exceeded, the congested nodes inside the parking lot are obtained based on the growth rate, the adjacent areas of the congested nodes are identified, and the diversion strategy of the garage passage is dynamically adjusted according to the adjacent areas. Based on the diversion strategy, the parking rights of the congested nodes within a preset range are restricted. The adjacent areas specifically include entrance ramps, exit convergence points and main roads, and the diversion strategy specifically includes internal diversion and external flow restriction.

[0019] In this embodiment, the system generates real-time traffic flow data corresponding to vehicles entering or leaving the parking lot based on pre-entered vehicle information at the parking lot entrances and exits. The system then determines whether this traffic flow data reaches a pre-set traffic flow threshold to execute corresponding steps. For example, if the system determines that the parking lot's traffic flow data has not reached the pre-set threshold, it considers the overall traffic flow in the parking lot to be acceptable and manageable, without potential overload or congestion risks. The system maintains a low-resource-occupancy monitoring state, performing only routine statistics on vehicle information at entrances and exits without activating channel-level pressure acquisition equipment. Simultaneously, the channels do not trigger dynamic interventions such as diversion, flow restriction, or bidirectional / unidirectional switching. Vehicles proceed according to default... The system recognizes autonomous traffic flow and continuously accumulates historical operational data under normal traffic conditions for later identification of abnormal fluctuations. For example, when the system determines that the traffic flow data in the parking lot has reached a pre-set traffic flow threshold, it considers the overall traffic flow in the parking lot to be relatively saturated and may experience traffic congestion. The system then constructs a traffic flow network map within the parking lot based on the pre-set internal traffic structure at the parking terminal, which includes turning nodes, pillar locations, and ramp entrances. Based on the pre-set basic attributes of the parking garage passages at the parking terminal, including width, speed limit, and capacity, the system dynamically activates the passage sensing devices in the parking garage passages. These devices specifically include vehicle detection geomagnetic sensors, laser rangefinders, and millimeter-wave detectors. Millimeter-wave radar and video recognition cameras, through these channel sensing devices, collect pressure indicators of garage passageways, specifically including local density, congestion index, throughput, and queue growth rate. The system constructs an internal traffic flow network model using internal traffic structure data (turning points, pillars, ramp entrances), transforming macro-statistical data into localized, structured predictive capabilities. This process upgrades the system from a weak monitoring mode that only monitors entrance traffic flow to a strong monitoring mode that tracks the actual flow of traffic throughout the entire garage, enabling early detection of congestion risks rather than reactive responses. Simultaneously, based on the basic attributes of the passageways (width, speed limit, capacity), different types of channel sensing devices are dynamically activated, resulting in more accurate and energy-efficient sensing. For example, millimeter-wave radar and ramp entrances are prioritized for narrow sections. Video detection nodes are activated at level crossings, while geomagnetic sensors can be used on wide, straight roads. This "on-demand activation" method improves the authenticity and completeness of sensor data while reducing redundant equipment work and improving system energy efficiency. The resulting traffic flow data more accurately reflects the degree of node constraint. Furthermore, the pressure indicators collected by the system include local density, congestion index, throughput, and queue growth rate. These indicators are no longer based on manual judgment or guesswork from entry data, but are measured based on real-time node data. This allows the system to form a quantifiable local pressure field characterization, thus providing a reliable basis for subsequent decisions such as "whether diversion is needed," "whether traffic restrictions are needed," "which area has the greatest pressure," and "whether secondary congestion may occur."The system then determines whether the growth rate of the pressure index in the parking garage aisle exceeds a pre-set threshold, and executes the corresponding steps accordingly. For example, if the system determines that the growth rate of the pressure index in a certain parking garage aisle does not exceed the pre-set threshold, the system considers the local density or congestion index of the parking lot to be at a moderate level, but the growth rate is slow, without any short-term sharp accumulation, belonging to "stable or slowly accumulating pressure." The system will continue to maintain the normal monitoring sampling frequency, without needing to intervene in vehicle parking permissions, while recording the current pressure index as a trend reference, focusing on observing the traffic flow growth trend rather than immediate intervention, and maintaining a low-energy observation mode. The system continues to await data sampling and analysis for the next cycle. For example, when the system determines that the growth rate of the pressure index for a certain parking garage aisle exceeds a pre-set threshold, the system considers a short-term, rapid accumulation of congestion in the parking lot. Based on different growth rates, the system identifies congested nodes within the parking lot and their adjacent areas, including entrance ramps, exit convergence points, and main roads. Based on these adjacent areas, the system dynamically adjusts the traffic diversion strategy for the parking garage aisles. This strategy includes internal diversion and external flow restriction. Depending on the specific diversion strategy, the system restricts vehicle parking access within a pre-defined range for congested nodes. The system then analyzes the increase in pressure index... Rather than simply relying on current pressure values, the system can identify early signs of "short-term, rapid congestion"—sudden congestion caused by an instantaneous influx of traffic exceeding capacity. By precisely locating congestion nodes based on different growth rates, the system can intervene promptly before congestion spreads significantly, avoiding the lag problem of traditional parking lots that only address congestion after it has already formed. This greatly improves the speed and sensitivity of internal traffic response. Furthermore, by identifying the area to which the congestion node belongs (entrance ramp, exit convergence point, main road), the system can dynamically select corresponding diversion strategies based on the area type. For example, entrance ramps trigger external flow control, main roads trigger internal detours, and exit convergence points prioritize adjusting queue formations. This regionalized and scenario-based traffic diversion approach avoids secondary congestion problems caused by a "one-size-fits-all" strategy, making intervention measures more targeted and efficient. It can accurately alleviate pressure at the source of congestion and dynamically restrict parking permissions in designated areas based on different types of diversion strategies (internal diversion and external flow restriction). This directly reduces the number of vehicles entering congested nodes, weakening the pressure growth trend at the source. This "proactive pressure reduction" method allows traffic flow to redistribute within the network, forming a more balanced pressure field and ensuring that key nodes are not overloaded. Compared to traditional methods of simply providing prompts or manual intervention, this system can automatically execute and maintain a stable traffic flow structure, improving the overall operational efficiency and safety of the parking lot.

[0020] It should be added that, based on the growth rate, congested nodes inside the parking lot are obtained, and adjacent areas of the congested nodes are identified. Based on these adjacent areas, the traffic diversion strategy of the parking garage access is dynamically adjusted. The adjacent areas specifically include entrance ramps, exit convergence points, and main roads. The traffic diversion strategy specifically includes internal diversion and external flow restriction, specifically as follows: I. Identifying congested nodes based on growth rate: The system identifies areas experiencing rapid traffic congestion within a short period by monitoring the growth rate of pressure indicators (such as the rate of increase in local density, the rate of increase in queue length, and the congestion index change curve). For example, if the pressure index of the main road A in the garage increases from 0.2 to 0.9 within 30 seconds, it indicates that vehicles are rapidly accumulating at this node and the traffic capacity is clearly insufficient; therefore, the system marks the main road A as a "congested node". Key technical point: Instead of judging congestion by absolute flow, it judges whether congestion is forming by the "growth rate", identifying "potential outbreak points" rather than just points that are already blocked; II. Identify the adjacent areas of congested nodes: The system then determines the structural area where the congestion node is located, such as: entrance ramp, exit convergence point, main road, because the traffic behavior in different areas is different, and the causes and solutions to congestion are also different. For example, if the congestion point is located at the exit convergence point, it means that vehicles from multiple lanes are converging towards the exit lane, and vehicles are queuing to leave the parking lot, resulting in an exit bottleneck. If the congestion point is located at the entrance ramp, it means that vehicles are converging to enter the parking lot, and there is dense activity in searching for parking spaces, causing traffic congestion at the entrance downhill. III. Dynamically adjust the diversion strategy based on regional characteristics: The system dynamically configures different traffic flow control strategies based on the congestion characteristics of different areas, including: internal diversion, which refers to redirecting vehicle routes within the parking lot. For example, if there is congestion at the exit convergence point, some vehicles will be guided to another exit; if there is congestion on main road A, vehicles will be guided to detour through lane B or area C. For example: When vehicles leaving the warehouse encounter congestion at Exit 1, the system displays the following message on the navigation screen: "Exit 1 is congested. We suggest using Exit 2 instead." Alternatively, the system may use the following guidance light system: Exit 1 → Red light, Exit 2 → Green light. External traffic control refers to restricting vehicles from entering parking lots or specific areas. For example, when the pressure on the entrance ramp increases rapidly, the system may take measures such as: reducing the frequency of entry gate releases, coordinating with the external traffic system to restrict the entry of external vehicles, or temporarily closing certain entrances. For example, if the growth rate of the entrance ramp surges, the system will increase the entry gate clearance interval from 1 second to 4 seconds, thereby slowing down the inflow of vehicles. IV. Restrict parking permissions at congested nodes. To prevent congested areas from continuing to attract vehicles to parking areas, the system implements parking restrictions in those areas. For example, if the algorithm determines that a parking space in zone D next to main road A is within the congestion area, the system will: pause vehicle navigation to zone D, display that the parking space in zone D is unavailable, and guide the vehicle to a more distant but more accessible parking space. V. Complete practical operation examples, Assuming the congestion point occurs at exit convergence point X: the system detects an abnormal growth rate in the queue of vehicles, identifies its adjacent area as an exit convergence point, and initiates a traffic diversion strategy. Internal diversion: directing vehicles to another exit Y; External traffic control: reducing the speed at which new cars enter the warehouse; The system restricts parking permissions in parking spaces around Exit X to prevent more vehicles from parking near the exit and exacerbating queuing. The system continuously monitors the growth rate until it falls back to the normal threshold. Result: The queuing pressure at Exit X was gradually relieved, the congestion dissipated, and the traffic flow was redistributed in a balanced manner.

