Charging station cloud operation and maintenance management method, system and equipment

Abstract

The invention relates to the technical field of operation and maintenance management, and discloses a cloud operation and maintenance management method, system and equipment for a charging station, and the method comprises the steps: monitoring the WiFi signal strength, 5G network quality and wired network delay parameters of each floor of a stereo garage in real time, and building a network environment database; when the charging pile detects that the SOC change amplitude exceeds a set threshold value, an optimal target transmission path set is screened out; carrying out differential bit coding on the collected charging data to generate a charging data frame, and sending the charging data frame to a cloud server in a multi-path concurrent mode; the cloud server receives and gathers the charging data frames from different floors of the stereo garage, and analyzes the charging load distribution condition of the whole garage; and calculating a power distribution scheme of the charging pile of each floor based on the whole garage charging load distribution condition, and issuing a scheduling instruction to the charging pile of the corresponding floor, thereby improving the intelligent operation and maintenance management level and resource utilization efficiency of the stereo garage charging station.

Description

Technical Field

[0001] The present invention relates to the field of operation and maintenance management technology, and in particular to a charging station cloud operation and maintenance management method, system and equipment. Background Art

[0002] Traditional charging data transmission methods based on a single network path cannot guarantee the stability and real-time performance of data transmission. Failures such as data loss, transmission interruption, and excessive delay often occur, seriously affecting the reliability of charging services and user experience.

[0003] Existing charging station data transmission technologies generally use fixed-precision data quantization and periodic upload mechanisms, lacking the adaptive capability to dynamically adjust according to changes in charging status and network environment. This leads to problems such as low data transmission efficiency, waste of network resources, and failure to upload key charging information in a stereoscopic parking garage environment with limited network bandwidth. Summary of the Invention

[0004] The present invention provides a cloud-based operation and maintenance management method, system, and equipment for charging stations, which improve the intelligent operation and maintenance management level and resource utilization efficiency of charging stations in stereo garages.

[0005] A first aspect of the present invention provides a charging station cloud operation and maintenance management method, the charging station cloud operation and maintenance management method comprising: Real-time monitoring of WiFi signal strength, 5G network quality, and wired network latency parameters on each floor of the parking garage to establish a network environment database; When the charging pile detects that the SOC change amplitude exceeds a set threshold, a target transmission path set with the best signal quality on the current floor is selected from the network environment database; Performing differentiated bit encoding on the collected charging data according to the current SOC range and charging power level of the vehicle to generate a charging data frame, and sending the charging data frame to the cloud server in a multi-path concurrent manner through the target transmission path set; The cloud server receives and aggregates charging data frames from different floors of the parking garage, and analyzes the charging load distribution of the entire garage. The power distribution plan for the charging piles on each floor is calculated based on the charging load distribution status of the entire warehouse, and the scheduling instructions are sent to the charging piles on the corresponding floor.

[0006] In combination with the first aspect, in a first implementation of the first aspect of the present invention, the real-time monitoring of WiFi signal strength, 5G network quality, and wired network delay parameters on each floor of the stereo garage and establishing a network environment database includes: Deploy network detection nodes on each floor of the stereo garage to collect WiFi signal strength, 5G network quality, and wired network delay parameters to form raw network parameter data; Classify and merge the original network parameter data according to floor number and network type, and calculate the comprehensive quality score of each network on each floor; When a change in vehicle load or electromagnetic interference fluctuation is detected in the stereo garage, the weight of the comprehensive quality score is adjusted and the priority ranking of the network paths on each floor is updated; Based on the priority sorting, the original network parameter data is stored in the corresponding database partition to establish a network environment database.

[0007] In combination with the first aspect, in a second implementation of the first aspect of the present invention, when the charging pile detects that the SOC change amplitude exceeds a set threshold, screening out a target transmission path set with the best signal quality on the current floor from the network environment database includes: The charging pile calculates the difference between the current SOC value and the last recorded SOC value in real time to obtain the SOC change range. When the SOC change range exceeds the set threshold, the path selection process is triggered; When triggering the path selection process, the WiFi, 5G and wired network path information of the corresponding floor are retrieved from the network environment database according to the floor identifier of the charging pile, and the set of available network paths on the current floor is obtained; Sorting the comprehensive quality score of each first candidate path in the set of available network paths on the current floor, and selecting the top three second candidate paths with the highest comprehensive quality scores; The first three second candidate paths are used as a target transmission path set, and the transmission parameter configuration of the target transmission path set is recorded.

[0008] In combination with the first aspect, in a third implementation of the first aspect of the present invention, performing differentiated bit encoding on the collected charging data according to the current SOC range and charging power level of the vehicle to generate a charging data frame, and sending the charging data frame to the cloud server in a multi-path concurrent manner through the target transmission path set, including: The charging pile determines the range to which the current SOC value belongs and obtains the current SOC range of the vehicle; The charging power level is divided according to the power value detected in real time by the charging pile, and the current SOC value is encoded by bit allocation based on the current SOC range of the vehicle to form SOC encoding data; Numerically compressing the charging power level according to the quantization step size of the corresponding power segment to generate compressed power coded data; The SOC coded data and the compressed power coded data are encapsulated and combined by adding a user identifier of a first identification digit, a charging pile number of a second identification digit, and a check code of a third check digit to form a charging data frame; The charging data frame is sent to the cloud server in a multi-path concurrent manner through the target transmission path set.

[0009] In combination with the first aspect, in a fourth implementation of the first aspect of the present invention, sending the charging data frame to the cloud server in a multi-path concurrent manner through the target transmission path set includes: The charging pile splits the charging data frame into several data slices according to the preset slice size, and adds a sequence identifier and a floor identifier to each data slice to form a slice data packet; Allocate the fragmented data packets to the WiFi path, the 5G path, and the wired network path according to the transmission parameter configuration of the target transmission path set; Simultaneously, multiple first network paths obtained after initialization of the second candidate path in the target transmission path set are started to execute concurrent sending processes, wherein WiFi paths among the multiple first network paths carry a first proportion of data slices, 5G paths among the multiple first network paths carry a second proportion of data slices, and wired network paths among the multiple first network paths carry a third proportion of data slices; The charging pile monitors the failure probability of each first network path in real time, and automatically switches the unsent fragments of the first network path to the second network path for continued transmission based on the failure probability; When the cloud server receives all the fragmented data packets of the same charging data frame, the charging pile receives a transmission completion confirmation signal from the cloud server and clears the locally cached fragmented data packets and path status records.

[0010] In combination with the first aspect, in a fifth implementation of the first aspect of the present invention, the charging pile monitors the failure probability of each first network path in real time, and automatically switches the unsent fragments of the first network path to the second network path for continued transmission based on the failure probability, including: The charging pile establishes a path quality attenuation prediction model based on the vehicle moving load change between floors of the stereo garage and the network signal attenuation law, and predicts the failure probability of each first network path through the path quality attenuation prediction model; Reordering the priority of the to-be-sent fragments on each first network path according to the failure probability to obtain a high-priority switching object; Adaptively re-encoding the high-priority switching object according to the bandwidth characteristics and delay characteristics of the second network path to generate path-adapted slice data; When the actual switching is triggered, the unsent fragments of the first network path are automatically switched to the second network path for continued transmission based on the path switching decision algorithm, and the parameter weights of the path quality attenuation prediction model are updated.

[0011] In combination with the first aspect, in a sixth implementation of the first aspect of the present invention, the cloud server receives and aggregates charging data frames from different floors of the stereo garage, and parses the charging load distribution status of the entire garage, including: The cloud server spatially classifies the received charging data frames according to the floor identifiers in the charging data frames and establishes a charging state matrix; Performing an inter-floor coupling relationship analysis on the power demand data of each floor in the charging state matrix to obtain a corrected charging load value; Based on the corrected charging load value, combined with the historical traffic flow pattern and charging behavior trajectory data of the three-dimensional parking garage, the charging demand change trend of each floor is predicted; Based on the overall power capacity limit of the stereo garage and the carrying capacity of the distribution lines on each floor, the charging demand change trend is optimized with power constraints to generate the charging load distribution status of the entire garage.

