A new energy automobile charging pile parking space state monitoring system

Abstract

The present application relates to the technical field of intelligent transportation and new energy vehicle supporting facilities, in particular to a new energy vehicle charging pile parking space state monitoring system, which comprises the following steps: multi-dimensional feature acquisition: a visual perception module acquires vehicle visual features and solves attributes, and an electrical monitoring module acquires equipment connection state signals; time sequence logic construction: in response to data changes, parking space time sequence data containing parking and connection time are generated; state spectrum solving: a central processing unit inputs attributes, signals and time sequence data into a parking space state analysis model, and calculates a real-time parking space state spectrum based on multi-dimensional logic closed loop; efficiency closed loop control: an efficiency control module generates traffic management instructions according to the state spectrum, and drives field devices to execute action strategies; the present application identifies inefficient states through a multi-dimensional coupling mechanism, and realizes the change from physical space occupation monitoring to value efficiency management.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of intelligent transportation and new energy vehicle supporting facilities, in particular to a new energy vehicle charging pile parking space state monitoring system. BACKGROUND

[0002] In the current new energy vehicle charging facility operation management environment, the parking space state monitoring system generally relies on visual recognition cameras or ground induction coils to perceive the physical existence of vehicles.

[0003] However, the existing solutions are mostly limited to single physical space occupation detection, and lack a mechanism for deeply coupling visual space features with electrical space load features. This single-dimensional monitoring logic cannot penetrate the physical appearance and cannot effectively distinguish between compliant charging behavior and inefficient occupation state, making it difficult to accurately identify behaviors such as illegal occupation of fuel vehicles, negative occupation of new energy vehicles without charging, and overtime parking after charging is completed. In addition, existing technologies typically use static thresholds for occupancy duration control, which cannot be dynamically adjusted based on vehicle battery capacity differences, charging power natural attenuation characteristics, and real-time congestion levels at the station, resulting in rigid resource allocation.

[0004] Therefore, how to accurately define and dynamically control the state of the parking space through multi-dimensional feature fusion to improve the turnover rate of charging resources and the accuracy of prediction and guidance services is a technical problem that needs to be solved. SUMMARY

[0005] To solve the above technical problems, the present application provides a new energy vehicle charging pile parking space state monitoring system. Specifically, the technical solution of the present application comprises:

[0006] A central processing unit, which is communicatively connected with a visual perception module, an electrical monitoring module, a timing logic module, and an efficiency control module;

[0007] The visual perception module is configured to collect vehicle visual feature data in the parking space area and calculate vehicle parking attribute information based on the vehicle visual feature data.

[0008] The electrical monitoring module is configured to collect device connection state signals on the side of the charging pile.

[0009] The timing logic module is configured to generate parking space timing data containing vehicle entry time, device connection time, and state duration information in response to changes in the vehicle visual feature data and the device connection state signals.

[0010] The central processing unit is pre-set with a parking space status analysis model, which is used to input the vehicle parking attribute information, the device connection status signal and the parking space time series data into the parking space status analysis model, calculate and output the real-time parking space status spectrum that represents the compliance and efficiency of the current parking space use;

[0011] The efficiency control module is used to generate traffic management instructions based on the real-time parking space status spectrum, and drive the on-site traffic indication equipment to execute corresponding action strategies based on the traffic management instructions.

[0012] Preferably, the process by which the visual perception module calculates vehicle parking attribute information includes:

[0013] Extract the license plate features and vehicle outline features from the vehicle visual feature data;

[0014] The vehicle's license plate characteristics are used to determine whether the vehicle possesses a valid parking space usage permit.

[0015] The vehicle's identity category information is obtained by matching the vehicle's outline features against a preset vehicle database.

[0016] Preferably, the process by which the central processing unit outputs the real-time parking space status spectrum includes:

[0017] Determine whether the vehicle possesses the aforementioned valid parking space use permit attributes;

[0018] If a vehicle does not have the legal parking space use permit attribute, and the duration of its entry into the parking space time series data is greater than the first preset time threshold, and the device connection status signal is disconnected, then the real-time parking space status spectrum will be marked as an illegal occupancy status.

[0019] If a vehicle possesses the legal parking space usage permit attribute, and the duration of its occupancy in the parking space time series data exceeds the second preset time threshold, and no effective device connection behavior occurs, then the real-time parking space status spectrum is marked as a passive occupancy state.

[0020] Preferably, the process of the central processing unit outputting the real-time parking space status spectrum further includes:

[0021] If the vehicle has the legal parking space use permit attribute and a valid device connection has occurred, then obtain the device connection duration from the parking space time sequence data.

[0022] A preset maximum allowed connection duration threshold is set. If the duration of the device connection exceeds the maximum allowed connection duration threshold, the real-time parking space status spectrum is marked as an overdue occupancy status.

[0023] Otherwise, the real-time parking space status spectrum will be marked as compliant occupancy.

[0024] Preferably, the process of the central processing unit outputting the real-time parking space status spectrum further includes:

[0025] Establish a differentiated connection duration policy based on vehicle identity category;

[0026] Based on the current vehicle's identity category information, the corresponding differentiated allowed connection duration is invoked to update the maximum allowed connection duration threshold.