[0021] In this embodiment, after step S3 of constructing the traffic flow network map inside the parking lot based on the preset internal traffic structure of the parking lot terminal, the method further includes: S301: Based on the pre-divided vehicle traffic lanes of the internal traffic structure, obtain the access area after a vehicle enters the parking lot through the parking lot entrance / exit; S302: Determine whether the access area intersects with the vehicle traffic lane; S303: If so, the frequency of the vehicles meeting in the traffic lanes is collected, and the traffic pattern of the traffic lanes is dynamically adjusted according to the meeting frequency. Based on the traffic pattern, the path guidance information of the vehicles inside the parking lot is generated. The traffic pattern specifically includes one-way traffic, temporary closure and detour.

[0022] In this embodiment, the system, based on pre-defined vehicle traffic lanes within the internal traffic structure, acquires the access areas after a vehicle passes through a parking lot entrance / exit and enters the parking lot. The system then determines whether these access areas intersect with vehicle traffic lanes to execute corresponding steps. For example, if the system determines that a vehicle's access areas after entering the parking lot do not intersect with vehicle traffic lanes, the system assumes that the vehicle's target route will not compete with or conflict with other major traffic flows. The system then reduces the real-time sampling frequency for that vehicle, meaning the monitoring index by cameras and radar decreases. Triggering additional stress analysis in this area, system resources are concentrated on other high-risk nodes, while the vehicle's navigation path is optimized, recommending shorter routes to facilitate faster parking. For example, when the system determines that a vehicle will converge on a traffic lane after entering the parking lot from the entrance / exit, the system considers this vehicle's target route to be competing with other major traffic flows. The system collects the vehicle's convergence frequency on the traffic lanes and dynamically adjusts the traffic patterns of these traffic lanes based on different convergence frequencies. The traffic patterns specifically include one-way traffic. The system employs various traffic management techniques, including temporary closures and detours, to generate route guidance information for vehicles within the parking lot, based on different traffic patterns. By identifying whether a vehicle's path passes through traffic lanes, the system can preemptively determine if a vehicle is a potential "conflict participant." It collects vehicle convergence frequencies and dynamically determines traffic patterns based on the merging characteristics between nodes, ensuring vehicles receive clear path assignments before entering the convergence area. This reduces random congestion caused by behaviors such as cutting in, vying for lanes, and yielding. Furthermore, it adjusts the traffic patterns of traffic lanes based on different vehicle convergence frequencies; for example, when the convergence frequency is too high... The system employs temporary one-way traffic, partially closing off areas or initiating detour guidance when traffic pressure is high in a particular direction. This flexible and variable traffic rule allows the passageway to operate in a flexible, adaptable manner, rather than being fixed. Based on the temporarily adjusted traffic patterns, the system generates real-time route guidance information for each vehicle, enabling vehicles to quickly select the optimal route in the complex environment of the parking garage, rather than relying on the driver's judgment. Drivers do not need to consider whether they encounter relative traffic flow, whether they need to yield, or whether they may enter a congested section, thus significantly reducing decision-making delays and the probability of incorrect route selection.

[0023] In this embodiment, step S3, which dynamically activates the channel sensing device of the garage passage based on the basic channel attributes preset by the parking terminal for the garage passage, further includes: S31: Based on the congestion contagion radius preset by the parking lot terminal, collect the sensing spread efficiency of the garage passage from one node to another; S32: Determine whether the sensing propagation efficiency reaches a preset efficiency threshold; S33: If so, then based on the target content pre-collected by the parking terminal for vehicles, construct the node traffic flow offset of the garage passage within a preset time period, and dynamically adjust the entry and exit weight of the garage passage at the parking terminal according to the node traffic flow offset and the passage geometry of the garage passage. The target content specifically includes the exit direction, parking blocks and charging pile distribution, and the passage geometry specifically includes narrow passages, blind corners, sharp bends and ramps.

[0024] In this embodiment, the system collects the congestion spread efficiency from one node to another in the parking garage passage based on the pre-set congestion spread radius of the parking garage terminal. The system then determines whether this spread efficiency reaches a pre-set efficiency threshold and executes corresponding steps accordingly. For example, if the system determines that the sensing spread efficiency from one node to another in the parking garage passage has not reached the pre-set efficiency threshold, the system considers the current congestion to be localized, and its impact range will not spontaneously expand. The system will mark this congestion as "locally controllable congestion," maintaining local monitoring of the area, not triggering cross-regional diversion strategies, while maintaining normal traffic flow patterns at the nodes and keeping existing traffic flow guidance unchanged. There is no need to implement one-way flow restriction or path blocking, and the risk of congestion is reduced. The data processing priority in this area is to monitor trends only within necessary limits. For example, when the system determines that the sensing spread efficiency of the parking garage aisle from one node to another has reached a pre-set efficiency threshold, the system considers the current congestion to be global and its impact range to be spontaneously expanding. The system will then construct the node traffic flow offset of the parking garage aisle within a pre-set time period based on the pre-collected entry target information from the parking terminal, specifically including exit direction, parking area, and charging pile distribution. Based on different node traffic flow offsets and the aisle geometry (including narrow passages, blind spots, sharp bends, and ramps), the system will dynamically adjust the entry and exit weights of the parking garage aisle at the parking terminal. When the spread efficiency reaches the threshold... After the system detects the congestion, it determines that the congestion is no longer a localized, isolated phenomenon, but rather a global congestion with a spreading trend. This means that the congestion pressure can be transmitted between multiple nodes in the parking garage, forming a chain reaction of congestion. By recognizing this spreading signal, the system upgrades its scheduling strategy from local intervention to global governance, enabling rapid blocking and buffering before the congestion fully develops. This prevents a blockage at one node from paralyzing traffic on the entire parking garage's main road or multi-level structure. Simultaneously, based on the vehicle's entry target (exit direction, parking area, charging station distribution), the system constructs traffic flow deviations for different nodes in the future time period. This means the system not only focuses on the vehicle's current location but also predicts in advance the area the vehicle will eventually move to, thus enabling proactive control. Predictive scheduling significantly enhances the system's proactive ability to handle congestion. Instead of passively responding to congestion, the system proactively restructures traffic flow, reducing the spread of congestion at its source and resulting in a more balanced traffic distribution. By considering the geometric characteristics of different lanes (narrow lanes, blind spots, sharp bends, ramps), the system can dynamically adjust the entry and exit weights of each lane. For example, narrow lanes are appropriately downweighted during congestion spread, while main roads with open views have higher entry and exit priorities. In this way, the system can guide traffic flow to routes with higher capacity and lower risk, making the fluid traffic flow structure conform to physical traffic characteristics, reducing the probability of "structural congestion," and ultimately improving overall traffic efficiency. This makes traffic flow scheduling more in line with actual lane characteristics and road safety requirements.

[0025] It should be noted that, based on the target content pre-collected by the parking lot terminal for vehicles, the node traffic flow offset of the garage passage within a preset time period is constructed. Based on the node traffic flow offset and the passage geometry of the garage passage, the entry and exit weights of the garage passage at the parking lot terminal are dynamically adjusted, specifically as follows: The system first predicts vehicle travel trends within a preset time period based on pre-collected data from parking lot terminals, including future exit directions, parking areas, and demand for charging station distribution. It then constructs a data structure for traffic flow offset at each node of the parking garage aisle during this period to determine if certain nodes will experience directional traffic concentration. Further, the system considers the aisle geometry, including narrow passages, blind spots, sharp bends, and ramps, to weightedly evaluate the traffic flow offset at each node, identifying which nodes, while experiencing increased traffic flow, also exhibit path characteristics that amplify congestion risks. Finally, based on the combined assessment of node traffic flow offset and aisle geometry, the system dynamically adjusts the entry and exit weights for each parking garage aisle. This allows high-capacity, high-quality aisles to handle more traffic, while reducing vehicle allocation to low-capacity and high-risk aisles. This proactively suppresses the spread of congestion and achieves early control and optimization of traffic flow within the parking lot. To illustrate the scenario, consider this example: In a five-story underground parking garage, 30 vehicles enter within a short period. Of these, 18 vehicles are heading towards area D, 7 vehicles are heading towards the charging station area, and 5 vehicles are near parking spaces at Exit 2. Based on the target information, the system predicts the traffic flow shift over the next 10 minutes, as shown in Table 1 below. Table 1:

[0026] The system considers the channel geometry and performs structural identification on the channels corresponding to these nodes, as shown in Table 2 below. Table 2:

[0027] The system comprehensively assesses and adjusts the entry and exit weights. Although P2 has the highest offset, its structure is wider, so its weight is increased, leading to more vehicles being guided into area D through this channel. R1 has a high offset and is a narrow ramp → weight is reduced → the system will prioritize guiding charging vehicles to the R2+R3 detour path. Q4 has a medium offset but has blind corners and sharp bends → weight is reduced moderately → vehicles are guided to connect to the exit from another straighter channel Q2. System behavior output (driver's actual experience): In the parking navigation interface, vehicles in Zone D will receive a recommended route: → "It is recommended to go straight from P2 to enter the main path of Zone D." Vehicles heading to the charging station will receive a prompt: → "High risk of congestion in R1 lane, please turn slightly right and take the R2-R3 route to reach the charging area." The 5 vehicles preparing to leave will receive guidance: → "Temporarily avoid the sharp bend at Q4, it is recommended to detour via the main road to Exit 2." Ultimately, the system achieves proactive control and diversion of potential congestion within parking lots by predicting future traffic flow trends, assessing the carrying capacity of the road structure, and dynamically adjusting the entry and exit weights of the lanes, resulting in a smoother, smarter, and more efficient distribution of traffic flow.