[0012] In combination with the first aspect, in a seventh implementation of the first aspect of the present invention, the power allocation scheme for the charging piles on each floor is calculated based on the charging load distribution status of the entire warehouse, and the scheduling instructions are sent to the charging piles on the corresponding floor, including: The cloud server performs vertical power flow modeling of the stereo garage based on the floor load weight and capacity margin analysis in the overall garage charging load distribution, and generates an inter-floor power coordination matrix; Performing spatiotemporal coupling on the inter-floor power coordination matrix to form a dynamic power redistribution strategy; Decomposing the dynamic power redistribution strategy according to the geographical location and electrical connection topology of the charging piles to obtain a power allocation plan for the charging piles on each floor; The power allocation scheme is differentially packaged and encoded according to the quality status of the target transmission path set on each floor to form a scheduling instruction, and the scheduling instruction is concurrently issued to the charging piles on the corresponding floor through the optimal path of each floor recorded in the network environment database.

[0013] A second aspect of the present invention provides a charging station cloud operation and maintenance management device, the charging station cloud operation and maintenance management device comprising: A real-time monitoring module is used to monitor the WiFi signal strength, 5G network quality, and wired network latency parameters on each floor of the parking garage in real time, and to establish a network environment database. A screening module is configured to screen out a target transmission path set with the best signal quality on the current floor from the network environment database when the charging pile detects that the SOC change amplitude exceeds a set threshold; A concurrent sending module, configured to perform differentiated bit encoding on the collected charging data according to the current SOC range and charging power level of the vehicle, generate a charging data frame, and send the charging data frame to the cloud server in a multi-path concurrent manner through the target transmission path set; The parsing module is used by the cloud server to receive and aggregate charging data frames from different floors of the parking garage and parse the charging load distribution status of the entire garage; The calculation module is used to calculate the power distribution plan of the charging piles on each floor based on the charging load distribution status of the entire warehouse, and send the scheduling instructions to the charging piles on the corresponding floor.

[0014] A third aspect of the present invention provides an electronic device, comprising: a memory and at least one processor, wherein the memory stores instructions; the at least one processor calls the instructions in the memory so that the electronic device executes the above-mentioned charging station cloud operation and maintenance management method.

[0015] Compared with the existing technology, the present invention has the following beneficial effects: by establishing a hierarchical network environment database and a dynamic path selection mechanism, intelligent switching and concurrent transmission of WiFi, 4G and wired networks are realized, significantly enhancing the reliability of data transmission in the complex environment of the stereo garage. The adaptive quantization coding technology based on SOC interval and charging power level effectively reduces the transmission load and optimizes bandwidth utilization efficiency while ensuring the accuracy of key information. The innovative path quality attenuation prediction model and adaptive recoding processing realize the prediction and seamless switching of transmission failures, avoiding the risk of data loss. The three-dimensional charging state matrix and inter-floor coupling relationship analysis technology realize the accurate analysis of load distribution in complex spatial environments, and the vertical power flow modeling and time-space coupling optimization algorithm ensure the coordinated scheduling and load balancing of charging piles on each floor. The closed-loop feedback control mechanism and real-time correction and update function realize the dynamic optimization of the scheduling strategy, improving the intelligent operation and maintenance management level and resource utilization efficiency of the stereo garage charging station. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0017] The structures, proportions, sizes, etc. depicted in the drawings of this specification are only used to match the contents disclosed in the specification so as to facilitate understanding and reading by persons familiar with this technology. They are not intended to limit the conditions under which the present invention can be implemented and therefore have no substantive technical significance. Any structural modifications, changes in proportional relationships, or adjustments in size should still fall within the scope of the technical contents disclosed in the present invention without affecting the effects and objectives that can be achieved by the present invention.

[0018] Figure 1 This is a flow chart of a charging station cloud operation and maintenance management method provided by an embodiment of the present invention; Figure 2 This is a schematic block diagram of the structure of a cloud-based operation and maintenance management device for a charging station provided by an embodiment of the present invention; Figure 3 It is a schematic block diagram of the structure of an electronic device provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0019] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of them. All other embodiments obtained by ordinary technicians in this field based on the embodiments of the present invention without making any creative efforts shall fall within the scope of protection of the present invention.

[0020] The flowcharts shown in the accompanying drawings are for illustrative purposes only and do not necessarily include all contents and operations / steps, nor must they be executed in the order described. For example, some operations / steps may be decomposed, combined, or partially merged, so the actual execution order may vary depending on the actual situation.

[0021] It should also be understood that the terminology used in this specification is for the purpose of describing particular embodiments only and is not intended to limit the present invention. As used in the specification and appended claims, the singular forms "a," "an," and "the" are intended to include the plural forms unless the context clearly indicates otherwise.

[0022] It should be further understood that the term "and / or" used in the present specification and the appended claims refers to any and all possible combinations of one or more of the associated listed items, and includes these combinations. Figure 1 , an embodiment of the charging station cloud operation and maintenance management method in an embodiment of the present invention includes: Step 100: Real-time monitoring of WiFi signal strength, 5G network quality, and wired network delay parameters on each floor of the stereo garage, and establishment of a network environment database; It is understandable that the execution subject of the present invention can be a charging station cloud operation and maintenance management device, or a terminal or a server, which is not limited here. The embodiment of the present invention is described by taking the server as the execution subject as an example.

[0023] Specifically, network detection nodes with multi-standard network monitoring capabilities are deployed in a targeted manner on each floor of the multi-story parking garage. The nodes integrate a WiFi receiving module, a 5G signal analysis module, and an Ethernet delay measurement unit, and continuously collect network parameter data in the target area within a set period. The WiFi signal strength is represented by the RSSI (Received Signal Strength Indicator) value, and the 5G network quality is comprehensively reflected based on RSRP (Reference Signal Received Power), SINR (Signal to Interference and Noise Ratio), and packet loss rate. The communication status of the wired network is quantified by detecting the minimum round-trip delay RTT and packet transmission jitter, thereby forming an original network parameter data set covering different physical link types. Raw network parameter data is structured and categorized and merged based on the floor number and network type corresponding to the location where it was collected. A multi-dimensional indicator matrix for different types of network connections on each floor is constructed. A comprehensive quality score is then calculated for each network path. This score utilizes a weighted normalization model, specifically including a normalized signal strength value, an inversely proportional value for packet loss rate, a normalized latency value, and a bandwidth utilization factor. These factors are then combined into a single score based on a pre-defined weight coefficient to represent the communication availability and stability of the path at the current point in time. When load detection sensors or electromagnetic interference detectors deployed within the garage detect changes in the distribution of vehicles within the garage or increased electromagnetic environment fluctuations in a local area, indicating signal obstruction or reflection interference, the original network score weights are dynamically adjusted. In areas with increased interference, the weight of the channel stability indicator is appropriately increased, while the weight of the signal strength indicator is reduced to more accurately reflect the actual availability of each path in this complex environment. Following this adjustment, the quality score of each network path is recalculated, and the priority of each path is reordered, updating the corresponding path's ranking on the current floor. All raw network parameter data, along with the calculated comprehensive scoring results and ranking information, are written into the network environment database. A partitioned storage strategy is adopted to divide the data table space using floor number and network type as the main index dimensions. The WiFi, 5G, and wired network performance indicators of each floor are stored in the corresponding database partitions. At the same time, each path is marked with the current priority level and timestamp, forming a network environment database that supports multi-path selection decisions.

[0024] Step 200: When the charging pile detects that the SOC change amplitude exceeds a set threshold, a target transmission path set with the best signal quality on the current floor is selected from the network environment database; Specifically, a SOC change monitoring mechanism is implemented, enabling the charging station to continuously record the current SOC (State of Charge) value at fixed sampling intervals during the charging process. This value is then numerically differentiated from the previously recorded SOC data to obtain the SOC change in real time. A 5% SOC threshold is set. When the system detects a SOC change greater than this threshold, it indicates a significant fluctuation in the vehicle's charging state, accompanied by power adjustments, billing updates, or user behavior responses. This triggers a path selection process to prioritize the transmission of critical data. Once the path selection process is triggered, the floor identifier associated with the charging station's unique number is used as a key search field to access a pre-built network environment database and retrieve the latest set of network path parameters for that floor. This information includes the RSSI value and bandwidth utilization of the WiFi channel, the RSRP and SINR metrics of the 5G channel, and the round-trip delay and jitter statistics of the wired network, forming the set of available network paths for the current floor. Each path is accompanied by the aforementioned comprehensive quality score, which serves as the primary basis for path quality assessment. The comprehensive score is obtained by integrating the four sub-indicators of signal strength, packet loss rate, delay and bandwidth, and normalizing them after weighting. The score range is between 0 and 1, and the higher the value, the better the communication performance. The comprehensive quality scores of all the first candidate paths in the above-mentioned current floor available path set are quickly sorted. The sorting algorithm is based on the ranking from high to low based on the score value. The top three paths are taken as the second candidate path set, and these three paths are used as the final target transmission path set. To ensure the traceability and reusability of the path selection decision process, the transmission parameter configuration corresponding to the target transmission path set screened out this time is recorded, including the current comprehensive score value, channel type, frequency band, delay average, maximum bandwidth utilization and path number of each path, and a timestamp and event type label are attached to each set of records.