[0027] Preferably, the process by which the efficiency control module generates traffic management instructions based on the real-time parking space status spectrum includes:

[0028] Extract the current state marker of the real-time parking space status spectrum;

[0029] If the current state is marked as an illegal occupancy state or a passive occupancy state, a parking space release prompt command or a parking space lock control command is generated.

[0030] If the current status is marked as an overdue occupancy status, a tiered rate billing instruction or a departure reminder instruction will be generated.

[0031] Preferably, the process by which the performance control module drives the on-site traffic indication equipment to execute action strategies includes:

[0032] Obtain the control interface of the status indicator unit set in the parking space area;

[0033] Establish a mapping relationship between the real-time parking space status spectrum and the displayed status;

[0034] In response to changes in the real-time parking space status spectrum, the status indicator unit is driven to switch to the corresponding display status through the control interface. The illegal occupancy status corresponds to the first warning display status, the compliant occupancy status corresponds to the second permitted display status, and the timed-out occupancy status corresponds to the third warning display status.

[0035] Preferably, the central processing unit is further configured to execute parking space usage prediction logic:

[0036] Based on historical parking space time series data, we analyze parking space usage patterns in different time periods;

[0037] By combining the real-time parking space status spectrum with the current time, the estimated occupancy time or the estimated release time of the parking space can be predicted.

[0038] The estimated occupancy time or estimated release time is sent to the data publishing port of the regional parking guidance system.

[0039] Compared with the prior art, the present invention has the following beneficial effects:

[0040] 1. This system introduces a parking space status analysis model and constructs a system architecture based on multi-dimensional feature coupling. By deeply coupling the physical features of the visual space, the load features of the electrical space, and the behavioral features of the time dimension, and using deterministic state machine logic, it can penetrate the physical appearance and accurately identify inefficient states where the vehicle is physically present but lacks value. This effectively solves the pain points of existing technologies where physical detection cannot distinguish between mistakenly entered vehicles and simple license plate recognition cannot determine the actual charging behavior. By setting a first preset time threshold and a second preset time threshold to construct a hierarchical filtering mechanism, it provides a reasonable fault tolerance buffer period for mistakenly entered vehicles and accurately combats the passive occupation behavior of parking without charging. This realizes the transformation of charging resource management from a single physical space occupancy monitoring to value efficiency management.

[0041] 2. This system achieves accurate physical reconstruction of vehicle identity by constructing a visual matching algorithm based on feature vector space; it generates a dense disparity map through a binocular stereo matching algorithm and solves the 3D point cloud by back-projection of the intrinsic parameter matrix, solving the black box problem of mapping from the 2D image domain to the 3D physical domain, and can accurately extract the physical dimensions of the vehicle (length, width, and height); by introducing the normalized Laplacian operator variance to calculate the texture complexity coefficient, and combining Z-Score normalization processing and Euclidean distance matching, it eliminates the dimensional differences of features in different dimensions, ensuring that the nominal battery capacity and power parameters of the vehicle can be accurately obtained even in the absence of electrical handshake data, providing definite physical parameter support for subsequent refined and timely management.

[0042] 3. This system achieves dynamic adaptive adjustment of the maximum allowable connection time threshold by establishing a capacity-congestion coupling model; by comprehensively considering vehicle battery capacity, current state of charge demand ratio, system bottleneck power, and charging curve equivalence coefficient, it can calculate a scientific theoretical charging time based on vehicle physical characteristics, avoiding misjudgments caused by one-size-fits-all management; by introducing a dynamic elastic buffer coefficient negatively correlated with the station congestion index, a linear congestion-duration feedback mechanism is established, which can automatically compress the allowable occupancy time slice under high station load conditions, realizing a dynamic control strategy for time-limited charging during congestion periods at the physical level, thereby maximizing service throughput under limited physical facility conditions;

[0043] 4. This system improves service transparency and guidance efficiency through dual-mode parking space usage prediction logic and closed-loop management strategies. By employing an electrochemical physical model for charging status and time series clustering analysis based on unit circle mapping for non-charging status, it solves the problem of traditional systems being unable to predict release times when faced with illegal parking. It can capture nighttime long-term occupancy patterns spanning midnight and output accurate predicted release times. By constructing a complete closed loop from monitoring to control, it drives on-site indication equipment to execute hierarchical action strategies, such as warning lights and ground lock control, based on the real-time parking space status spectrum. This not only creates a psychological deterrent to potential violators but also provides intuitive visual guidance for users looking for parking spaces, significantly improving the overall operational order of the parking lot. Attached Figure Description

[0044] The present invention will be further explained below with reference to the accompanying drawings and embodiments:

[0045] Figure 1 This is a structural diagram of the system of the present invention. Detailed Implementation

[0046] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to specific embodiments. Example 1:

[0047] Please see Figure 1 A new energy vehicle charging pile parking space status monitoring system includes a central processing unit, which is communicatively connected to a visual perception module, an electrical monitoring module, a timing logic module and an efficiency control module.

[0048] The visual perception module is used to collect visual feature data of vehicles in the parking space area and calculate vehicle parking attribute information based on the vehicle visual feature data.

[0049] The electrical monitoring module is used to collect equipment connection status signals on the charging pile side;

[0050] The timing logic module is used to generate parking space timing data that includes information on the vehicle's entry time, the device connection time, and the duration of the status, in response to changes in vehicle visual feature data and device connection status signals.