[0028] In this embodiment, step S5, which identifies the adjacent areas of the congested node and dynamically adjusts the traffic diversion strategy of the garage passage based on the adjacent areas, further includes: S51: Based on the vehicle passage clearance pre-detected by the parking terminal for the garage passage, identify the vehicle flow rate reduction rate of the congestion node; S52: Determine whether the vehicle flow rate decrease rate continues to decrease; S53: If not, obtain the vehicle congestion type of the congestion node, guide the vehicle queue length of the congestion node according to the vehicle congestion type, and dynamically introduce the internal diversion mechanism preset by the parking terminal according to the vehicle queue length. The vehicle congestion type specifically includes flow-type congestion, behavior-type congestion and structure-type congestion, and the internal diversion mechanism specifically includes volume restriction, speed restriction and target restriction.

[0029] In this embodiment, the system, based on the pre-detected vehicle throughput capacity of the parking garage lanes by the parking terminal, identifies the vehicle speed reduction rate at congested nodes. The system then determines whether this vehicle speed reduction rate is continuously decreasing and executes corresponding steps accordingly. For example, when the system determines that the vehicle speed reduction rate at a congested node is continuously decreasing, it considers the congestion level to be gradually decreasing. Instead of maintaining strict flow control measures, the system gradually restores the throughput weight of the lane corresponding to the congested node and appropriately reduces external flow restriction or detour strategies for that area. Simultaneously, the system maintains a low-level monitoring mode to observe the vehicle speed recovery and retains only a mild, flexible diversion strategy to allow traffic flow to... Without excessive intervention, the system naturally restores the balance of traffic flow in the affected area, improving overall traffic efficiency and avoiding the traffic disturbance effects caused by excessive scheduling. For example, if the system determines that the vehicle speed reduction rate at a congested node has not continued to decrease, it considers the congestion level to be still severe. The system will then obtain the type of vehicle congestion at the congested node, which specifically includes flow-related congestion, behavioral congestion, and structural congestion. Based on different vehicle congestion types, the system guides the length of the vehicle queue at the congested node. Based on different queue lengths, it dynamically introduces a pre-set internal diversion mechanism from the parking lot terminal. This internal diversion mechanism specifically includes volume limits, speed limits, and target limits. The system further manages the congested nodes... Traffic congestion type identification can determine whether the current congestion is due to "flow-based congestion" caused by excessive traffic volume, "behavioral congestion" caused by individual vehicles making slow decisions, temporarily stopping, or waiting to react, or "structural congestion" caused by bottlenecks in the channel geometry. This congestion type differentiation allows the system to adopt different response strategies based on the specific causes, improving scheduling efficiency, avoiding the use of a single, crude flow-limiting method, reducing misjudgments and excessive intervention, and dynamically performing queue compression, queue transfer, or queue segmentation based on the length of the vehicle queue at the congested node. This effectively reduces the fixed length of the congestion queue and reduces the pressure accumulation caused by vehicles stagnating due to long waiting times. This orderly queuing... Queue guidance can prevent vehicles from randomly cutting in, vying for lanes, hesitating, and making blind choices in congested areas, turning a chaotic state into a controllable one, thereby improving traffic flow and promoting the recovery of local traffic flow. Furthermore, once the congestion type and queue length are determined, the system introduces three types of internal diversion mechanisms: volume restriction (controlling the entry of large vehicles), speed restriction (limiting the impact of cutting in and sudden stops), and target restriction (re-planning parking target areas). This allows traffic flow to be redistributed to more suitable channels or target areas, thereby achieving global traffic optimization. This dynamic diversion can minimize the structural pressure in local areas, allowing traffic flow to form a flexible distribution within the parking lot, improving overall throughput and capacity efficiency.

[0030] It should be noted that the process involves obtaining the vehicle congestion type of the congestion node, determining the vehicle queue length based on the vehicle congestion type, and dynamically introducing the parking lot terminal's preset internal diversion mechanism based on the vehicle queue length. Specifically: I. Obtaining Vehicle Congestion Types at Congested Nodes: The system first identifies the vehicle congestion types at congested nodes in the parking lot; vehicle congestion types mainly include three categories: Traffic flow-related congestion: The number of vehicles passing through a node exceeds the designed capacity of the channel, forcing vehicles to slow down or wait in line; Behavioral congestion: Abnormal behavior of one or a small number of vehicles causes localized slowdown or stagnation, such as vehicles hesitating, reversing, or making temporary stops; Structural congestion: Vehicle movement is restricted due to the geometry of the passage (narrow, sharp bends, ramps, blind spots, etc.) or obstacles; By identifying the type of congestion, the system can determine the main cause of the congestion and decide on the appropriateness of subsequent intervention measures. For example: The long queue and dense traffic on the main road P1 → traffic flow congestion; vehicles frequently reversing in the charging station area E2 → behavioral congestion; and vehicles driving at low speed for a long time on the narrow ramp node R3 → structural congestion. 2. The system guides vehicle queue length based on congestion type. Utilizing congestion type information, the system dynamically adjusts the length of the vehicle queue at each node, achieving orderly passage through queue management. For traffic-intensive congestion nodes: control the vehicle entry rate and limit the queue length within a controllable range to prevent downstream nodes from being overwhelmed; For behavioral congestion nodes: alleviate queue accumulation caused by localized deceleration by guiding vehicle spacing, changing lane order, or prompting operation; For structurally congested nodes: guide vehicles to alternative lanes with wider structures or higher traffic efficiency to shorten the queue length at the main node; For example: P1 flow-type congestion → the system limits the queue length to no more than 15 vehicles by restricting the flow at the entrance; E2 behavioral congestion → the system prompts "Please enter the parking space in order" through navigation to reduce queue jumping or reversing conflicts; R3 structural congestion → the system guides some vehicles to detour through the wider R4 ramp. 3. An internal diversion mechanism is dynamically introduced based on the vehicle queue length. The system determines whether to trigger the internal diversion mechanism based on the queue length and selects a specific strategy: Size restrictions: Large vehicles are restricted from entering congested nodes or narrow passages to avoid further reducing traffic capacity; Speed ​​limit: Control vehicle speed to prevent vibration transmission caused by sudden acceleration or deceleration; Target constraints: Adjust the target parking area for vehicles, guide vehicles to alternative areas or vacant parking spaces, and distribute the pressure; For example: P1 Traffic Congestion → Temporary Restriction on Large SUVs Entering the Entrance (Size Restriction). E2 behavioral congestion → Navigation instructions for vehicles to slow down by 5 km / h (speed limit). R3 structural congestion → redirect some vehicles to R4 or R5 areas (target restriction). In this way, the system can dynamically and accurately adjust the distribution of traffic flow based on actual congestion characteristics and queue status, thereby optimizing traffic flow within the parking lot, reducing pressure on congested nodes, and improving overall traffic efficiency.

[0031] In this embodiment, step S2, which determines whether the traffic flow data has reached a preset traffic flow threshold, further includes: S21: Based on the number of vehicles pre-identified in the garage passage, identify the spacing information between a single vehicle and another vehicle; S22: Determine whether the spacing information reaches a preset spacing threshold; S23: If not, then based on the traffic flow data, detect the average speed of vehicles in the garage passage, and based on the average speed, identify abnormal vehicle behavior in the garage passage, wherein the abnormal vehicle behavior specifically includes abnormal driving, abnormal path and abnormal vehicle characteristics.

[0032] In this embodiment, the system identifies the distance information between a single vehicle and another vehicle based on the pre-identified number of vehicles in the garage passage. The system then determines whether this distance information meets a pre-set distance threshold to execute corresponding steps. For example, when the system determines that the distance information between a vehicle and another vehicle meets the pre-set distance threshold, the system considers the vehicle distance to meet safe passage requirements, meaning that the vehicles maintain a controllable and safe distance to avoid collisions or brake shock transmission due to excessive closeness. The system will not restrict or close the passage or related nodes, maintaining the current passage priority and passage access rights. The system emphasizes real-time monitoring of vehicle spacing, but can appropriately reduce the sampling frequency to avoid excessive intervention that could cause unnecessary detours or deceleration. It also provides optimal route suggestions to ensure stable traffic flow in the garage. If a vehicle's target area is far away, it can suggest a reasonable route to avoid potential future congestion. For example, if the system determines that the distance between two vehicles does not meet a pre-set threshold, it considers the distance between vehicles too close, possibly due to excessive traffic congestion in the garage aisle. The system will then detect the average speed of vehicles in the garage aisle based on different traffic flow data. These average speeds identify abnormal vehicle behavior within the parking garage aisles, specifically including abnormal driving, abnormal routing, and abnormal vehicle characteristics. By monitoring traffic flow and vehicle spacing in real time, the system can detect potential congestion points in advance, allowing for intervention before collisions or sudden braking occur. This early warning mechanism helps reduce the probability of accidents and ensures safe vehicle passage within the parking lot. Furthermore, upon detecting insufficient vehicle spacing, the system calculates the average vehicle speed within the parking garage aisles based on traffic flow data, identifying abnormal vehicle behavior, including abnormal driving (such as sudden acceleration or braking) and abnormal routing (such as arbitrary lane changes or detours). 1. Identifying abnormal vehicle characteristics (such as excessively large vehicles affecting traffic flow): By recognizing such abnormal behavior, the system can take intervention measures for specific vehicles, reducing the contagion effect of local congestion on the overall traffic flow, improving the efficiency of the passage. After identifying abnormal vehicle behavior, the system can take dynamic adjustment measures based on the characteristics of congested nodes and passages, such as adjusting traffic weights, implementing internal diversion or route guidance. Through precise control of abnormal vehicles and congested areas, the system can not only alleviate the current congestion, but also optimize the traffic flow distribution within the entire parking lot, making vehicle traffic more balanced and smooth, thereby improving the overall operational efficiency of the parking lot and the user experience.