[0025] Step 300: Differentiate the number of bits of the collected charging data according to the vehicle's current SOC range and charging power level to generate a charging data frame, and send the charging data frame to the cloud server in a multi-path concurrent manner through the target transmission path set; Specifically, the system detects the vehicle's current SOC (State of Charge) value in real time and divides it into the SOC interval based on the range of the value. For example, when the SOC is less than 20%, it is determined to be in the low-charge interval, when the SOC is between 20% and 80%, it is in the medium-charge interval, and if the SOC is greater than 80%, it is classified as the high-charge interval. The current SOC state is classified by this interval judgment. Combined with the power value obtained in real time by the built-in power detection module of the charging pile, it is divided into set power level segments according to the power size. For example, 0 to 30kW is the low-power segment, 30 to 80kW is the medium-power segment, and 80 to 120kW is the high-power segment. Different quantization step sizes are set for different power segments. The low-power segment uses a 0.1kW step size to improve accuracy, the medium-power segment uses a 0.5kW step size, and the high-power segment uses a 1kW step size, thereby achieving the most reasonable data compression with minimal redundancy. This power value is compressed by the step size to form the corresponding compressed power encoding data. At the same time, based on the SOC range determined above, different encoding bit counts are assigned to the SOC value. For example, 12-bit encoding is used for low-battery ranges to ensure high-precision monitoring during the initial charging phase; 10-bit encoding is used for medium-battery ranges to reduce bandwidth pressure while maintaining accuracy; and 8-bit encoding is used for high-battery ranges to further compress the data volume, resulting in SOC-encoded data that is inversely proportional to the transmission pressure. To ensure the identifiability and integrity of charging data transmitted over the network, the SOC-encoded data and compressed power-encoded data are encapsulated in data frames. Several identification fields are added to the header of the data frame, including a first identification bit (e.g., a 16-bit user ID) to identify the current user; a second identification bit (e.g., an 8-bit charging pile number) to uniquely identify the charging pile; and a third check bit (e.g., a 16-bit CRC checksum) to verify data validity. This creates a complete and structured data frame. After the data frame is encapsulated, a multi-path concurrent transmission mechanism is initiated based on the target transmission path set previously selected from the network environment database. This mechanism involves concurrently pushing copies of the same data frame over multiple links, such as WiFi channels, 5G cellular networks, or wired Ethernet. Each path maintains an independent communication session and sends it through different network interfaces. After receiving the data, the cloud server performs deduplication and sorting operations based on the timestamp and identification fields in the data frame, thereby achieving complete data integration.

[0026] Step 400: The cloud server receives and aggregates charging data frames from different floors of the parking garage, and analyzes the charging load distribution of the entire garage. Specifically, a charging load calculation and prediction module is constructed on the cloud server side, combining hierarchical identification, time series analysis, and coupling optimization. At the data receiving layer, this module uniformly receives and processes charging data frames uploaded from different paths and floors. Based on the floor identification field, charging station number, and timestamp information contained in the data frames, it classifies and groups them according to the spatial dimension. A three-dimensional charging state matrix is ​​constructed, with each dimension corresponding to the floor number, time window, and corresponding power demand value, thus constructing a charging state map covering the entire parking garage. Correlation analysis is performed on the instantaneous power demand data of each floor in this charging state matrix to identify the coupling of power fluctuations between floors caused by power scheduling, user behavior migration, or vehicle movement. To improve modeling accuracy, an inter-floor dependency matrix model based on weighted correlation coefficients and covariance analysis is introduced. By examining the synchronization and deviation of power changes between adjacent floors within the same time period, a power transfer influencing factor is derived. The instantaneous raw power data is then adjusted to a corrected charging load value that reflects actual usage trends. Based on the corrected charging load value, it is time-series fused with the data on vehicle flow patterns, entry and exit frequencies, user charging preferences, and behavioral trajectories stored in the historical database of the multi-story parking garage. A recurrent neural network-based prediction model is then used to estimate the changing trend of charging demand on each floor over the next period of time. This model performs multi-factor fitting of charging load changes based on factors such as daily peak hours, floor preference distribution, and holiday behavioral fluctuations, and outputs a floor-level charging demand prediction curve based on time periods. The above prediction results are input into the power capacity constraint optimization module, which uses the upper limit of the garage's overall distribution capacity, the capacity of the main power supply circuit, and the carrying capacity parameters of each floor's branches to construct a multi-objective constrained optimization equation set. Combined with a weighted minimization method or a distributed linear programming algorithm, the predicted load values ​​for each floor are corrected for power feasibility and prioritized, thereby generating a final charging load distribution for the entire garage that meets the physical conditions of the power supply and the actual distribution of demand.

[0027] Step 500: Calculate the power allocation plan for the charging piles on each floor based on the charging load distribution of the entire warehouse, and send the scheduling instructions to the charging piles on the corresponding floor.

[0028] Specifically, in the cloud-based dispatch control unit, based on the overall charging load distribution, parameters such as load pressure, user demand density, and residual power margin on each floor are quantitatively analyzed. Based on this, a vertical power flow model for the parking garage is established. This model uses floor numbers as nodes and inter-floor power transfer possibilities as edges. Taking into account the main distribution circuit capacity, vertical busbar transmission capacity, and the redundancy of each floor's distribution branches, a multi-dimensional coupled inter-floor power coordination matrix is ​​generated. Each element in this matrix represents the upper limit of power adjustment that a floor can release or request from adjacent floors at the current moment, enabling dynamic coordination of power resources across floors. This inter-floor power coordination matrix is ​​introduced into a spatiotemporal coupling modeling framework. By combining the changing charging behavior patterns and dynamic load profiles of the parking garage over different time periods, a dynamic evolution rule for power allocation is constructed. Furthermore, by predicting inter-floor demand fluctuations and power bottlenecks within a certain period of time, the static coordination matrix is ​​expanded into a dynamic power reallocation strategy with time-recurring characteristics. This strategy adaptively adjusts the upper and lower limits of available power resources on each floor, without exceeding the total power capacity or local load limits of each floor, achieving local load shifting and global resource rebalancing. To ensure the effective implementation of this strategy at the physical level, the cloud system performs a spatial decomposition of the reallocation strategy based on the actual location and electrical connection topology of each charging station in the parking garage. The system calculates the target power output value for each charging station on each floor, resulting in a power allocation plan at the individual station level. During the dispatch instruction construction phase, the dispatch instruction is differentiated by selecting appropriate encoding compression methods and transmission encapsulation formats, taking into account the current network transmission conditions of the target floor. Specifically, the quality of the target transmission path set for each floor, previously obtained from the network environment database, includes key indicators such as signal strength, stability, latency, and packet loss rate. High-quality paths can carry high-density command data packets and use standard compression and full redundancy checking. In the case of poor path quality, a high-redundancy, low-rate, segmented encapsulation method is used to enhance data arrival rate and parsing error tolerance, resulting in a complete dispatch instruction packet with adaptive transmission characteristics. Based on the optimal transmission path table recorded in the network environment database, the cloud server calls the corresponding path interface and concurrently sends the generated dispatch instructions to all charging piles on the target floor. Each path is transmitted independently, and the dispatch command includes the pile number, target power value, execution timestamp, and checksum field. The receiving end completes command parsing and parameter update by comparing the number and checksum, achieving dynamic power allocation and refined control across the entire storage area.