[0051] The central processing unit is pre-set with a parking space status analysis model, which is used to input vehicle parking attribute information, equipment connection status signals and parking space time series data into the parking space status analysis model, calculate and output the real-time parking space status spectrum that represents the compliance and efficiency of the current parking space use;

[0052] The efficiency control module is used to generate traffic management instructions based on the real-time parking space status spectrum, and drive the on-site traffic indication equipment to execute corresponding action strategies according to the traffic management instructions.

[0053] This embodiment details the system architecture based on multi-dimensional feature coupling; the visual perception module collects vehicle visual feature data within the parking space area through a binocular wide dynamic range camera deployed directly in front of the charging parking space and an embedded image processing chip. This data aims to construct an identity mapping for physical space; the electrical monitoring module is directly coupled to the CAN bus or PLC communication interface inside the charging pile to collect real-time device connection status signals on the charging pile side. To obtain the true state of energy interaction;

[0054] The timing logic module utilizes high-precision timers and event loggers to respond to... and The changes in these data generate parking space time-series data, which includes the time when a vehicle enters a parking space. Device connection time And the duration of the state measured in minutes; based on this, the central processing unit calls a preset parking space state analysis model, which is a deterministic state machine based on a multi-dimensional logical closed loop, and its state space is defined as:

[0055] ;

[0056] in, Indicates a waiting state. This indicates an unauthorized space allocation status. This indicates a passive, placeholder state. Indicates the charging status. Indicates a timeout status;

[0057] The input vector space is composed of Composition, among which, This indicates the vehicle's parking space usage permit attribute. This indicates the device connection status signal. Indicates the duration of the vehicle being parked. Indicates the duration of the device connection. hour The state transition logic is rigorously described mathematically: when ,in, When the first preset time threshold is used to define the tolerance time for illegal occupancy, the state... ;when ,in, When the second preset time threshold is used to define the determination time of negative occupancy, the state... ;when And connection duration ,in, When the maximum allowed connection duration threshold is reached, the state transitions to... ;against Regarding the subsequent evolution of the state, the system further defines the following transition logic: If in Connection duration monitored under status Exceeding the maximum allowed connection duration threshold Then a state transition is triggered. This incorporates timeout behavior into the closed-loop control of the state machine; at any time before the aforementioned trigger threshold is met, the model remains... That is, the observed state output; this model combines vehicle parking attribute information and signals. Using parking space time-series data as input vectors, calculate and output the real-time parking space state spectrum. ;

[0058] Performance management module based on Generate commands to drive the smart lock or guide screen to perform actions; among which... Derived from the output of the parking space status analysis model, its physical meaning is a comprehensive evaluation index characterizing the compliance and efficiency of the current parking space use;

[0059] This embodiment introduces a parking space status analysis model, which deeply couples the physical characteristics of the visual space, the load characteristics of the electrical space, and the behavioral characteristics of the time dimension. This multi-dimensional coupling mechanism can penetrate the physical appearance and identify inefficient states where the physical space is in place but the value is missing, such as a fuel vehicle occupying the space or not leaving after charging. This realizes the transformation of charging resource management from a single physical space occupancy monitoring to value efficiency management. Example 2:

[0060] The process by which the visual perception module calculates vehicle parking attribute information includes:

[0061] Extract license plate features and vehicle outline features from vehicle visual feature data;

[0062] Determine whether a vehicle has a valid parking space use permit based on its license plate characteristics;

[0063] The vehicle's identity category information is obtained by matching the vehicle's outline features against a pre-set vehicle database.

[0064] This embodiment specifies the logic for solving visual features, focusing on constructing a matching algorithm based on feature vector space; the system utilizes a convolutional neural network (CNN) to process the acquired vehicle visual feature data. Convolutional operations are performed to extract license plate features, including the license plate color HSV histogram and character OCR recognition results, and the region of interest (ROI) of the vehicle is located. Based on this ROI, the vehicle contour features are specifically quantized into a set of multi-dimensional physical feature vectors. ,in, The estimated width and height of the vehicle body after pixel mapping are given, with the unit uniformly in meters. Since the vehicle body length is difficult to obtain directly from the forward view, this embodiment uses width and height features combined with texture features for dimensionality reduction matching. To solve the black-box problem of mapping from the two-dimensional image domain to the three-dimensional physical domain, the system performs back-projection calculation based on the calibration parameters of the binocular camera: a semi-global stereo matching algorithm is performed on the left and right views acquired by the binocular camera to generate a dense disparity map. Using disparity maps Pixel disparity value Calculation depth:

[0065] ;

[0066] in, For disparity map The pixel disparity value at the corresponding coordinate point; Focal length The baseline is shown in meters. This parameter, representing the pixel size of the image sensor in meters, is introduced to ensure the accuracy of the calculation results. It has the correct physical length dimensions, combined with the intrinsic parameter matrix. contour pixel coordinates The data is converted into a 3D point cloud set in the camera coordinate system, and the minimum bounding box algorithm is applied to this point cloud set to extract its physical dimensions. ,in, Indicates the physical length of the vehicle; The texture complexity coefficient for the front face feature region is defined; to ensure the reproducibility of the feature extraction algorithm, the system defines the calculation... The required region of interest for the front grille is defined as follows: a rectangular area, 60% of the width of the vehicle's outline bounding box and 30% of its height, located in the lower-middle vertical direction, specifically within the 35%-65% height range. This area is used to physically isolate the license plate character area from the road background area, thus preventing interference with texture calculation. This is done to address the confusion between physical meaning and dimensions caused by direct variance calculation. Specifically, the variance is calculated using the normalized Laplace operator:

[0067] ;

[0068] in, Representing an image In pixel coordinates The second-order spatial derivative response value at a point represents the local edge gradient strength at that point; The Laplacian response value of all pixels in this region The arithmetic mean; The sum of the absolute response coefficients of the Laplacian operator convolution kernel is the gain normalization factor of the convolution kernel. For example, for the 8-neighborhood operator, the value is 8. This coefficient is introduced to ensure that the result is strictly mapped to the [0,1] interval. To determine the maximum quantized grayscale value of the image sensor, for example, 255 for an 8-bit image, we introduce... The denominator is used to normalize the texture energy to the upper limit of the sensor's dynamic range, thereby eliminating the square dimension of grayscale. This becomes a dimensionless coefficient that conforms to the definition, taking values ​​in the range [0,1]; where, This represents the pixel coordinates within the region of interest of the front grille. This represents the total number of pixels in the region. The Laplacian response value of all pixels in this region Image exist The arithmetic mean of the output results after the points are convolved by the Laplacian operator; this metric is used to quantify the texture density of the vehicle's front grille area; the system executes attribute determination logic to identify whether the vehicle has a valid parking space use permit attribute based on the license plate features, for example, a green license plate is determined as True;

[0069] To acquire identity category information, the system executes vector space matching logic: Considering the significant dimensional difference between physical size (meter-level values) and texture coefficients (typically dimensionless values ​​between 0 and 1), directly calculating the Euclidean distance would cause large-scale features to mask small-scale features. Therefore, the system utilizes the corresponding statistical mean vectors of each dimension from a pre-set database. and standard deviation vector For eigenvectors Each component is independently normalized using Z-Score: Here, a small regularization constant is introduced. For example, 1e-6, to prevent division by zero errors when the standard deviation is zero due to a single sample; and Perform Euclidean distance calculations to obtain the distance set. ,in, This is the normalized vehicle physical feature vector. This is a standard vehicle model template vector set from a pre-defined vehicle database; Note: The template vectors in the database need to be pre-processed to reduce their length dimension to match the input vector. The structure of this set has been pre-processed with the same normalization; ,in, This represents the normalized feature vector compared to the first feature vector in the preset vehicle database. Standard vehicle template vector Euclidean distance between them; select The index corresponding to the minimum value As a preliminary matching result, if the minimum value is less than the preset similarity tolerance... For example, 0.15, then the output will be... Corresponding vehicle identity category information For example, A00-class pure electric vehicles or B-class plug-in hybrids, and retrieve the nominal battery capacity associated with that category from the database. With power parameters; if the minimum value is greater than If the model is not specified, it is marked as an unknown model and the default general parameters are used. This process concretizes abstract feature matching into computable vector operations, ensuring the determinism of physical parameter acquisition. Example 3:

[0070] The process by which the central processing unit outputs the real-time parking space status spectrum includes:

[0071] Determine whether the vehicle has a valid parking space use permit.

[0072] If a vehicle does not have a valid parking space use permit, and the duration of its entry into the parking space in the time series data exceeds the first preset time threshold, and the device connection status signal is disconnected, then the real-time parking space status spectrum will be marked as an illegal occupancy status.

[0073] If a vehicle has a valid parking space use permit, and the duration of its occupancy in the parking space time series data exceeds the second preset time threshold, and no effective device connection behavior occurs, then the real-time parking space status spectrum will be marked as a passive occupancy state.

[0074] This embodiment details the logic for determining abnormal parking space occupancy; the central processing unit determines whether the vehicle has a valid parking space use permit; the system executes the illegal occupancy determination logic, responding when the vehicle does not have a permit and the timing logic module records the duration of the occupancy. Greater than the first preset time threshold Meanwhile, the device connection status signal The system will display the parking space status spectrum in real time as disconnected. Marked as an illegal occupancy state; among which, Derived from a preset fault tolerance time, its physical meaning is a buffer period, such as 5 minutes, allowing a fuel-powered vehicle to mistakenly enter and then leave; the system executes passive occupancy judgment logic, responding to the vehicle having permission attributes, but Greater than the second preset time threshold And if no valid device connection behavior is detected, the system will Marked as a negative placeholder state; among which, Derived from operational strategy settings, the physical meaning is the maximum allowable preparation time for a new energy vehicle to start charging after parking, such as 15 minutes; to ensure the continuity and completeness of system logic on the timeline, this embodiment further defines the pre-observation period logic: at the moment the vehicle enters the parking space... Beginning, to If the above-mentioned conditions for determining illegal or passive occupation are not met within the time window, i.e. During the threshold accumulation process, the central processing unit will display the real-time parking space status spectrum. Forced lockout is in compliance monitoring state; this state, as the logical default branch, explicitly blocks the punitive command output of the performance control module, thereby providing users with a reasonable buffer period for parking adjustments and equipment operation, and avoiding system anomalies or false alarms caused by undefined logic.