[0033] It should be noted that abnormal vehicle behavior can be: I. Abnormal driving behavior (operational abnormality) refers to behaviors that cause obstruction or affect flow speed due to improper operation by the driver, including: repeatedly reversing, adjusting the direction of the vehicle, hesitating for a long time, frequently braking suddenly or stopping to observe, giving way unnecessarily, and temporarily stopping and occupying the lane. This type mainly reflects the risk of obstruction caused by driving behavior. 2. Path deviation and non-compliant behavior (rule-based anomalies) refers to vehicles not traveling according to the preset traffic flow path or prescribed direction, including: driving in the wrong direction after missing the turn, driving in the wrong direction in violation of regulations, using non-designated exits / entrances, entering restricted traffic lanes, illegal parking or driving over the line. Such behaviors are highly dangerous and can easily cause multi-node chain congestion. III. Behaviors caused by abnormal vehicle characteristics (characteristic anomalies) refer to behaviors that differ from normal traffic flow due to the characteristics of the vehicle itself. These include: large vehicles causing slow turning and difficulty in cornering, long-wheelbase vehicles causing longer transit times, novice drivers causing slow speeds and excessive caution, uncoordinated deceleration caused by the mixing of autonomous and human-driven vehicles, and abnormal stagnation of heavily loaded / trailer vehicles when passing through nodes. This type is a special flow model anomaly caused by the superposition of physical characteristics and behavioral patterns.

[0034] In this embodiment, step S4, which determines whether the growth rate of the pressure index exceeds a preset rate threshold, further includes: S41: Based on the pre-detected vehicle types in the garage passage, collect the pressure contribution values ​​of different vehicles to the garage passage, wherein the vehicle types specifically include standard vehicles, large vehicles and special vehicles; S42: Determine whether the pressure contribution values ​​are balanced; S43: If not, then based on the vehicle type, obtain the pressure boosting information of different vehicles on the garage passage, and based on the pressure boosting information, generate the time required for the garage passage to go from single-point congestion to patchy congestion. Specifically, the pressure boosting information refers to the decrease in the passage capacity of the garage passage caused by differences in the size of different vehicles occupying the lane, their speed, and their operating behavior.

[0035] In this embodiment, the system collects the pressure contribution values ​​of different vehicle types pre-detected within the garage aisle, specifically including standard vehicles, large vehicles, and special vehicles. The system then determines whether these pressure contribution values ​​are balanced within the same garage aisle to execute corresponding steps. For example, if the system determines that the pressure contribution values ​​of different vehicles within the same garage aisle are balanced, it considers that the vehicle combination in that aisle does not generate significant unbalanced pressure under the current traffic conditions, the traffic flow is stable, and the aisle capacity is not occupied or compressed by a single vehicle type. The system maintains the originally assigned entry and exit weights and does not restrict specific vehicle types from entering the aisle, while continuing... The system collects vehicle type and pressure contribution values, but can reduce intervention priority, focusing on monitoring potentially new vehicle types entering the garage to prevent future imbalances. It also continues to provide navigation guidance to ensure vehicles travel along optimal paths, improving overall traffic efficiency. For example, when the system determines that the pressure contribution values ​​of different vehicles to the same garage aisle are unbalanced, it considers the traffic flow abnormal under current conditions, with the aisle's capacity easily occupied by a single vehicle type. The system will then acquire pressure boosting information for different vehicle types, specifically information on how vehicle size, speed, and operational behavior cause a decrease in the garage aisle's capacity. Based on this pressure boosting information, the system generates the time required for a garage lane to evolve from single-point congestion to widespread congestion. When the system determines that the pressure contribution values ​​of different vehicles to the same garage lane are unbalanced, it indicates that the vehicle composition in that lane is unbalanced under current traffic conditions. This may result in large or special vehicles occupying the lane and blocking the passage of standard vehicles. By acquiring pressure boosting information for different vehicle types (size, lane occupation, speed, and operational behavior differences), the system can accurately identify potential congestion nodes and high-risk lanes, detect abnormal traffic flow conditions in advance, and prevent local congestion from spreading into global congestion. Furthermore, based on the pressure boosting information of different vehicles, the system generates the time required for a garage lane to develop from single-point congestion to widespread congestion. The system enables quantitative prediction of congestion evolution. Through this predictive capability, the system can identify which nodes are likely to form continuous congestion chains in a short period of time, providing data support for subsequent control measures. This allows for intervention strategies to be implemented before congestion develops, effectively reducing the risk of congestion propagation. Furthermore, by understanding the contribution of different vehicle types to the pressure increase of the passage and the duration of congestion evolution, the system can adjust vehicle distribution, traffic weights, or activate internal diversion mechanisms. This not only alleviates the pressure on current congested nodes but also optimizes traffic flow distribution in advance, making the vehicle combination within the parking garage more reasonable, improving overall traffic efficiency, reducing the risk of systemic congestion caused by localized traffic occupying passages, and enhancing the safety and stability of parking lot operations.

[0036] It should be noted that the different vehicles are specifically: 1. Standard vehicles (ordinary passenger cars), small cars, and compact SUVs. These vehicles usually account for the largest proportion, have a small turning radius and occupy a small road area, and are the main reference objects for the baseline traffic flow model. 2. Large vehicles (enlarged size category), large SUVs, commercial vehicles (MPV), pickup trucks, and extended models. These types of vehicles take longer to pass through narrow passages, sharp bends, and slopes, which can easily cause localized and suppressive congestion. 3. Vehicles with special operating behaviors (flow velocity disturbance type): vehicles driven by novice drivers, taxis or ride-hailing vehicles that frequently stop, electric vehicles looking for charging stations, and vehicles that may be reversing or frequently hesitating. These vehicles are not necessarily large in size, but their behavior will significantly disturb the local flow velocity, thereby producing a non-size-based pressure boosting effect on the node pressure.

[0037] It should be added that the pressurization information of different vehicles on the garage passage is obtained, and based on the pressurization information, the time required for the garage passage to change from single-point congestion to patchy congestion is generated, specifically as follows: First, based on the vehicle type information detected in the garage aisle, the system obtains the pressurization information for each vehicle type in the aisle; the pressurization information specifically includes: Size-related lane encroachment: Large or special vehicles occupy lane widths exceeding those of standard vehicles, resulting in a reduction in the actual usable space of the passageway; Traffic speed differences: Different types of vehicles have different speeds, and large vehicles or slow-moving vehicles may slow down the overall traffic flow. Differences in operational behavior: Some vehicles may engage in behaviors such as parking, reversing, sudden braking, or detouring, which may reduce the instantaneous throughput of the passage. The system comprehensively quantifies these boosting factors and generates the pressure contribution of each vehicle to the channel, providing basic data for subsequent congestion evolution analysis; II. Based on boost information, the system generates the time required for congestion to spread from a single point to a patchy congestion. After acquiring boost information, the system simulates or calculates the evolution of vehicle flow within the lane to determine the time required for congestion at a single node to spread to surrounding nodes and form a patchy congestion. Specific methods include: The channel is divided into several nodes, and the instantaneous throughput capacity of each node (affected by boost information) is calculated. The speed at which congestion spreads from a single point (single node) to adjacent nodes is predicted, and the total time required to form a patchy congestion is accumulated. Through this calculation, the system can quantify the speed of congestion spread in advance, providing a basis for decision-making for dynamic diversion, flow restriction and path guidance. III. Examples of Scenarios Hypothetical scenario: The following vehicles are simultaneously present on main road A of the three-level underground parking garage: 10 standard vehicles, each with a boost pressure of 1. Three large SUVs, each with a boost pressure of 3. Two special vehicles (such as trucks), each with a boost pressure of 4. The system calculates based on vehicle type and boost information: Single-point congestion (node ​​P1) formation time: 3 minutes. The total time required for this node to spread to adjacent nodes P2 and P3, forming a patchy congestion, is 7 minutes. The rapid spread was mainly due to large SUVs and special vehicles significantly encroaching on the road and traveling at speeds below average. Example of system response measures: By adjusting the traffic weights in advance at node P1 and its neighboring nodes, reducing the influx of new vehicles, and activating the internal diversion mechanism to redirect some vehicles to alternative lanes, the system provides optimized routes for specific vehicle types (such as large vehicles) to avoid slowing down traffic flow on the main road. In this way, the system can proactively intervene before patchy congestion forms, alleviate lane pressure, and optimize the overall traffic flow distribution.

[0038] In this embodiment, step S1, which generates traffic flow data corresponding to the vehicle information in real time based on the vehicle information pre-recorded at the parking lot entrance and exit, further includes: S11: Based on the parking terminal's preset entrance and exit points for the parking lot entrance and exit, identify the corresponding parking space area when a vehicle passes through the entrance and exit point; S12: Determine whether the number of vehicles parked in the parking spaces of the area has reached the preset parking saturation value; S13: If so, the parking terminal guides the vehicles to park and divert traffic. Based on the parking diversion, the total number of vehicles inside the parking lot is dynamically updated. Specifically, the parking diversion includes guiding vehicles to turn around and re-enter the parking lot from other entrances and exits, guiding vehicles from the current parking space to other parking spaces, and guiding vehicles to leave the parking lot.