[0029] In a specific embodiment, the process of executing step 100 may specifically include the following steps: Deploy network detection nodes on each floor of the parking garage to collect WiFi signal strength, 5G network quality, and wired network latency parameters to generate raw network parameter data. The original network parameter data is classified and merged according to the floor number and network type, and the comprehensive quality score of each network type on each floor is calculated; When changes in vehicle load or electromagnetic interference fluctuations are detected in the parking garage, the overall quality score is weighted and the priority ranking of network paths on each floor is updated; The original network parameter data is stored in the corresponding database partition based on priority sorting to establish a network environment database.

[0030] Specifically, network monitoring nodes are installed on each floor of the multi-story parking garage. These nodes integrate interface modules for multiple communication protocols, including a WiFi channel receiver for receiving and demodulating wireless signals, a 5G module supporting NSA / SA standards, and an Ethernet port module with precise time synchronization. Each network monitoring node periodically samples and actively measures raw communication parameters at its location, including WiFi signal strength (RSSI), 5G network metrics (such as RSRP, RSRQ, and SINR), and Ethernet communication latency characteristics (such as minimum round-trip time and delay jitter). The collected data is cached in a local cache in a high-frequency format and aggregated and transmitted to a cloud-based analysis module via an internal encrypted channel, forming a set of raw network parameter data. Using the floor number and network type identifier fields included with each set of raw parameter data as classification keys, this data is structured into a multidimensional index table, with each sub-table corresponding to a specific floor and network type (e.g., WiFi on the third floor, 5G on the fifth floor). For each subtable's network parameter sample sequence, a unified comprehensive quality evaluation model is constructed using a weighted scoring mechanism. This model uses four performance metrics as input parameters: signal strength (normalized RSSI or RSRP), packet loss rate (calculated using a sliding statistical window), latency (e.g., mean RTT), and bandwidth utilization (reflecting network load). Weight coefficients are assigned to each metric, with the sum of these four coefficients equaling 1, reflecting the varying sensitivity of different network types to these performance metrics. Using these weight coefficients, a real-time comprehensive quality score is calculated for each network channel. A sliding average mechanism is used to smooth data fluctuations, resulting in a network availability description vector for the current period. Furthermore, to enable the network channel scoring to dynamically respond to environmental changes within the parking garage, vehicle load sensing and electromagnetic interference detection mechanisms are introduced as triggers for weight adjustment. Vehicle load fluctuations are detected by intelligent cameras, geomagnetic sensors, or laser scanning modules deployed at entrances and exits and in parking areas on each floor. Significant changes in vehicle density on a particular floor can obstruct signal propagation paths. Similarly, when electromagnetic interference detectors detect increased interference power in specific frequency bands (e.g., due to high-power devices or external radiation intrusion), the effectiveness of wireless communication links can be affected. When the system detects any of these changes, it recalibrates the weighting of various performance indicators in the quality scoring model. For example, it can reduce the weight of signal strength in areas of increased interference, increase the weight of jitter tolerance, or increase the weight of bandwidth utilization during peak load periods, making the scoring system more adaptable to real-time conditions. The revised network scores are used to construct a prioritized list of network paths for each floor. Within each floor, candidate channels—such as WiFi, 5G, and Ethernet—are reordered by their scores to produce a target path sequence prioritized by real-time performance.At the storage level, the system writes the corresponding raw network parameter data to dedicated partitions within the database based on the sorting results. Each partition is indexed by floor number and network type, and includes the latest score, weight configuration, interference event identifier, and update timestamp information. To support cross-period data analysis and forecasting, each partition maintains historical score traces and trend curves for training path switching prediction models or modeling communication interruption warning logic.

[0031] In a specific embodiment, the process of executing step 200 may specifically include the following steps: The charging pile calculates the difference between the current SOC value and the last recorded SOC value in real time to obtain the SOC change range. When the SOC change range exceeds the set threshold, the path selection process is triggered; When triggering the path selection process, the WiFi, 5G, and wired network path information of the corresponding floor are retrieved from the network environment database according to the floor identifier of the charging pile, and the set of available network paths on the current floor is obtained; Sort the comprehensive quality scores of each first candidate path in the set of available network paths on the current floor, and select the top three second candidate paths with the highest comprehensive quality scores; The first three second candidate paths are used as a target transmission path set, and the transmission parameter configuration of the target transmission path set is recorded.

[0032] Specifically, first, in the status detection phase, the charging pile's internal control system is equipped with a real-time SOC monitoring module. This module reads the current SOC value from the battery management system at fixed time intervals (for example, every 30 seconds) and performs a numerical difference operation with the SOC value recorded in the previous time period to obtain the current SOC change amplitude. The system has a preset SOC threshold between 5% and 10% to determine whether the current charging state is in a significant change phase. When the SOC change amplitude is greater than this threshold, it indicates that the vehicle is in a critical phase such as a high-current charging stage, a power change point, or a critical billing stage. Therefore, the reliability of the data transmission path is re-evaluated, thereby triggering the path selection process. After the path selection process is initiated, the system extracts the floor identifier of the current charging pile from the local or cloud-based pile location mapping table based on its unique identification code. The floor number is used as the key index field to access the network environment database and retrieve all available communication path records corresponding to that floor. This record includes core transmission parameters such as WiFi path information (such as SSID number, RSSI value, and connection stability), 5G path information (such as NR cell number, RSRP, SINR), and wired Ethernet path information (such as link bandwidth, mean latency, and packet loss rate). All path information is accompanied by a timestamp, device port number, channel availability status indicator, and network score. Based on these parameters, the system establishes a set of network paths for the current floor, marking it as the first candidate path set. To select the optimal channel group from this first candidate path set, the system incorporates a multi-factor integrated path quality scoring model, calculating a quality score for each path in the set. This score is a normalized value between 0 and 1, constructed using the weighting function Q = α·S + β·(1-L) + γ·(1-D) + δ·B, where S is the normalized signal strength value, L is the packet loss rate, D is the normalized latency value, and B is the bandwidth utilization. α, β, γ, and δ are weighting parameters configured by the system based on the current network type and service priority. After the scoring calculation is complete, the system sorts all candidate paths in descending order using a standard quick sort or stable heap sort algorithm. The top three paths with the highest scores are retained to form the second set of candidate paths. This second set of candidate paths becomes the final set of target transmission paths. The system registers this set of paths in the path task scheduling table, along with its corresponding detailed transmission parameter configuration, including each path's score, path type (Wi-Fi / 5G / Ethernet), maximum available bandwidth, average latency, port number, channel ID, channel idleness, and estimated availability duration.

[0033] In a specific embodiment, the process of executing step 300 may specifically include the following steps: The charging pile determines the range to which the current SOC value belongs and obtains the current SOC range of the vehicle; The charging power level is divided according to the power value detected by the charging pile in real time, and the current SOC value is encoded by bit allocation based on the current SOC range of the vehicle to form SOC encoding data; Compress the charging power level numerically according to the quantization step size of the corresponding power segment to generate compressed power encoding data; The SOC coded data and the compressed power coded data are encapsulated and combined by adding the user identification of the first identification digit, the charging pile number of the second identification digit, and the check code of the third check digit to form a charging data frame; The charging data frame is sent to the cloud server in a multi-path concurrent manner through the target transmission path set.