[0075] This embodiment introduces a dual time threshold. and A hierarchical filtering mechanism was constructed; this mechanism, on the one hand, through... It provides necessary margin for error, avoiding false alarms for fuel-powered vehicles that enter briefly or unintentionally; on the other hand, through It accurately identifies the passive behavior of parking without charging, filling the loophole in existing technologies that rely solely on green license plates for compliance determination. This effectively combats the inefficient use of charging resources as ordinary parking spaces and significantly improves the turnover rate of charging piles. Furthermore, regarding the aforementioned first preset time threshold... Second preset time threshold The specific determination method disclosed in this embodiment is a statistical optimization scheme based on historical data: the system establishes a statistical histogram of transit behavior, statistically analyzes all samples from the past three months that are not connected to any device and have a stay time of less than 30 minutes, and takes the 85th percentile value of this distribution as... The baseline value, for example, is that if 85% of the mistakenly entered vehicles leave within 4 minutes and 30 seconds, then Set to 5 minutes;

[0076] Simultaneously, a statistical histogram of charging preparation behavior was established, and the time difference from input to connection was statistically analyzed for all samples that successfully established a charging connection. The 95th percentile value of this distribution was taken as the... The baseline value is calculated and updated quarterly by the backend server, so that the threshold parameter can adapt to the spatial layout complexity of different sites and the operational proficiency of user groups. Example 4:

[0077] The process of the central processing unit outputting the real-time parking space status spectrum also includes:

[0078] If the vehicle has a valid parking space use permit and a valid device connection has been established, then obtain the duration of the device connection from the parking space time series data.

[0079] A preset maximum allowed connection duration threshold is set. If the device connection duration exceeds the maximum allowed connection duration threshold, the real-time parking space status spectrum will be marked as timed out.

[0080] Otherwise, mark the real-time parking space status spectrum as compliantly occupied.

[0081] This embodiment further defines the compliance boundaries in the connected state; the system confirms that the vehicle has the necessary permits and detects valid device connection behavior, such as detecting a CC1 resistor signal or a successful BMS handshake; the system obtains the device connection duration from the parking space timing data. The system introduces a preset maximum allowed connection duration threshold. and perform comparison logic; in response to Greater than The system determines that the current behavior has exceeded the reasonable range of resource usage and will display the real-time parking space status spectrum. Mark as timeout status; otherwise, if Less than or equal to If it is, then it is marked as a compliant occupancy status; among which, It originates from the resource rotation benchmark set by the operator, and its physical meaning is the maximum time slice that a single service is allowed to occupy;

[0082] This embodiment achieves refined time-slice management of charging resources by setting a maximum allowed connection duration threshold. This logic clearly distinguishes between normal charging services and resource stagnation caused by not leaving after charging or long-term trickle charging, ensuring that public charging resources are not monopolized by a single user for a long time, thereby maximizing service throughput under limited physical infrastructure conditions. Example 5:

[0083] The process of the central processing unit outputting the real-time parking space status spectrum also includes:

[0084] Establish a differentiated connection duration policy based on vehicle identity category;

[0085] Based on the current vehicle's identity category information, the corresponding differentiated allowed connection duration is invoked to update the maximum allowed connection duration threshold.

[0086] This embodiment dynamically upgrades the duration threshold setting mechanism and introduces real-time feedback adjustment based on station congestion; the system establishes a differentiated allowed connection duration strategy based on vehicle identity category; and uses the vehicle identity category information obtained in Embodiment 2. Call the corresponding differentiated allowed connection duration to update the maximum allowed connection duration threshold. Based on this, the system adopts the following capacity-congestion coupling model. Precise calculation:

[0087] ;

[0088] in, Sourced from a pre-defined database and identity categories. The physical meaning of "matching" is the vehicle's nominal battery capacity, measured in kWh. Defined as the energy demand ratio for the current session; to avoid individual calculation biases caused by relying solely on historical averages, the system executes dual-mode value retrieval logic: it prioritizes reading the current state of charge from the BMS handshake message through the electrical monitoring module. Compared with the target state of charge set by the user If not set, the default value is 0.95. Calculation If BMS communication has not yet been established or data fields are missing, the system will automatically downgrade to using historical big data statistics. To ensure the availability and real-time nature of this statistical value, the system maintains a vehicle type-energy characteristic mapping table in local storage. The table is constructed and updated as follows: whenever a certain identity category is detected... Once a vehicle completes a full charging session, encompassing the entire process from BMS handshake to disconnection, the system records its initial SOC value and updates the corresponding table for that category using a sliding window algorithm with a window size of N=50. Value; if no record of this category exists in the table, it is initialized to the industry default value of 0.3; Defined as the system bottleneck power for the current charging session, the calculation formula is:

[0089] ;

[0090] in, The rated output power of the current charging pile is identified from the device connection status signal. The vehicle's identity category is derived from a pre-set database. The maximum supported DC receiving power is limited to 40kW for A00-class vehicles; if the vehicle category is unknown, the default is used. The reduction factor is introduced here to prevent the risk of incorrectly triggering the timeout judgment in scenarios where the unknown vehicle model is a low-power vehicle and it is parked at a high-power charging station, due to the use of full power calculation resulting in an excessively short theoretical charging time. : This is the preset equivalent coefficient for the charging curve, for example, a value of 0.75, used to correct for the average power being lower than the rated power due to natural power decay during constant current and constant voltage charging. The physical phenomena are used to prevent misjudgment and timeout due to insufficient theoretical calculation time;