[0039] In this embodiment, the system identifies the corresponding parking spaces when a vehicle passes through a pre-set entrance / exit point at the parking lot terminal. The system then determines whether the number of vehicles parked in these areas has reached a pre-set parking saturation value, and executes the corresponding steps accordingly. For example, if the system determines that the number of vehicles parked in the corresponding area when a vehicle passes through an entrance / exit point has not reached the pre-set parking saturation value, the system considers that there are still available parking spaces in that area, the parking resources are not fully utilized, the local parking pressure is low, and the passageway and parking space allocation are still in normal operating condition. The system will continue to update the information for that area in real time. The system monitors the number of parked vehicles in a given area, maintaining early warning capabilities to detect impending saturation. It also provides optimal parking routes and space navigation to improve the efficiency of finding parking spaces and reduce unnecessary driving within the garage. For example, when the system determines that the number of parked vehicles in the corresponding area has reached a pre-set saturation value when a vehicle passes through an entrance / exit, the system considers parking spaces in that area scarce. The system will then guide the vehicle to alternative parking areas via the parking terminal. This redirection includes guiding the vehicle to turn around and re-enter the parking lot from other entrances / exits, guiding the vehicle from its current parking space to another area, and... The system guides vehicles out of the parking lot and dynamically updates the total number of vehicles inside the parking lot based on different parking diversion measures. When the system determines that the parking spaces in a certain area have reached a pre-set saturation value, it indicates that parking resources in that area are scarce. Through parking diversion measures, the system can promptly guide new vehicles from high-pressure areas to other parking spaces or entrances / exits, avoiding vehicles blindly waiting or repeatedly searching for parking spaces in congested areas, thereby alleviating local parking pressure and preventing local bottlenecks. At the same time, by guiding vehicles from the current area to other parking spaces or re-entering the parking lot from other entrances / exits, the system can achieve a balanced distribution of vehicle flow and parking spaces. By balancing the allocation of parking spaces and making full use of the remaining capacity in each area of ​​the parking lot, this dynamic control method ensures that parking resources are used efficiently, reduces the situation where some areas are over-occupied while others are vacant, and improves the overall utilization rate of the parking lot. In addition, when parking spaces are scarce, the system provides a variety of diversion strategies (turning around to re-enter, going to other areas of parking space, or leaving the parking lot), which can effectively reduce the waiting and detouring time of vehicles in the garage. The system also dynamically updates the total number of vehicles in the parking lot, providing real-time data support for subsequent vehicle entry and internal control. This not only ensures smooth traffic flow but also improves the convenience and satisfaction of users during the parking process.

[0040] It should be noted that allowing vehicles to leave the parking lot and re-enter through other entrances / exits is because there may be congestion inside the parking lot at this time, and re-entering through other entrances / exits can avoid congestion; moving from the current parking area to another parking area is because there is no congestion inside the parking lot at this time, and going directly there is faster; and leaving the parking lot is because the current parking lot is very congested, so vehicles are being guided to other parking lots outside.

[0041] Reference Appendix Figure 2 A dynamic traffic flow pressure prediction system for parking lots, as described in one embodiment of the present invention, includes: The generation module 10 is used to generate traffic flow data corresponding to the vehicle information in real time based on the vehicle information pre-entered at the parking lot entrance and exit. The judgment module 20 is used to determine whether the traffic flow data has reached a preset traffic flow threshold; The execution module 30 is used to, if so, construct a traffic flow network map inside the parking lot based on the internal traffic structure preset by the parking lot terminal, dynamically activate the channel sensing device of the parking lot channel according to the channel basic attributes preset by the parking lot terminal for the parking lot channel, and collect the pressure index of the parking lot channel through the channel sensing device. The internal traffic structure specifically includes turning nodes, pillar positions and ramp entrances, the channel basic attributes specifically include width, speed limit and capacity, and the pressure index specifically includes local density, congestion index, throughput and queue growth rate. The second judgment module 40 is used to determine whether the growth rate of the pressure index exceeds a preset rate threshold. The second execution module 50 is used to, if the growth rate is exceeded, obtain the congested nodes inside the parking lot, identify the adjacent areas of the congested nodes, dynamically adjust the diversion strategy of the garage passage according to the adjacent areas, and restrict the vehicle parking rights of the congested nodes within a preset range according to the diversion strategy. The adjacent areas specifically include entrance ramps, exit convergence points and main roads, and the diversion strategy specifically includes internal diversion and external flow restriction.

[0042] In this embodiment, the generation module 10 generates real-time traffic flow data corresponding to the vehicles entering or leaving the parking lot based on the pre-entered vehicle information at the parking lot entrances and exits. Then, the judgment module 20 determines whether this traffic flow data reaches a pre-set traffic flow threshold to execute corresponding steps. For example, if the system determines that the parking lot's traffic flow data has not reached the pre-set threshold, the system considers the overall traffic flow in the parking lot to be acceptable and manageable, without potential overload or congestion risks. The system maintains a low-resource-occupancy monitoring state, only performing routine statistics on vehicle information at the entrances and exits without activating the channel-level pressure acquisition equipment. Simultaneously, the channels do not trigger dynamic interventions such as diversion, flow restriction, or bidirectional / unidirectional switching. Vehicles flow autonomously along default routes, and historical operational data under normal flow conditions is continuously accumulated for future use. The system identifies abnormal fluctuations. For example, when the system determines that the traffic flow data of the parking lot has reached the preset traffic flow threshold, the execution module 30 will consider that the overall traffic flow of the parking lot is relatively saturated and traffic pressure may occur. The system will construct a traffic flow network map inside the parking lot based on the internal traffic structure preset by the parking lot terminal, which specifically includes turning nodes, pillar positions and ramp entrances. Based on the basic channel attributes preset by the parking lot terminal for the garage passage, which specifically include width, speed limit and capacity, the system will dynamically activate the channel sensing devices of the garage passage. The channel sensing devices specifically include vehicle detection geomagnetic sensors, laser rangefinders, millimeter-wave radar and video recognition cameras. Through these channel sensing devices, the system collects pressure indicators of the garage passage, which specifically include local density, congestion index, throughput and queue growth rate.The system constructs an internal traffic flow network model using internal traffic structure data (turning points, pillars, ramp entrances), transforming macroscopic statistical data into localized structured predictive capabilities. This process upgrades the system from a weak monitoring mode that only monitors entrance traffic flow to a strong monitoring mode that tracks the actual flow of traffic throughout the entire parking garage. This enables early detection of congestion risks, rather than reactive responses. Simultaneously, it dynamically activates different types of passage sensing devices based on basic passage attributes (width, speed limit, capacity), resulting in more accurate and energy-efficient sensing. For example, millimeter-wave radar is prioritized for narrow sections, video detection nodes are activated at ramp entrances, while geomagnetic sensors can be used for wide, straight passages. This "on-demand activation" approach... This approach improves the authenticity and completeness of sensor data while reducing redundant equipment and enhancing system energy efficiency. The resulting traffic flow data more accurately reflects the degree of node constraint. Furthermore, the system collects pressure indicators including local density, congestion index, throughput, and queue growth rate. These indicators are no longer based on manual judgment or guesswork from entry data, but rather on real-time node data measurements. This allows the system to form a quantifiable local pressure field representation, providing a reliable basis for subsequent decisions such as "whether diversion is needed," "whether traffic restrictions are needed," "which area has the greatest pressure," and "whether secondary congestion is likely to occur." The second judgment module then... 40. Determine whether the growth rate of the pressure index in a parking garage aisle exceeds a pre-set threshold to execute corresponding steps. For example, if the system determines that the growth rate of the pressure index in a parking garage aisle does not exceed the pre-set threshold, the system will consider the local density or congestion index of the parking lot to be at a moderate level, but the growth rate is slow and there is no short-term sharp backlog. This is considered "stable or slowly accumulating pressure." The system will continue to maintain the normal monitoring and sampling frequency without additional intervention on vehicle parking permissions. At the same time, it will record the current pressure index as a trend reference, focusing on observing the traffic flow growth trend rather than immediate intervention, and maintaining a low-energy observation mode. The system continues to wait for the next cycle of data sampling and judgment. For example, when the system determines that the growth rate of the pressure index of a certain garage passage exceeds the preset rate threshold, the second execution module 50 will consider that the parking lot congestion has caused a short-term and rapid backlog. Based on different growth rates, the system will obtain the congestion nodes inside the parking lot and identify the adjacent areas of these congestion nodes. The adjacent areas specifically include the entrance ramp, the exit convergence point, and the main road. According to different adjacent areas, the system will dynamically adjust the diversion strategy of the garage passage. The diversion strategy specifically includes internal diversion and external flow restriction. Based on different diversion strategies, the system will restrict the vehicle parking rights of the congestion nodes within the preset range.The system identifies early signs of "short-term rapid congestion" by judging the growth rate of pressure indicators rather than simply the current pressure value. This refers to sudden congestion caused by the instantaneous arrival volume exceeding the traffic capacity. Based on different growth rates, the system accurately locates congestion nodes, enabling timely intervention before large-scale congestion spreads. This avoids the lag problem of traditional parking lots that "only deal with congestion after it has already formed," significantly improving the speed and sensitivity of internal traffic response. Furthermore, by identifying the area to which the congestion node belongs (entrance ramp, exit convergence point, main road), the system can dynamically select the corresponding diversion strategy according to the area type. For example, entrance ramps trigger external flow control, main roads trigger internal detours, and exit convergence points are prioritized for diversion. By optimizing queuing formations, this regionalized and scenario-based traffic diversion approach avoids secondary congestion problems caused by a "one-size-fits-all" strategy. It makes intervention measures more targeted and efficient, accurately alleviating pressure at the source of congestion. Furthermore, by dynamically restricting parking permissions in designated areas based on different diversion strategies (internal diversion and external flow restriction), it directly reduces the number of vehicles entering congested nodes, weakening the pressure growth trend at the source. This "proactive pressure reduction" method allows traffic flow to redistribute within the network, forming a more balanced pressure field and ensuring that key nodes are not overloaded. Compared to traditional methods of simple prompts or manual intervention, this system can automatically execute and maintain a stable traffic flow structure, improving the overall operational efficiency and safety of the parking lot.