[0034] Specifically, the vehicle's current State of Charge (SOC) value is collected in real time via the charging pile's internal battery management interface. This SOC value is compared against a preset threshold range to determine the accuracy control zone to which it belongs. This range is divided based on the sensitivity of the charging state to accuracy requirements and includes three segments: low charge (SOC less than 20%), medium charge (SOC between 20% and 80%), and high charge (SOC above 80%). The low charge range is extremely sensitive to fluctuations in remaining charge, so the system uses a 12-bit SOC encoding; the medium charge range uses 10 bits; and the high charge range uses only 8 bits due to its stable voltage and minimal fluctuations, enabling dynamic bit width allocation based on demand. The system then calls the charging pile's power monitoring unit to collect real-time charging power data and classifies it into predefined charging power levels based on this data. These power levels are divided into multiple segments, for example, 0-30kW for low power, 30-80kW for medium power, and 80-120kW for high power. Each power segment is matched to a quantization precision setting, which compresses the power values ​​using a specific step size. The low-power segment uses a smaller step size, such as 0.1kW, the medium-power segment uses 0.5kW, and the high-power segment uses 1kW. This ensures a balance between accuracy and code length. The compression result is converted to an unsigned fixed-point format to generate compressed power encoded data. The bit width of this encoded result can be set uniformly or adaptively adjusted based on the compression segment. The charging data frame structure is constructed. The frame structure consists of five basic segments: user identification segment, charging pile number segment, SOC data segment, power data segment, and checksum segment. The user identification segment is used to identify the vehicle-account binding relationship and is represented using 16 bits. The charging pile number segment uniquely identifies the charging pile and is configured as an 8-bit or 12-bit field. The SOC data segment and power data segment respectively carry the encoding results generated above. The checksum segment uses a CRC-16 or CRC-32 checksum algorithm to calculate all the aforementioned fields and then fill in the checksum to ensure that the data frame is correctly recognized by the error detection module during transmission. The data frame structure has a fixed frame header and tail control bytes, and supports the insertion of timestamp-based extension fields to enhance timing consistency and traceability during transmission. After completing the data frame construction, the system starts the multi-path concurrent transmission mechanism based on the target transmission path set stored in the current path scheduling module. This set consists of WiFi channels, 5G channels, and wired Ethernet paths that have been previously screened and sorted through the network environment database, including the top three communication paths with the highest scores. The system copies the same data frame three times and transmits them concurrently to the cloud server. The concurrent mechanism adopts asynchronous non-blocking transmission control. Each path is managed by an independent task thread and has its own retransmission strategy and confirmation mechanism to ensure that even if a path is delayed or interrupted, the remaining paths can complete data transmission within the specified time window, thereby improving the redundancy and reliability of data transmission.After receiving charging data frames from different paths, the cloud server merges and deduplicates the data using the timestamp, user ID, and pile number fields in the frame header. The first copy of the data that passes legal verification is retained as the valid record, and subsequent copies are automatically discarded or marked as retransmission results. This mechanism not only ensures data uniqueness and transmission efficiency, but also enhances the fault tolerance of the data uplink in high-interference environments such as parking garages, preventing data loss or status omissions at key nodes due to network quality fluctuations.

[0035] In a specific embodiment, the process of executing the step of sending the charging data frame to the cloud server in a multi-path concurrent manner through the target transmission path set may specifically include the following steps: The charging pile splits the charging data frame into several data slices according to the preset slice size, and adds a sequence identifier and a floor identifier to each data slice to form a slice data packet; Allocate fragmented data packets to WiFi paths, 5G paths, and wired network paths based on the transmission parameter configuration of the target transmission path set; Simultaneously, multiple first network paths obtained after initialization of the second candidate path in the target transmission path set are started to execute concurrent sending processes, wherein WiFi paths among the multiple first network paths carry a first proportion of data slices, 5G paths among the multiple first network paths carry a second proportion of data slices, and wired network paths among the multiple first network paths carry a third proportion of data slices; The charging pile monitors the failure probability of each first network path in real time, and automatically switches the unsent fragments of the first network path to the second network path for continued transmission based on the failure probability; When the cloud server receives all the fragmented data packets of the same charging data frame, the charging pile receives a transmission completion confirmation signal from the cloud server and clears the locally cached fragmented data packets and path status records.

[0036] Specifically, the charging station splits the charging data frame into several data segments according to a preset segment size. Each segment contains a portion of the original data payload and is appended with two structural identification fields: a segment sequence identifier and a floor identifier. The segment sequence identifier uses a consecutive, ascending number to mark the position of each data segment within the entire data frame for reassembly at the receiving end. The floor identifier specifies which floor of the garage the data frame originates from, facilitating spatial data aggregation, classification, and scheduling analysis by the cloud server. These identifiers are added to the segmented data packet as header fields, forming a data segment set. The path adaptation engine reads the transmission parameter configurations of all members in the target transmission path set. These parameters include each path's maximum bandwidth, current available bandwidth ratio, round-trip delay, congestion factor, packet loss rate, and path scoring weight. Based on these parameters, a weighted ratio is assigned to each path type. For example, a WiFi path, due to its ample bandwidth and low latency, is assigned a first weighting (e.g., 50%); a 5G path, due to its moderate latency and high interference immunity, is assigned a second weighting (e.g., 30%); and a wired network path, due to its high stability and low jitter, is assigned a third weighting (e.g., 20%). Based on this allocation ratio, the system divides all data shards to be sent into three categories. Each category of shards is associated with a set of designated paths, thereby achieving distributed load scheduling. The system also launches a separate sending process for each path group. Each sending process, defined as a thread pool or asynchronous event loop, is responsible for sequentially delivering all shards in its group to its corresponding primary network path channel. Within the same path category, multiple physical channels are dynamically distributed using a send queue using a round-robin or minimum queuing delay strategy. This concurrent process is non-blocking and interleaved. The system uses a transmission buffer to temporarily store in-flight but unacknowledged shards. A path status monitoring module records the transmission success rate and failure probability of each path in real time. The failure probability is calculated by periodically counting metrics such as packet loss, retransmission counts, and delay variation within a specific time window. When the failure probability of any primary network path exceeds a threshold (e.g., 10%), the path is deemed to no longer have stable transmission capabilities, triggering a failover mechanism. This mechanism transfers the list of fragment records that have not been delivered on this path to the available second network path pool, and then redistributes the sending tasks of these fragments according to the real-time transmission pressure and path redundancy priority, and ensures that the fragment sequence number is not changed. During the path switching process, soft interrupt technology is used to prevent the current session state from being lost, ensuring a smooth transition of data transfer. After the entire fragment set is transmitted, the cloud server parses, verifies and reassembles all the received data fragments. The server reconstructs the original data frame based on the sequence identifier in the fragment data packet, and performs CRC verification and structural integrity verification to confirm that all fragments have arrived and are correct.If verification succeeds, the cloud server generates a transmission completion confirmation signal containing the original data frame number, the reassembly completion status flag, and the confirmation timestamp, and returns it to the source charging station via the optimal available channel marked in the original path. When the charging station receives this confirmation signal, it performs a data cleanup process, releasing all fragment copies of the corresponding data frame, sequence identification records, path status tables, and related transmission history logs in the local cache to free up local memory space and transmission task resources. If the confirmation signal is not received within the set time limit, the system triggers the timeout retransmission logic and selectively reschedules the lost fragments until success or the maximum number of attempts is reached.

[0037] In a specific embodiment, the process of executing the step of the charging pile monitoring the failure probability of each first network path in real time and automatically switching the unsent fragments of the first network path to the second network path for continued transmission based on the failure probability may specifically include the following steps: The charging pile establishes a path quality attenuation prediction model based on the vehicle moving load changes between floors of the stereo garage and the network signal attenuation law, and uses the path quality attenuation prediction model to predict the failure probability of each first network path; Re-order the priority of the fragments to be sent on each first network path according to the failure probability to obtain a high-priority switching object; Adaptively re-encode the high-priority switching object according to the bandwidth characteristics and delay characteristics of the second network path to generate path-adapted slice data; When the actual switching is triggered, the unsent fragments of the first network path are automatically switched to the second network path for continued transmission based on the path switching decision algorithm, and the parameter weights of the path quality attenuation prediction model are updated.