[0091] The specific method for determining this is based on the vehicle model's identity category. Standard charging test curve The curve data is pre-stored in the system's vehicle physical characteristic database. The data comes from the GB / T27930 communication protocol conformance test reports published by various automakers or third-party test data. The corresponding power values ​​are stored at a resolution of 1% SOC, and the actual energy demand from the current state of charge to the fully charged state is calculated by integration. And calculate if peak power Theoretical energy required for constant charging Take the ratio of the two. This serves as the basis for quantifying the charging efficiency degradation characteristics of this vehicle model; : is a time unit conversion factor used to unify physical calculation results to the system's timing logic baseline; due to the first part of the formula The calculation results are in hours, while the time thresholds and countdown logic in this system are measured in minutes. Therefore, the physical dimension is set as follows: To achieve alignment of physical dimensions from hours to minutes; : This is a dynamic, elastic buffer coefficient; the system calculates the station congestion index in real time. Determine The value;

[0092] The specific calculation logic is as follows: obtain the total number of parking spaces at the current depot. The number of parking spaces that are not idle in the real-time parking space status spectrum To avoid logical loops, i.e., the state spectrum depends on the threshold, and the threshold depends on the state spectrum, this is explicitly defined here. The visual perception module detects the number of physical parking spaces where a vehicle exists, independent of subsequent compliance determination results; calculation ,in, Substitute into the negative correlation mapping function To ensure the effectiveness of the congestion control mechanism, the parameter settings must be guaranteed to work during congestion. It can enter the negative range to accelerate the flow, for example, let's say... That is, the idle reward coefficient. That is, the congestion sensitivity coefficient. This refers to the maximum compression lower limit; the method for determining the above parameters is as follows: It depends on the operator's tolerance for marketing strategies that attract users to stay during off-peak periods, and is usually set at 0.2-0.3; Depending on the intensity of demand during peak periods for drastically reducing time to increase table turnover, it is typically set to -0.1 to -0.3. Then set to ,in, The saturation congestion level that achieves the maximum compression effect is physically defined as the critical density threshold at which the service throughput of the station changes from increasing to decreasing. In this example, it is set to 0.75. This value is determined based on the critical inflection point of the station's historical flow-speed macroscopic basic graph, that is, the critical density value at which the flow throughput begins to decrease with increasing density.

[0093] The specific calculation steps are as follows: The system collects the vehicle entry and exit records of the station for the past 30 natural days, and calculates the equivalent traffic flow in each time slot with a time granularity of 15 minutes. The unit is vehicles / 15min, and the average parking space occupancy rate. ,in, ;Build dataset The least squares method was used to fit the parabolic model:

[0094] ;

[0095] constraint Calculate the x-coordinate of the vertex of the parabola. , will Value as saturation congestion level If the calculation result exceeds Interval or goodness of fit If the congestion index is set to 0.75, a linear congestion-duration feedback mechanism is constructed. Under this parameter setting, when the congestion index... When it exceeds 0.4167, It begins to turn negative; when When it reaches 0.75, or 75% occupancy, When the maximum compression limit is reached, the allowed duration is... This will be forcibly shortened by 20%, thereby implementing a dynamic management strategy for time-limited charging during congestion periods at the physical level; this formula not only considers the physical charging needs of the vehicle models, but also introduces... The variables mathematically couple the micro-level bicycle management with the macro-level station congestion status. Example 6:

[0096] The process by which the efficiency management module generates traffic management instructions based on the real-time parking space status spectrum includes:

[0097] Extract the current state marker of the real-time parking space status spectrum;

[0098] If the current status is marked as illegal occupancy or passive occupancy, a parking space release prompt command or a parking space lock control command will be generated.

[0099] If the current status is marked as timeout occupied, a tiered rate billing instruction or a departure reminder instruction will be generated.

[0100] The process by which the performance management module drives the on-site traffic indication equipment to execute action strategies includes:

[0101] Obtain the control interface of the status indicator unit set in the parking space area;

[0102] Establish a mapping relationship between the real-time parking space status spectrum and the displayed status;

[0103] In response to changes in the real-time parking space status spectrum, the status indicator unit is driven to switch to the corresponding display status through the control interface. The illegal occupancy status corresponds to the first warning display status, the compliant occupancy status corresponds to the second permitted display status, and the timed-out occupancy status corresponds to the third warning display status.

[0104] This embodiment details the closed-loop control and on-site interaction logic based on state spectrum; the efficiency control module extracts the real-time parking space state spectrum. The current state flag;

[0105] The system executes hierarchical instruction generation logic: in response to If the parking space is illegally or passively occupied, the system determines it as malicious occupation and generates a parking space release prompt or parking space lock control command, such as controlling the smart lock to rise to restrict departure or even incur a penalty fee; in response to If the parking space is in an overdue occupancy state, the system determines it as a non-malicious but inefficient behavior and generates a tiered pricing instruction or a departure reminder instruction; the system obtains the control interface of the status indicator unit set in the parking space area and establishes... Mapping relationship with display state; responsive to The changes in the drive status indicator unit switch the display: illegal occupation status corresponds to the first warning display status, such as a high-frequency flashing red light; compliant occupation status corresponds to the second permitted display status, such as a breathing blue light; and timeout occupation status corresponds to the third warning display status, such as a constantly lit yellow light.