[0043] In this embodiment, it also includes: The acquisition module is used to acquire the access area of ​​a vehicle after it enters the parking lot through the parking lot entrance / exit, based on the vehicle traffic lanes pre-divided by the internal traffic structure. The third judgment module is used to determine whether the access area intersects with the vehicle traffic lane; The third execution module is used to collect the frequency of the vehicles' intersections in the vehicle traffic lanes if the condition is met, dynamically adjust the traffic patterns of the vehicle traffic lanes based on the intersection frequencies, and generate path guidance information for the vehicles within the parking lot based on the traffic patterns. The traffic patterns specifically include one-way traffic, temporary closures, and detours.

[0044] In this embodiment, the system, based on pre-defined vehicle traffic lanes within the internal traffic structure, acquires the access areas after a vehicle passes through a parking lot entrance / exit and enters the parking lot. The system then determines whether these access areas intersect with vehicle traffic lanes to execute corresponding steps. For example, if the system determines that a vehicle's access areas after entering the parking lot do not intersect with vehicle traffic lanes, the system assumes that the vehicle's target route will not compete with or conflict with other major traffic flows. The system then reduces the real-time sampling frequency for that vehicle, meaning the monitoring index by cameras and radar decreases. Triggering additional stress analysis in this area, system resources are concentrated on other high-risk nodes, while the vehicle's navigation path is optimized, recommending shorter routes to facilitate faster parking. For example, when the system determines that a vehicle will converge on a traffic lane after entering the parking lot from the entrance / exit, the system considers this vehicle's target route to be competing with other major traffic flows. The system collects the vehicle's convergence frequency on the traffic lanes and dynamically adjusts the traffic patterns of these traffic lanes based on different convergence frequencies. The traffic patterns specifically include one-way traffic. The system employs various traffic management techniques, including temporary closures and detours, to generate route guidance information for vehicles within the parking lot, based on different traffic patterns. By identifying whether a vehicle's path passes through traffic lanes, the system can preemptively determine if a vehicle is a potential "conflict participant." It collects vehicle convergence frequencies and dynamically determines traffic patterns based on the merging characteristics between nodes, ensuring vehicles receive clear path assignments before entering the convergence area. This reduces random congestion caused by behaviors such as cutting in, vying for lanes, and yielding. Furthermore, it adjusts the traffic patterns of traffic lanes based on different vehicle convergence frequencies; for example, when the convergence frequency is too high... The system employs temporary one-way traffic, partially closing off areas or initiating detour guidance when traffic pressure is high in a particular direction. This flexible and variable traffic rule allows the passageway to operate in a flexible, adaptable manner, rather than being fixed. Based on the temporarily adjusted traffic patterns, the system generates real-time route guidance information for each vehicle, enabling vehicles to quickly select the optimal route in the complex environment of the parking garage, rather than relying on the driver's judgment. Drivers do not need to consider whether they encounter relative traffic flow, whether they need to yield, or whether they may enter a congested section, thus significantly reducing decision-making delays and the probability of incorrect route selection.

[0045] In this embodiment, the execution module further includes: The data acquisition unit is used to collect the sensing spread efficiency of the garage passage from one node to another based on the congestion spread radius preset by the parking lot terminal. The judgment unit is used to determine whether the sensing spread efficiency reaches a preset efficiency threshold. The execution unit is configured to, if so, construct the node traffic flow offset of the garage passage within a preset time period based on the target content pre-collected by the parking terminal, and dynamically adjust the entry and exit weight of the garage passage at the parking terminal based on the node traffic flow offset and the passage geometry of the garage passage. The target content specifically includes the exit direction, parking blocks and charging pile distribution, and the passage geometry specifically includes narrow passages, blind corners, sharp bends and ramps.

[0046] In this embodiment, the system collects the congestion spread efficiency from one node to another in the parking garage passage based on the pre-set congestion spread radius of the parking garage terminal. The system then determines whether this spread efficiency reaches a pre-set efficiency threshold and executes corresponding steps accordingly. For example, if the system determines that the sensing spread efficiency from one node to another in the parking garage passage has not reached the pre-set efficiency threshold, the system considers the current congestion to be localized, and its impact range will not spontaneously expand. The system will mark this congestion as "locally controllable congestion," maintaining local monitoring of the area, not triggering cross-regional diversion strategies, while maintaining normal traffic flow patterns at the nodes and keeping existing traffic flow guidance unchanged. There is no need to implement one-way flow restriction or path blocking, and the risk of congestion is reduced. The data processing priority in this area is to monitor trends only within necessary limits. For example, when the system determines that the sensing spread efficiency of the parking garage aisle from one node to another has reached a pre-set efficiency threshold, the system considers the current congestion to be global and its impact range to be spontaneously expanding. The system will then construct the node traffic flow offset of the parking garage aisle within a pre-set time period based on the pre-collected entry target information from the parking terminal, specifically including exit direction, parking area, and charging pile distribution. Based on different node traffic flow offsets and the aisle geometry (including narrow passages, blind spots, sharp bends, and ramps), the system will dynamically adjust the entry and exit weights of the parking garage aisle at the parking terminal. When the spread efficiency reaches the threshold... After the system detects the congestion, it determines that the congestion is no longer a localized, isolated phenomenon, but rather a global congestion with a spreading trend. This means that the congestion pressure can be transmitted between multiple nodes in the parking garage, forming a chain reaction of congestion. By recognizing this spreading signal, the system upgrades its scheduling strategy from local intervention to global governance, enabling rapid blocking and buffering before the congestion fully develops. This prevents a blockage at one node from paralyzing traffic on the entire parking garage's main road or multi-level structure. Simultaneously, based on the vehicle's entry target (exit direction, parking area, charging station distribution), the system constructs traffic flow deviations for different nodes in the future time period. This means the system not only focuses on the vehicle's current location but also predicts in advance the area the vehicle will eventually move to, thus enabling proactive control. Predictive scheduling significantly enhances the system's proactive ability to handle congestion. Instead of passively responding to congestion, the system proactively restructures traffic flow, reducing the spread of congestion at its source and resulting in a more balanced traffic distribution. By considering the geometric characteristics of different lanes (narrow lanes, blind spots, sharp bends, ramps), the system can dynamically adjust the entry and exit weights of each lane. For example, narrow lanes are appropriately downweighted during congestion spread, while main roads with open views have higher entry and exit priorities. In this way, the system can guide traffic flow to routes with higher capacity and lower risk, making the fluid traffic flow structure conform to physical traffic characteristics, reducing the probability of "structural congestion," and ultimately improving overall traffic efficiency. This makes traffic flow scheduling more in line with actual lane characteristics and road safety requirements.

[0047] In this embodiment, the second execution module further includes: The identification unit is used to identify the vehicle flow rate reduction rate of the congestion node based on the vehicle passage clearance pre-detected by the parking terminal of the parking lot terminal. The second judgment unit is used to determine whether the vehicle flow rate decrease rate continues to decrease. The second execution unit is used to, if not, obtain the vehicle congestion type of the congestion node, guide the vehicle queue length of the congestion node according to the vehicle congestion type, and dynamically introduce the internal diversion mechanism preset by the parking terminal according to the vehicle queue length. The vehicle congestion type specifically includes flow-type congestion, behavioral congestion and structural congestion, and the internal diversion mechanism specifically includes volume restriction, speed restriction and target restriction.

[0048] In this embodiment, the system, based on the pre-detected vehicle throughput capacity of the parking garage lanes by the parking terminal, identifies the vehicle speed reduction rate at congested nodes. The system then determines whether this vehicle speed reduction rate is continuously decreasing and executes corresponding steps accordingly. For example, when the system determines that the vehicle speed reduction rate at a congested node is continuously decreasing, it considers the congestion level to be gradually decreasing. Instead of maintaining strict flow control measures, the system gradually restores the throughput weight of the lane corresponding to the congested node and appropriately reduces external flow restriction or detour strategies for that area. Simultaneously, the system maintains a low-level monitoring mode to observe the vehicle speed recovery and retains only a mild, flexible diversion strategy to allow traffic flow to... Without excessive intervention, the system naturally restores the balance of traffic flow in the affected area, improving overall traffic efficiency and avoiding the traffic disturbance effects caused by excessive scheduling. For example, if the system determines that the vehicle speed reduction rate at a congested node has not continued to decrease, it considers the congestion level to be still severe. The system will then obtain the type of vehicle congestion at the congested node, which specifically includes flow-related congestion, behavioral congestion, and structural congestion. Based on different vehicle congestion types, the system guides the length of the vehicle queue at the congested node. Based on different queue lengths, it dynamically introduces a pre-set internal diversion mechanism from the parking lot terminal. This internal diversion mechanism specifically includes volume limits, speed limits, and target limits. The system further manages the congested nodes... Traffic congestion type identification can determine whether the current congestion is due to "flow-based congestion" caused by excessive traffic volume, "behavioral congestion" caused by individual vehicles making slow decisions, temporarily stopping, or waiting to react, or "structural congestion" caused by bottlenecks in the channel geometry. This congestion type differentiation allows the system to adopt different response strategies based on the specific causes, improving scheduling efficiency, avoiding the use of a single, crude flow-limiting method, reducing misjudgments and excessive intervention, and dynamically performing queue compression, queue transfer, or queue segmentation based on the length of the vehicle queue at the congested node. This effectively reduces the fixed length of the congestion queue and reduces the pressure accumulation caused by vehicles stagnating due to long waiting times. This orderly queuing... Queue guidance can prevent vehicles from randomly cutting in, vying for lanes, hesitating, and making blind choices in congested areas, turning a chaotic state into a controllable one, thereby improving traffic flow and promoting the recovery of local traffic flow. Furthermore, once the congestion type and queue length are determined, the system introduces three types of internal diversion mechanisms: volume restriction (controlling the entry of large vehicles), speed restriction (limiting the impact of cutting in and sudden stops), and target restriction (re-planning parking target areas). This allows traffic flow to be redistributed to more suitable channels or target areas, thereby achieving global traffic optimization. This dynamic diversion can minimize the structural pressure in local areas, allowing traffic flow to form a flexible distribution within the parking lot, improving overall throughput and capacity efficiency.