[0038] Specifically, a path quality degradation prediction model covering the spatial structure of a multi-story parking garage is established. This model, by modeling the coupling relationship between inter-floor vehicle load and the network propagation environment, predicts the temporal evolution of channel quality for each path in real time. Vehicle movement up and down floors in a multi-story parking garage introduces significant spatial occlusion and reflection disturbances from metal objects. This variability can have a time-dependent impact on wireless communication signals, particularly those for Wi-Fi and 5G paths, resulting in multipath fading, interference overlap, and reflection shielding. The system collects inter-floor vehicle migration data using geomagnetic sensors and visual flow detectors deployed at entrances, ramps, and berths. Causal modeling is then performed based on historical signal strength curves, packet loss trajectories, and RTT fluctuation trends for each network path. This model constructs a set of time-series prediction models that describe the degradation behavior of path channels as load fluctuates. The model uses a sliding window regression algorithm, time-weighted exponential filtering, or a decay state prediction method based on a hidden Markov process to estimate the future short-term signal quality of each primary network path. The failure probability of the current path is derived using metrics such as the predicted signal attenuation rate, packet loss growth rate, and delay expansion rate. Based on failure probability, the system re-prioritizes the pending segments on each primary network path. This re-prioritization strategy follows the scheduling principle of "the higher the risk of path failure, the higher the task processing priority." This means that when a potential outage occurs on the current path, unsent data tasks are prioritized to reduce retransmission costs and the impact of transmission disruptions on the overall system. To achieve this, the system adds a path failure risk weight field to the segment scheduling queue. Each segment is bound to its corresponding upstream path and sorted by failure probability. A re-priority index is constructed, combining the segment's original sending priority with its content importance tag (e.g., SOC data takes precedence over billing data). This index is then used to output a set of high-priority switching targets—the most urgent and most important pending data segments to be transferred. High-priority switching targets are adaptively re-encoded based on the bandwidth and latency characteristics of the secondary network path to ensure that the migrating segments maximize their transmission performance on the new path. This process includes operations such as encoding bit width adjustment, compression level adaptation, and protocol format conversion. For example, when the original fragment was designed for a high-rate WiFi channel but needs to be switched to a bandwidth-constrained 5G channel, the system automatically adjusts the encoding structure, reducing the proportion of redundant fields, introducing differential encoding, or using a more efficient compression dictionary. Furthermore, a sliding send window mechanism is used to control the cadence based on the latency characteristics of the second path, thereby reducing the risk of sudden congestion and cumulative packet loss on the new path. This encoding reconstruction process is completed by the encoding adaptation engine, which converts adaptation instructions based on the different channel performance thresholds stored in the path configuration parameter library, and outputs the re-encoded path-adapted data fragment.When a handover event is triggered, the system uses a path switching decision algorithm based on real-time status parameters of the current path (such as the consecutive failure count, packet loss rate spike threshold, and link idleness percentage) as input, along with trigger conditions such as a failure probability exceeding 0.6 or a number of consecutive retransmissions exceeding five. When these conditions are met, the system determines that the first network path has become unavailable and migrates its unfinished data fragment transmission tasks to the second available path for continued transmission. During this migration, the system does not change the data frame number or fragment sequence number; it only updates the path binding and reestablishes the transmission tasks at the new path interface layer. This switching process ensures transaction integrity, preventing frame crossover or out-of-order delivery during the path switch. Furthermore, to continuously improve the adaptability of the prediction model, the system feeds the actual parameters from the path switch event back into the path quality degradation prediction model, updating the weighting parameters and path environment response function involved, thereby optimizing the model's response accuracy for specific path and floor structure combinations.

[0039] In this embodiment, the charging pile establishes a path quality attenuation prediction model based on the vehicle movement load changes and network signal attenuation rules between floors of the stereo garage, specifically including: the charging pile collects vehicle weight distribution data and parking density change information on each floor in real time through floor load sensors and parking space occupancy detectors, and monitors the elevator operation frequency and vehicle in and out flow, establishes a vehicle load dynamic change database, and performs time series correlation analysis on the load change moment and the corresponding network signal strength fluctuation; based on the electromagnetic wave propagation characteristics of the reinforced concrete structure of the stereo garage, the multipath reflection effect caused by the vehicle metal body and the floor deformation caused by the load increase are quantitatively modeled and processed, and the load-signal attenuation correlation function is established by analyzing the WiFi signal penetration loss, 4G signal diffraction attenuation and the loose probability of the wired network connector under different load conditions; the load-signal attenuation correlation function is subjected to machine learning training processing with the historical network quality change data, and the long short-term memory neural network algorithm is used to learn the load transmission between floors. Based on the law of transmission and signal attenuation propagation mode, a time series prediction model is constructed to predict the changes in the quality of each network path in the future time window; according to the vehicle flow peak period and the periodic law of load change in the daily operation of the stereo garage, the time series prediction model is calibrated in different time periods, and differentiated prediction parameter sets for four time periods: morning peak, noon, evening peak and night are established, and the model weight coefficient is dynamically adjusted through the Bayesian optimization algorithm; a prediction accuracy evaluation mechanism is set up to analyze the deviation between the network quality change trend predicted by the model and the actual monitoring data. When the prediction error exceeds the threshold, the model retraining process is automatically triggered, and the number of neural network layers and node parameters are adjusted according to the error distribution characteristics; an adaptive update strategy for the prediction model is established. When the structural modification, equipment upgrade or environmental change of the stereo garage is detected, new load-signal attenuation sample data is automatically collected for incremental learning processing to ensure that the prediction model always adapts to the current garage environment characteristics, and the updated model parameters are synchronized to the charging piles on each floor.

[0040] In a specific embodiment, the process of executing step 400 may specifically include the following steps: The cloud server spatially classifies the received charging data frames according to the floor identifiers in the charging data frames and establishes a charging state matrix; Perform inter-floor coupling analysis on the power demand data of each floor in the charging state matrix to obtain the corrected charging load value; Based on the corrected charging load value, combined with the historical traffic flow pattern and charging behavior trajectory data of the three-dimensional parking garage, the charging demand change trend of each floor is predicted; Based on the overall power capacity limit of the multi-story parking garage and the carrying capacity of the distribution lines on each floor, power constraints are optimized for the changing trend of charging demand to generate the charging load distribution status of the entire garage.

[0041] Specifically, the cloud server performs a structural analysis operation on each charging data frame, which includes extracting key fields such as timestamp, charging pile number, user ID, SOC value, power data, floor ID, etc. The server uses the floor ID field as the classification basis, groups all received data frames according to the floor to which they belong, and maps the grouped data into a structured space-time two-dimensional matrix based on a unified time axis to form a charging state matrix. The matrix uses the floor number as the row index and the sampling time period as the column index. Each matrix unit is filled with the total power request value or average charging load value of all charging piles on the floor at the corresponding moment, forming a charging behavior heat map covering the entire stereo garage. The inter-floor coupling relationship analysis is performed on the power demand data of each floor in the charging state matrix to solve the problem of inter-floor load disturbance coupling caused by vehicle flow, user selection preferences, power regulation behavior, etc. Considering the mobility and fluctuating correlation of charging demand between adjacent floors in a multi-story parking garage—for example, a vehicle moving from the first floor to the second floor and starting charging will cause a sudden drop in load on the first floor and a sudden increase on the second—the system uses a time-correlation coefficient matrix and spatial covariance analysis to model the power variation trends between floors in the charging state matrix. A correction factor based on multivariate regression is then introduced to adjust the original load values ​​in the matrix. This correction process eliminates isolated outliers caused by localized sudden behavior. By modeling the inter-floor influence path structure, the power values ​​on each floor more accurately reflect the actual load level under the coordinated overall behavior, thereby outputting a corrected charging load matrix that incorporates structural coupling factors. Based on these corrected charging load values, a multidimensional historical behavior feature vector set is constructed, combining historical parking pattern and charging behavior trajectory data from the multi-story garage, including peak-hour inbound vehicle density distribution, floor parking time statistics, charging frequency temporal distribution, and user charging station location preference models. By inputting the modified load matrix and the historical vector set into a time series prediction model based on a recurrent neural network or a long short-term memory network, a predicted charging load change curve for each floor within several future time windows is output. The trend direction (increase, plateau, decay) and degree of variation (stability, amplitude of fluctuation) are simultaneously evaluated, forming a three-dimensional prediction structure covering both spatial and temporal dimensions. After obtaining the predicted trend, the optimal scheduling phase further considers the physical constraints of the parking garage's power supply. This phase comprehensively considers the overall garage transformer capacity, the maximum power of the main power supply circuit, and the load-carrying capacity of branch power cables and protective devices on each floor. The system defines the overall power supply capacity of the parking garage as the total available power upper limit, and defines the upper limit of power allocation and the instantaneous power adjustment rate constraint for each floor. A regulation cost function is also introduced to quantify the energy loss caused by rapid load transfers and increased user waiting time. Based on these constraints, a multi-objective power constraint optimization model is constructed. The objective function is to minimize the deviation between the predicted load and the physical boundary. Power is redistributed among the floors using a linear programming or greedy scheduling algorithm.A weighting function is introduced into the scheduling process to dynamically adjust resource constraints and user waiting risks on different floors, shifting resources toward floors experiencing significant load surges while maintaining global scheduling balance. The final output is a charging load distribution across the entire storage system that conforms to power boundaries, considers load migration and user behavior predictions, and comprehensively optimizes energy efficiency and service quality.