[0106] This embodiment constructs a complete closed loop from monitoring to control. By distinguishing between malicious and inefficient behaviors and adopting a tiered control strategy, it maintains charging order while demonstrating management flexibility. At the same time, the visual status indicators make on-site information intuitive and transparent, which creates a psychological deterrent to potential violators and provides long-distance visual guidance for users looking for parking spaces, greatly improving the overall operational efficiency and sense of order of the station. Example 7:

[0107] The central processing unit is also used to execute parking space usage prediction logic:

[0108] Based on historical parking space time series data, we analyze parking space usage patterns in different time periods;

[0109] By combining the real-time parking space status spectrum with the current time, the estimated occupancy time or the estimated release time of the parking space can be predicted.

[0110] The expected occupancy time or expected release time will be sent to the data publishing port of the regional parking guidance system.

[0111] This embodiment endows the system with the ability to predict future states, focusing on solving the prediction problem in non-charging states; the central processing unit acquires the real-time parking space state spectrum. and according to Different values ​​trigger branch prediction logic: Branch 1: When When a connection is marked as compliant or timed out, indicating a valid connection, the system uses an electrochemical physical model for prediction; this takes into account the real-time power during the initial charging start-up or trickle charging phase. The power value may approach zero, and using it directly as the denominator could lead to computational overflow or abnormally high predicted values. Therefore, the system introduces a minimum effective power threshold. For example, 2.0kW;

[0112] The threshold is determined by obtaining the baseline standby power of the vehicle with the air conditioning and BMS turned on when it is not charging. Combined with the minimum reliable resolution of the charging pile metering module ,set up ,in, The signal-to-noise ratio coefficient, with a value of 1.5-2.0, accurately distinguishes between effective charging trickle current and standby virtual connection power consumption; the calculation formula is revised as follows:

[0113] ;

[0114] in, This refers to the nominal capacity of the vehicle's battery. The target state of charge is derived from the user's charging order settings. If not obtained, it will be set to 0.95 by default. For real-time SOC, it is obtained by parsing BMS messages through the electrical monitoring module. For real-time charging power, same as above. The time unit conversion factor is set to 60 to output minute-level data. The average dwell time for this vehicle type when drawing a gun is calculated based on historical data, in minutes. The calculation method is as follows: the vehicle type is selected from historical parking space time-series data. For all compliant charging records, the time of device disconnection and the time of charging completion in each record are calculated, defined as the moment when the current drops to 0A. The specific logical criterion is: the electrical monitoring module samples the output current at a fixed frequency, such as 1Hz. Set the zero current noise floor threshold When continuous One, such as The current values ​​at the sampling points all satisfy At that time, the timestamp of the first sampling point in the continuous sequence is marked as the charging end time to filter out instantaneous fluctuation interference; the time difference between the charging end time and the device disconnection time is calculated, and the arithmetic mean of the most recent 100 records is taken as the charging end time. And updated monthly;

[0115] Branch 2: When When a parking space is marked as illegally occupied or passively occupied (i.e., without a valid connection), and electrical parameters cannot be obtained, the system switches to the behavior pattern statistics library in the historical parking space time-series data. The historical parking space time-series data is constructed as follows: the system extracts complete session records containing entry, connection, disconnection, and departure times from the original operation log daily. After outlier cleaning, records shorter than 1 minute or longer than 24 hours are removed and stored in the time-series database as the basic data source for the behavior pattern statistics library. The system then performs time-series clustering analysis to construct feature vectors. Considering time data The 0-24 hour timeframe exhibits a periodic characteristic of consecutive beginnings and endings; that is, 23:59 and 00:01 are extremely close in physical time, but their numerical differences are significant in traditional Euclidean distance calculations, leading to clustering failure. This embodiment introduces a unit circle mapping mechanism: transforming the one-dimensional linear time... Mapped to two-dimensional coordinates ,in, , ; eigenvectors reconstructed as The K-Means algorithm was used to cluster historical data on illegal occupation of land into... For each typical pattern center, the expected value of the occupancy time is calculated. At that time, calculate the arithmetic mean of the sine and cosine components of all samples within the cluster. Then, the time center is restored using the arctangent function. That is, the moment when the centroid of the cluster on the time cycle circle is projected back onto the linear time axis:

[0116] (twenty four);

[0117] in, This represents the modulo operation; the addition of 24 followed by the modulo operation is introduced in this formula to solve... The problem of incorrect physical time calculation caused by negative function output values ​​was resolved by normalizing the results to the 0-24 range, thereby accurately capturing the long-term nighttime occupation pattern that crosses midnight.

[0118] During real-time prediction, the system calculates the current vehicle entry time. Similarly, perform the trigonometric function mapping described above, and use the Euclidean distance in the mapping space to identify the nearest cluster index based on the time dimension distance to each cluster center. Extract the expected value of the cluster. As a baseline duration; calculate the estimated remaining usage time:

[0119] ;

[0120] in, The system uses a dual-mode prediction mechanism to accurately predict the release time of vehicles that are charging, and also provides a probabilistic prediction of the release time of illegally occupied vehicles based on the law of large numbers. This provides dynamic time data for the entire parking guidance system, avoiding the problem of prediction function failure when encountering illegal parking in traditional systems.