[0049] In this embodiment, the determination module further includes: The second identification unit is used to identify the spacing information between a single vehicle and another vehicle based on the number of vehicles pre-identified in the garage passage. The third judgment unit is used to determine whether the spacing information reaches a preset spacing threshold. The third execution unit is used to, if not, detect the average speed of vehicles in the garage passage based on the traffic flow data, and identify abnormal vehicle behavior in the garage passage based on the average speed, wherein the abnormal vehicle behavior specifically includes abnormal driving, abnormal path and abnormal vehicle characteristics.

[0050] In this embodiment, the system identifies the distance information between a single vehicle and another vehicle based on the pre-identified number of vehicles in the garage passage. The system then determines whether this distance information meets a pre-set distance threshold to execute corresponding steps. For example, when the system determines that the distance information between a vehicle and another vehicle meets the pre-set distance threshold, the system considers the vehicle distance to meet safe passage requirements, meaning that the vehicles maintain a controllable and safe distance to avoid collisions or brake shock transmission due to excessive closeness. The system will not restrict or close the passage or related nodes, maintaining the current passage priority and passage access rights. The system emphasizes real-time monitoring of vehicle spacing, but can appropriately reduce the sampling frequency to avoid excessive intervention that could cause unnecessary detours or deceleration. It also provides optimal route suggestions to ensure stable traffic flow in the garage. If a vehicle's target area is far away, it can suggest a reasonable route to avoid potential future congestion. For example, if the system determines that the distance between two vehicles does not meet a pre-set threshold, it considers the distance between vehicles too close, possibly due to excessive traffic congestion in the garage aisle. The system will then detect the average speed of vehicles in the garage aisle based on different traffic flow data. These average speeds identify abnormal vehicle behavior within the parking garage aisles, specifically including abnormal driving, abnormal routing, and abnormal vehicle characteristics. By monitoring traffic flow and vehicle spacing in real time, the system can detect potential congestion points in advance, allowing for intervention before collisions or sudden braking occur. This early warning mechanism helps reduce the probability of accidents and ensures safe vehicle passage within the parking lot. Furthermore, upon detecting insufficient vehicle spacing, the system calculates the average vehicle speed within the parking garage aisles based on traffic flow data, identifying abnormal vehicle behavior, including abnormal driving (such as sudden acceleration or braking) and abnormal routing (such as arbitrary lane changes or detours). 1. Identifying abnormal vehicle characteristics (such as excessively large vehicles affecting traffic flow): By recognizing such abnormal behavior, the system can take intervention measures for specific vehicles, reducing the contagion effect of local congestion on the overall traffic flow, improving the efficiency of the passage. After identifying abnormal vehicle behavior, the system can take dynamic adjustment measures based on the characteristics of congested nodes and passages, such as adjusting traffic weights, implementing internal diversion or route guidance. Through precise control of abnormal vehicles and congested areas, the system can not only alleviate the current congestion, but also optimize the traffic flow distribution within the entire parking lot, making vehicle traffic more balanced and smooth, thereby improving the overall operational efficiency of the parking lot and the user experience.

[0051] In this embodiment, the second determination module further includes: The second acquisition unit is used to acquire the pressure contribution value of different vehicles to the garage passage based on the pre-detected vehicle types in the garage passage, wherein the vehicle types specifically include standard vehicles, large vehicles and special vehicles; The fourth judgment unit is used to determine whether the pressure contribution values ​​are balanced; The fourth execution unit is used to, if not, obtain the pressure boosting information of different vehicles on the garage passage according to the vehicle type, and generate the time required for the garage passage to change from single-point congestion to patchy congestion based on the pressure boosting information. Specifically, the pressure boosting information refers to the decrease in the passage capacity of the garage passage caused by differences in the size of different vehicles occupying the lane, their speed, and their operating behavior.

[0052] In this embodiment, the system collects the pressure contribution values ​​of different vehicle types pre-detected within the garage aisle, specifically including standard vehicles, large vehicles, and special vehicles. The system then determines whether these pressure contribution values ​​are balanced within the same garage aisle to execute corresponding steps. For example, if the system determines that the pressure contribution values ​​of different vehicles within the same garage aisle are balanced, it considers that the vehicle combination in that aisle does not generate significant unbalanced pressure under the current traffic conditions, the traffic flow is stable, and the aisle capacity is not occupied or compressed by a single vehicle type. The system maintains the originally assigned entry and exit weights and does not restrict specific vehicle types from entering the aisle, while continuing... The system collects vehicle type and pressure contribution values, but can reduce intervention priority, focusing on monitoring potentially new vehicle types entering the garage to prevent future imbalances. It also continues to provide navigation guidance to ensure vehicles travel along optimal paths, improving overall traffic efficiency. For example, when the system determines that the pressure contribution values ​​of different vehicles to the same garage aisle are unbalanced, it considers the traffic flow abnormal under current conditions, with the aisle's capacity easily occupied by a single vehicle type. The system will then acquire pressure boosting information for different vehicle types, specifically information on how vehicle size, speed, and operational behavior cause a decrease in the garage aisle's capacity. Based on this pressure boosting information, the system generates the time required for a garage lane to evolve from single-point congestion to widespread congestion. When the system determines that the pressure contribution values ​​of different vehicles to the same garage lane are unbalanced, it indicates that the vehicle composition in that lane is unbalanced under current traffic conditions. This may result in large or special vehicles occupying the lane and blocking the passage of standard vehicles. By acquiring pressure boosting information for different vehicle types (size, lane occupation, speed, and operational behavior differences), the system can accurately identify potential congestion nodes and high-risk lanes, detect abnormal traffic flow conditions in advance, and prevent local congestion from spreading into global congestion. Furthermore, based on the pressure boosting information of different vehicles, the system generates the time required for a garage lane to develop from single-point congestion to widespread congestion. The system enables quantitative prediction of congestion evolution. Through this predictive capability, the system can identify which nodes are likely to form continuous congestion chains in a short period of time, providing data support for subsequent control measures. This allows for intervention strategies to be implemented before congestion develops, effectively reducing the risk of congestion propagation. Furthermore, by understanding the contribution of different vehicle types to the pressure increase of the passage and the duration of congestion evolution, the system can adjust vehicle distribution, traffic weights, or activate internal diversion mechanisms. This not only alleviates the pressure on current congested nodes but also optimizes traffic flow distribution in advance, making the vehicle combination within the parking garage more reasonable, improving overall traffic efficiency, reducing the risk of systemic congestion caused by localized traffic occupying passages, and enhancing the safety and stability of parking lot operations.

[0053] In this embodiment, the generation module further includes: The third identification unit is used to identify the corresponding parking space when a vehicle passes through the entrance / exit of the parking lot based on the preset entrance / exit points of the parking lot terminal. The fifth judgment unit is used to determine whether the number of vehicles parked in the parking spaces of the area has reached a preset parking saturation value; The fifth execution unit is used to guide the vehicles to park through the parking terminal if the condition is met, and dynamically update the total number of vehicles inside the parking lot based on the parking guidance. The parking guidance specifically includes guiding vehicles to turn around and re-enter the parking lot from other entrances and exits, guiding vehicles from the current parking space to other parking spaces, and guiding vehicles to leave the parking lot.

[0054] In this embodiment, the system identifies the corresponding parking spaces when a vehicle passes through a pre-set entrance / exit point at the parking lot terminal. The system then determines whether the number of vehicles parked in these areas has reached a pre-set parking saturation value, and executes the corresponding steps accordingly. For example, if the system determines that the number of vehicles parked in the corresponding area when a vehicle passes through an entrance / exit point has not reached the pre-set parking saturation value, the system considers that there are still available parking spaces in that area, the parking resources are not fully utilized, the local parking pressure is low, and the passageway and parking space allocation are still in normal operating condition. The system will continue to update the information for that area in real time. The system monitors the number of parked vehicles in a given area, maintaining early warning capabilities to detect impending saturation. It also provides optimal parking routes and space navigation to improve the efficiency of finding parking spaces and reduce unnecessary driving within the garage. For example, when the system determines that the number of parked vehicles in the corresponding area has reached a pre-set saturation value when a vehicle passes through an entrance / exit, the system considers parking spaces in that area scarce. The system will then guide the vehicle to alternative parking areas via the parking terminal. This redirection includes guiding the vehicle to turn around and re-enter the parking lot from other entrances / exits, guiding the vehicle from its current parking space to another area, and... The system guides vehicles out of the parking lot and dynamically updates the total number of vehicles inside the parking lot based on different parking diversion measures. When the system determines that the parking spaces in a certain area have reached a pre-set saturation value, it indicates that parking resources in that area are scarce. Through parking diversion measures, the system can promptly guide new vehicles from high-pressure areas to other parking spaces or entrances / exits, avoiding vehicles blindly waiting or repeatedly searching for parking spaces in congested areas, thereby alleviating local parking pressure and preventing local bottlenecks. At the same time, by guiding vehicles from the current area to other parking spaces or re-entering the parking lot from other entrances / exits, the system can achieve a balanced distribution of vehicle flow and parking spaces. By balancing the allocation of parking spaces and making full use of the remaining capacity in each area of ​​the parking lot, this dynamic control method ensures that parking resources are used efficiently, reduces the situation where some areas are over-occupied while others are vacant, and improves the overall utilization rate of the parking lot. In addition, when parking spaces are scarce, the system provides a variety of diversion strategies (turning around to re-enter, going to other areas of parking space, or leaving the parking lot), which can effectively reduce the waiting and detouring time of vehicles in the garage. The system also dynamically updates the total number of vehicles in the parking lot, providing real-time data support for subsequent vehicle entry and internal control. This not only ensures smooth traffic flow but also improves the convenience and satisfaction of users during the parking process.