[0042] In this embodiment, based on the influence of the gravity load of vehicles in the stereo garage on the floor structure, the power demand data of each floor in the three-dimensional charging state matrix is ​​analyzed and processed for the inter-floor coupling relationship, and a modified charging load value considering the influence of structural bearing is calculated, including: real-time collection of vehicle gravity load distribution data through the load-bearing sensor array on each floor of the stereo garage, combined with the vehicle weight information obtained by the vehicle type recognition system, to establish a floor load density distribution diagram, and calculate the deflection and stress distribution state of each floor under different load conditions according to the elastic deformation theory of reinforced concrete structure; based on the structural continuity and load transfer path between the floors of the stereo garage, a multi-floor coupling mechanical model is established, and the additional stress and deformation influence of the upper vehicle load on the lower floor is calculated by the finite element analysis method, and the inter-floor load influence coefficient matrix is ​​constructed to quantify the mechanical coupling relationship between the floors; the power demand data in the three-dimensional charging state matrix and the inter-floor load influence coefficient matrix are subjected to tensor operation processing, and the influence of floor deformation on charging pile installation is analyzed. The power output correction coefficient of each charging pile under structural deformation conditions is calculated by taking into account the influence of the inclination of the installation foundation and the stability of the electrical connection; according to the creep effect of the concrete floor of the stereo garage and the structural performance degradation law under long-term load, the power output correction coefficient is adjusted with time correlation, and a dynamic correction model considering structural aging and fatigue accumulation is established to generate time-varying floor coupling influence parameters; a multi-objective optimization algorithm is used to perform constraint correction operations on the original power demand data of each floor, with maximizing the structural safety margin and optimizing the charging efficiency as the objective functions, while considering the load balance constraints and deformation control constraints between floors, and calculating the corrected charging load values ​​of each floor that meet the structural safety requirements; a structural response verification mechanism is established, and the load conditions corresponding to the corrected charging load values ​​are input into the stereo garage structure simulation model for safety verification. When the predicted maximum deformation exceeds the allowable value of the structural design, the charging load limit of the corresponding floor is automatically reduced, and the safety constraint parameters of the inter-floor coupling relationship analysis model are updated.

[0043] In a specific embodiment, the process of executing step 500 may specifically include the following steps: The cloud server models the vertical power flow of the parking garage based on the floor load weight and capacity margin analysis of the charging load distribution in the entire garage, and generates an inter-floor power coordination matrix. Perform spatiotemporal coupling of the inter-floor power coordination matrix to form a dynamic power redistribution strategy; The dynamic power redistribution strategy is decomposed according to the geographical location and electrical connection topology of the charging piles to obtain the power allocation plan for the charging piles on each floor; The power allocation plan is differentially packaged and encoded according to the quality status of the target transmission path set on each floor to form a scheduling instruction, and the scheduling instruction is concurrently sent to the charging piles on the corresponding floor through the optimal path of each floor recorded in the network environment database.

[0044] Specifically, based on the overall charging load distribution, the power demand value for each floor is extracted, and the load weight and capacity margin for each floor are calculated. The load weight refers to the floor's proportion of the total load, while the capacity margin represents the remaining adjustable headroom between the upper limit of the electrical branch's load capacity and the current demand. Together, these two factors form the fundamental indicators for evaluating power dispatch capability. The cloud-based dispatch core module uses these as input to construct a graph model with floors as nodes and power transfer relationships as edges. This model represents the possibility and cost of directed power flow between floors, and builds a vertical power flow model based on this graph. This model analyzes the actual connectivity between floors (i.e., the presence of a shared bus, reversible power lines, or branching structures under the same power supply unit) and the upstream and downstream current switching capabilities to extract a feasibility matrix for power transfer channels between floors. The model then constructs an inter-floor power coordination matrix based on parameters such as power difference, capacity bounds, cable load capacity, and switch transition times. Each element in the matrix represents the maximum power transfer capacity that a floor can transfer to or from another floor. The system then extends the inter-floor power coordination matrix by introducing a time dimension, forming a dispatch modeling framework with spatiotemporal coupling characteristics. This process incorporates dynamic indicators such as time slice sequences, floor load change rates, user behavior trend curves, and burst load probability distributions. It dynamically adjusts the weights of power coordination capabilities based on a multi-period sliding window, forming a dynamic power redistribution strategy generator with dual inputs: the feasibility of power flow in the current period and future load change trends. The strategy generator uses a greedy allocation or linear programming algorithm to output the adjusted power quota and corresponding target allocation path for each floor within each future control cycle. The dynamic power redistribution strategy is decomposed based on the charging pile's geographic location and electrical connection topology, converting the floor-level power allocation results into specific instructions accurate to individual charging piles. The system accesses a data table of charging pile location information and cable connection topology. Based on the actual connection relationships of each charging pile within the branch network, the system allocates floor power values ​​proportionally to each branch's resistance, cable load capacity, and predicted power demand at the charging pile location. This process balances electrical stability and task balance, ensuring that different charging stations share a reasonable amount of power regulation, ultimately forming a final power allocation plan. This plan uses the charging station number as an index field and records the target power setpoint, execution timestamp, and a flag indicating whether the instruction can be delayed. The dispatch control system encapsulates this power allocation plan based on the actual communication conditions on each floor. To achieve path adaptation and self-adjustment of instruction redundancy, the system calls the network environment database to determine the quality status of each floor's current target transmission path set and determines the transmission capabilities of its WiFi, 5G, and Ethernet channels.For floors with excellent network quality, a standard encapsulation format is used to compress transmission content and improve efficiency. For floors with channel interference or high packet loss risk, a highly redundant encapsulation strategy is employed, introducing a redundant structure with multiple copies of instructions or splitting instructions into multiple small packets for batch delivery. Transmission priority tags and timeout retransmission strategies are also implemented. All scheduling instructions are uniformly encoded as structured data packets, containing the charging station ID, power command value, checksum field, routing priority flags, and failure handling logic. The system's path distribution module invokes the optimal path interface for each floor maintained in the network environment database and concurrently sends the encapsulated scheduling instructions to the corresponding floors. Task packets on each path are managed by a dedicated thread, and their status is recorded in real time to support confirmation feedback and scheduling rollback mechanisms. Upon receiving the scheduling instructions, the charging stations on each floor adjust their power output based on the protocol parsing content and provide feedback on the execution status. The cloud then uses this information to adjust the path channel availability score and update the path selection model for the next cycle.

[0045] In this embodiment, the cloud server performs vertical power flow modeling of the stereo garage based on the floor load weight and capacity margin analysis in the whole-garage charging load distribution, calculates the power support capacity and load transfer potential between floors, and generates an inter-floor power coordination matrix, including: the cloud server extracts the load weight coefficient and capacity margin data of each floor in the whole-garage charging load distribution, combines the bus connection topology and transformer capacity distribution of the vertical power distribution system of the stereo garage, establishes the floor power supply and demand state vector, and calculates the power margin and power gap values ​​of each floor at the current moment; based on the impedance characteristics and power transmission loss law of the vertical power distribution network of the stereo garage, performs network flow analysis on the power transmission path between floors, and establishes the inter-floor power transmission cost matrix considering transmission loss by calculating the line impedance, voltage drop and transmission efficiency between different floors; takes the charging pile distribution density, cable cross-sectional area and distribution equipment capacity of each floor as constraints, and adopts graph theory algorithm to coordinate the power supply and demand of the stereo garage. The connectivity analysis and processing of the power network of the stereo garage is carried out to identify the direct connection paths and multi-hop transmission paths between floors, and an inter-floor power connectivity assessment model is constructed. According to the load variation pattern of the stereo garage in different time periods and the imbalanced power consumption characteristics between floors, the power transmission cost matrix is ​​time-weighted. By analyzing the inter-floor power mutual assistance pattern in the historical load data, a dynamic power flow prediction algorithm based on time series characteristics is established. The minimum cost maximum flow algorithm is used to perform global optimization calculation on the vertical power network of the stereo garage. With the optimization goals of minimizing total transmission loss and maximizing load balance, the optimal power allocation strategy between floors is solved, and the safety constraints of distribution equipment and voltage stability constraints are taken into account. The optimal power allocation strategy is converted into the inter-floor power coordination matrix format, and the matrix elements represent the power support capacity and priority weight from floor i to floor j. At the same time, a dynamic update mechanism for the matrix elements is established. When it is detected that the load status change of a floor exceeds the threshold, the coordination matrix parameters are automatically recalculated.