[0121] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention.

Claims

1. A new energy vehicle charging pile parking space status monitoring system, characterized in that, It includes a central processing unit, which is communicatively connected to a visual perception module, an electrical monitoring module, a timing logic module, and a performance control module; The visual perception module is used to collect vehicle visual feature data within the parking space area and calculate vehicle parking attribute information based on the vehicle visual feature data. The electrical monitoring module is used to collect equipment connection status signals on the charging pile side; The timing logic module is used to generate parking space timing data containing information on vehicle entry time, device connection time, and status duration in response to changes in the vehicle visual feature data and the device connection status signal. The central processing unit is pre-set with a parking space status analysis model, which is used to input the vehicle parking attribute information, the device connection status signal and the parking space time series data into the parking space status analysis model, calculate and output the real-time parking space status spectrum that represents the compliance and efficiency of the current parking space use; The efficiency control module is used to generate traffic management instructions based on the real-time parking space status spectrum, and drive the on-site traffic indication equipment to execute corresponding action strategies based on the traffic management instructions. The process by which the visual perception module calculates vehicle parking attribute information includes: Extract the license plate features and vehicle outline features from the vehicle visual feature data; The vehicle's license plate characteristics are used to determine whether the vehicle possesses a valid parking space usage permit. The vehicle's identity category information is obtained by matching the vehicle's outline features against a preset vehicle database. The process of the central processing unit outputting the real-time parking space status spectrum also includes: Establish a differentiated connection duration policy based on vehicle identity category; Based on the current vehicle's identity category information, the corresponding differentiated allowed connection duration is invoked to update the maximum allowed connection duration threshold; The vehicle contour features are specifically quantized into a set of multi-dimensional physical feature vectors. ,in, These correspond to the estimated width and height of the vehicle body after pixel mapping, respectively. Let be the texture complexity coefficient of the front face feature region, and The specific method for calculating variance using the normalized Laplace operator is as follows: ； in, Representing an image In pixel coordinates The second-order spatial derivative response value at a point represents the local edge gradient strength at that point; The Laplacian response value of all pixels in this region The arithmetic mean; This is the sum of the absolute response coefficients of the Laplacian operator convolution kernel; represents the maximum quantized grayscale value of the image sensor; where, This represents the pixel coordinates within the region of interest of the front grille. This represents the total number of pixels in the region. The Laplacian response value of all pixels in this region Image exist The arithmetic mean of the output results after a point is convolved by the Laplacian operator.

2. The new energy vehicle charging pile parking space status monitoring system according to claim 1, characterized in that, The process by which the central processing unit outputs the real-time parking space status spectrum includes: Determine whether the vehicle possesses the aforementioned valid parking space use permit attributes; If a vehicle does not have the legal parking space use permit attribute, and the duration of its entry into the parking space time series data is greater than the first preset time threshold, and the device connection status signal is disconnected, then the real-time parking space status spectrum will be marked as an illegal occupancy status. If a vehicle possesses the legal parking space usage permit attribute, and the duration of its occupancy in the parking space time series data exceeds the second preset time threshold, and no effective device connection behavior occurs, then the real-time parking space status spectrum is marked as a passive occupancy state.

3. The new energy vehicle charging pile parking space status monitoring system according to claim 2, characterized in that, The process of the central processing unit outputting the real-time parking space status spectrum also includes: If the vehicle has the legal parking space use permit attribute and a valid device connection has occurred, then obtain the device connection duration from the parking space time sequence data. A preset maximum allowed connection duration threshold is set. If the duration of the device connection exceeds the maximum allowed connection duration threshold, the real-time parking space status spectrum is marked as an overdue occupancy status. Otherwise, the real-time parking space status spectrum will be marked as compliant occupancy.

4. The new energy vehicle charging pile parking space status monitoring system according to claim 3, characterized in that, The process by which the efficiency control module generates traffic management instructions based on the real-time parking space status spectrum includes: Extract the current state marker of the real-time parking space status spectrum; If the current state is marked as an illegal occupancy state or a passive occupancy state, a parking space release prompt command or a parking space lock control command is generated. If the current status is marked as an overdue occupancy status, a tiered rate billing instruction or a departure reminder instruction will be generated.

5. The new energy vehicle charging pile parking space status monitoring system according to claim 4, characterized in that, The process by which the efficiency control module drives the on-site traffic indication equipment to execute action strategies includes: Obtain the control interface of the status indicator unit set in the parking space area; Establish a mapping relationship between the real-time parking space status spectrum and the displayed status; In response to changes in the real-time parking space status spectrum, the status indicator unit is driven to switch to the corresponding display status through the control interface. The illegal occupancy status corresponds to the first warning display status, the compliant occupancy status corresponds to the second permitted display status, and the timed-out occupancy status corresponds to the third warning display status.

6. The new energy vehicle charging pile parking space status monitoring system according to claim 1, characterized in that, The central processing unit is also used to execute parking space usage prediction logic: Based on historical parking space time series data, we analyze parking space usage patterns in different time periods; By combining the real-time parking space status spectrum with the current time, the estimated occupancy time or the estimated release time of the parking space can be predicted. The estimated occupancy time or estimated release time is sent to the data publishing port of the regional parking guidance system.

Images