[0055] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art 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 appended claims and their equivalents.

Claims

1. A dynamic traffic pressure prediction method for a parking lot, characterized by, Includes the following steps: Based on the vehicle information pre-entered at the parking lot entrance and exit, traffic flow data corresponding to the vehicle information is generated in real time. Determine whether the traffic flow data has reached a preset traffic flow threshold; If so, then based on the internal traffic structure preset by the parking terminal, a traffic flow network map inside the parking lot is constructed. Based on the basic channel attributes preset by the parking terminal for the garage passage, the channel sensing device of the garage passage is dynamically activated. Through the channel sensing device, the pressure index of the garage passage is collected. The internal traffic structure specifically includes turning nodes, pillar positions and ramp entrances. The basic channel attributes specifically include width, speed limit and capacity. The pressure index specifically includes local density, congestion index, throughput and queue growth rate. Determine whether the growth rate of the pressure index exceeds a preset rate threshold; If the growth rate is exceeded, the congested nodes inside the parking lot are obtained based on the growth rate, the adjacent areas of the congested nodes are identified, and the diversion strategy of the garage passage is dynamically adjusted according to the adjacent areas. Based on the diversion strategy, the parking rights of the congested nodes within a preset range are restricted. The adjacent areas specifically include entrance ramps, exit convergence points and main roads, and the diversion strategy specifically includes internal diversion and external flow restriction.

2. The dynamic traffic pressure prediction method of a parking lot according to claim 1, wherein, After the step of constructing a traffic flow network map inside the parking lot based on the preset internal traffic structure of the parking lot terminal, the method further includes: Based on the pre-divided vehicle traffic lanes of the internal traffic structure, the access area after a vehicle enters the parking lot through the parking lot entrance / exit is obtained. Determine whether the access area intersects with the vehicle traffic lane; If so, the frequency of the vehicles meeting in the traffic lanes is collected, and the traffic patterns of the traffic lanes are dynamically adjusted according to the meeting frequency. Based on the traffic patterns, path guidance information for the vehicles inside the parking lot is generated, wherein the traffic patterns specifically include one-way traffic, temporary closure, and detour.

3. The dynamic traffic pressure prediction method of a parking lot according to claim 1, wherein, The step of dynamically activating the lane sensing device of the garage lane based on the preset basic attributes of the lane by the parking terminal further includes: Based on the pre-set congestion contagion radius of the parking lot terminal, the sensing spread efficiency of the garage passage from one node to another is collected. Determine whether the sensing spread efficiency reaches a preset efficiency threshold; If so, then based on the target content pre-collected by the parking terminal for vehicles, the node traffic flow offset of the garage passage within a preset time period is constructed. Based on the node traffic flow offset and the passage geometry of the garage passage, the entry and exit weight of the garage passage at the parking terminal is dynamically adjusted. The target content specifically includes the exit direction, parking blocks and charging pile distribution, and the passage geometry specifically includes narrow passages, blind corners, sharp bends and ramps.

4. The dynamic traffic pressure prediction method of a parking lot according to claim 1, wherein, The step of identifying the adjacent areas of the congested node and dynamically adjusting the traffic diversion strategy of the garage passage based on the adjacent areas further includes: Based on the vehicle passage clearance pre-detected by the parking terminal for the garage passage, the vehicle flow rate reduction rate of the congestion node is identified. Determine whether the vehicle flow rate decreases continuously; If not, then obtain the vehicle congestion type of the congestion node, guide the vehicle queue length of the congestion node according to the vehicle congestion type, and dynamically introduce the internal diversion mechanism preset by the parking terminal according to the vehicle queue length. The vehicle congestion type specifically includes flow-type congestion, behavior-type congestion and structure-type congestion, and the internal diversion mechanism specifically includes volume restriction, speed restriction and target restriction.

5. The method for predicting dynamic traffic flow pressure in a parking lot according to claim 1, characterized in that, The step of determining whether the traffic flow data has reached a preset traffic flow threshold further includes: Based on the number of vehicles pre-identified in the garage passage, the spacing information between a single vehicle and another vehicle is identified; Determine whether the spacing information reaches a preset spacing threshold; If not, then based on the traffic flow data, the average speed of vehicles in the garage passage is detected, and based on the average speed, abnormal vehicle behavior in the garage passage is identified, wherein the abnormal vehicle behavior specifically includes abnormal driving, abnormal path and abnormal vehicle characteristics.

6. The method for predicting dynamic traffic flow pressure in a parking lot according to claim 1, characterized in that, The step of determining whether the growth rate of the pressure index exceeds a preset rate threshold further includes: Based on the pre-detected vehicle types in the garage passage, the pressure contribution values ​​of different vehicles to the garage passage are collected, wherein the vehicle types specifically include standard vehicles, large vehicles and special vehicles; Determine whether the pressure contribution values ​​are balanced; If not, then based on the vehicle type, obtain the pressure boosting information of different vehicles on the garage passage, and based on the pressure boosting information, generate the time required for the garage passage to go from single-point congestion to patchy congestion. Specifically, the pressure boosting information refers to the decrease in the passage capacity of the garage passage caused by differences in the size of different vehicles occupying the lane, their speed, and their operating behavior.

7. The method for predicting dynamic traffic flow pressure in a parking lot according to claim 1, characterized in that, The step of generating traffic flow data corresponding to the vehicle information in real time based on the vehicle information pre-entered at the parking lot entrance and exit also includes: Based on the parking terminal's preset entrance and exit points for the parking lot entrance and exit, the corresponding area parking space is identified when a vehicle passes through the entrance and exit point; Determine whether the number of vehicles parked in the parking spaces of the area has reached a preset parking saturation value; If so, the parking terminal guides the vehicles for parking and diversion. Based on the parking diversion, the total number of vehicles inside the parking lot is dynamically updated. Specifically, the parking diversion includes guiding vehicles to turn around and re-enter the parking lot from other entrances and exits, guiding vehicles from the current parking space to other parking spaces, and guiding vehicles to leave the parking lot.

8. A dynamic traffic flow pressure prediction system for parking lots, characterized in that, include: The generation module is used to generate traffic flow data corresponding to the vehicle information pre-entered at the parking lot entrance and exit in real time. The judgment module is used to determine whether the traffic flow data has reached a preset traffic flow threshold; The execution module is used to, if so, construct a traffic flow network map inside the parking lot based on the internal traffic structure preset by the parking lot terminal, dynamically activate the channel sensing device of the parking lot channel according to the channel basic attributes preset by the parking lot terminal for the parking lot channel, and collect the pressure index of the parking lot channel through the channel sensing device. The internal traffic structure specifically includes turning nodes, pillar positions and ramp entrances, the channel basic attributes specifically include width, speed limit and capacity, and the pressure index specifically includes local density, congestion index, throughput and queue growth rate. The second judgment module is used to determine whether the growth rate of the pressure index exceeds a preset rate threshold. The second execution module is used to, if the growth rate is exceeded, obtain the congested nodes inside the parking lot, identify the adjacent areas of the congested nodes, dynamically adjust the diversion strategy of the garage passage according to the adjacent areas, and restrict the vehicle parking rights of the congested nodes within a preset range according to the diversion strategy. The adjacent areas specifically include entrance ramps, exit convergence points and main roads, and the diversion strategy specifically includes internal diversion and external flow restriction.

9. The dynamic traffic flow pressure prediction system for parking lots according to claim 8, characterized in that, Also includes: The acquisition module is used to acquire the access area of ​​a vehicle after it enters the parking lot through the parking lot entrance / exit, based on the vehicle traffic lanes pre-divided by the internal traffic structure. The third judgment module is used to determine whether the access area intersects with the vehicle traffic lane; The third execution module is used to collect the frequency of the vehicles' intersections in the vehicle traffic lanes if the condition is met, dynamically adjust the traffic patterns of the vehicle traffic lanes based on the intersection frequencies, and generate path guidance information for the vehicles within the parking lot based on the traffic patterns. The traffic patterns specifically include one-way traffic, temporary closures, and detours.

10. The dynamic traffic flow pressure prediction system for parking lots according to claim 8, characterized in that, The execution module further includes: The data acquisition unit is used to collect the sensing spread efficiency of the garage passage from one node to another based on the congestion spread radius preset by the parking lot terminal. The judgment unit is used to determine whether the sensing spread efficiency reaches a preset efficiency threshold. The execution unit is configured to, if so, construct the node traffic flow offset of the garage passage within a preset time period based on the target content pre-collected by the parking terminal, and dynamically adjust the entry and exit weight of the garage passage at the parking terminal based on the node traffic flow offset and the passage geometry of the garage passage. The target content specifically includes the exit direction, parking blocks and charging pile distribution, and the passage geometry specifically includes narrow passages, blind corners, sharp bends and ramps.

Images