[0046] Among them, based on the operation mode of the elevator in the stereo garage and the distribution of vehicle parking density, the power coordination matrix between floors is optimized by time-space coupling operation. When a power demand peak occurs on a certain floor, the load sharing ratio of the adjacent floors is automatically calculated to form a dynamic power redistribution strategy, including: the cloud server collects the operation status data of the elevator in the stereo garage in real time, including the lifting frequency, the distribution of the parking floors, the load change and the operation time mode, and combines the vehicle parking density information fed back by the parking sensors on each floor to establish a time-space correlation model between the elevator operation trajectory and the vehicle distribution on the floor, and identify the concentrated parking period and floor preference pattern of vehicles; according to the instantaneous impact effect of the elevator operation on the power system of the stereo garage, the power fluctuation at the start and stop of the elevator is quantitatively analyzed and processed, and by monitoring the starting current and braking feedback power of the elevator motor, the dynamic impact coefficient of the elevator operation on the available power of the charging piles on each floor is calculated, and an elevator-charging power coupling interference model is established; the vehicle parking density distribution data is converted into a floor charging demand prediction vector, and by analyzing the vehicle arrival time, expected stay time and charging The probability of electricity demand is determined by using a Markov chain model to predict the charging pile occupancy rate and power demand change trend of each floor within the time window, and a floor charging demand time series prediction matrix is ​​generated. The inter-floor power coordination matrix is ​​weighted and corrected according to the elevator operation status. When the elevator runs to a certain floor, the power support capability weight of that floor is reduced. When the elevator is in an idle state, the power transmission capacity between floors is increased, and a dynamic coordination matrix considering the elevator operation constraints is constructed. A time-space coupling optimization algorithm is used to jointly solve the dynamic coordination matrix and the floor charging demand time series prediction matrix. With minimizing the power imbalance between floors and maximizing the elevator operation efficiency as the dual objective function, the particle swarm optimization algorithm is used to solve the optimal power allocation ratio of each floor at different times. A real-time adjustment mechanism for the dynamic power redistribution strategy is established. When a sudden increase in power demand on a certain floor or a change in the elevator operation mode is detected, the strategy recalculation process is automatically triggered, and the prediction parameters and coordination weights are updated through a sliding time window to ensure that the power redistribution strategy always adapts to the real-time operation status of the stereo parking garage.

[0047] The above describes the charging station cloud operation and maintenance management method according to the embodiment of the present invention. The following describes the charging station cloud operation and maintenance management device according to the embodiment of the present invention. Figure 2 In one embodiment of the present invention, a cloud-based operation and maintenance management device for a charging station includes: The real-time monitoring module 11 is used to monitor the WiFi signal strength, 5G network quality and wired network delay parameters of each floor of the stereo garage in real time and establish a network environment database; The screening module 12 is used to screen out the target transmission path set with the best signal quality on the current floor from the network environment database when the charging pile detects that the SOC change amplitude exceeds a set threshold; The concurrent sending module 13 is used to perform differential bit encoding on the collected charging data according to the current SOC range and charging power level of the vehicle, generate charging data frames, and send the charging data frames to the cloud server in a multi-path concurrent manner through the target transmission path set; The analysis module 14 is used for the cloud server to receive and summarize the charging data frames from different floors of the stereo garage and analyze the charging load distribution status of the entire garage; The calculation module 15 is used to calculate the power allocation plan of the charging piles on each floor based on the charging load distribution status of the entire warehouse, and send the scheduling instructions to the charging piles on the corresponding floor.

[0048] Through the collaborative efforts of these components, a hierarchical network environment database and a dynamic path selection algorithm enable intelligent switching and concurrent transmission between WiFi, 4G, and wired networks. This significantly enhances the reliability and stability of data transmission within the complex environment of a multi-story parking garage and effectively addresses the technical challenge of single-path transmission being prone to interruption. A differentiated bit encoding scheme based on SOC range and charging power level achieves an intelligent balance between data accuracy and transmission efficiency. While ensuring the accuracy of critical charging information, it effectively reduces data transmission load and optimizes network bandwidth utilization. A path quality degradation prediction model and adaptive recoding processing enable predictive detection of transmission failures and seamless switching, avoiding the risk of data loss associated with traditional passive switching and ensuring the continuity and integrity of charging data transmission. A three-dimensional charging state matrix constructed based on the structural characteristics of the multi-story parking garage and analysis of inter-floor coupling relationships enable precise analysis of charging load distribution within complex spatial environments. Vertical power flow modeling and a spatiotemporal coupling optimization algorithm enable coordinated scheduling and load balancing of charging stations on each floor, effectively improving the overall charging service capabilities and resource utilization efficiency of the multi-story parking garage. The established instruction execution status feedback mechanism and real-time correction and update functions realize the dynamic optimization and self-improvement of the scheduling strategy, ensuring the intelligence and adaptability of the charging station operation and maintenance management.

[0049] See also Figure 3 , Figure 3 This is a schematic block diagram of the structure of an electronic device 300 provided in an embodiment of the present invention. The electronic device 300 includes a processor 301 and a memory 302. The processor 301 and the memory 302 are connected via a device bus 303, wherein the memory 302 may include a non-volatile storage medium and an internal memory.

[0050] The non-volatile storage medium may store a computer program. The computer program includes program instructions, which, when executed by the processor 301, may cause the processor 301 to execute any of the above-mentioned charging station cloud operation and maintenance management methods.

[0051] The processor 301 is used to provide computing and control capabilities to support the operation of the entire electronic device 300 .

[0052] The internal memory provides an environment for the operation of the computer program in the non-volatile storage medium. When the computer program is executed by the processor 301, the processor 301 can execute any of the above-mentioned charging station cloud operation and maintenance management methods.

[0053] Those skilled in the art will understand that Figure 3 The structure shown in the figure is only a block diagram of a part of the structure related to the solution of the present invention, and does not constitute a limitation on the electronic device 300 involved in the solution of the present invention. The specific electronic device 300 may include more or fewer components than shown in the figure, or combine certain components, or have a different component arrangement.

[0054] It should be understood that the processor 301 may be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor may be a microprocessor or any conventional processor.

[0055] It should be noted that, those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working process of the electronic device 300 described above can refer to the corresponding process of the aforementioned charging station cloud operation and maintenance management method, and will not be repeated here.

[0056] Those skilled in the art will clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, devices and units described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0057] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the portion that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions for enabling an electronic device (which can be a personal computer, server, or network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes various media that can store program code, such as a USB flash drive, a mobile hard drive, a read-only memory (ROM), a random access memory (RAM), a magnetic disk, or an optical disk.

[0058] As described above, the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit the same. Although the present invention has been described in detail with reference to the above embodiments, those skilled in the art should understand that the technical solutions described in the above embodiments can still be modified, or some of the technical features thereof can be replaced by equivalents. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A charging station cloud operation and maintenance management method, characterized in that: include: Real-time monitoring of WiFi signal strength, 5G network quality, and wired network latency parameters on each floor of the parking garage to establish a network environment database; When the charging pile detects that the SOC change amplitude exceeds a set threshold, a target transmission path set with the best signal quality on the current floor is selected from the network environment database; Performing differentiated bit encoding on the collected charging data according to the current SOC range and charging power level of the vehicle to generate a charging data frame, and sending the charging data frame to the cloud server in a multi-path concurrent manner through the target transmission path set; The cloud server receives and aggregates charging data frames from different floors of the parking garage, and analyzes the charging load distribution of the entire garage. The power distribution plan for the charging piles on each floor is calculated based on the charging load distribution status of the entire warehouse, and the scheduling instructions are sent to the charging piles on the corresponding floor.

2. The charging station cloud operation and maintenance management method according to claim 1, characterized in that: The real-time monitoring of WiFi signal strength, 5G network quality, and wired network delay parameters on each floor of the stereo garage and the establishment of a network environment database include: Deploy network detection nodes on each floor of the stereo garage to collect WiFi signal strength, 5G network quality, and wired network delay parameters to form raw network parameter data; Classify and merge the original network parameter data according to floor number and network type, and calculate the comprehensive quality score of each network on each floor; When a change in vehicle load or electromagnetic interference fluctuation is detected in the stereo garage, the weight of the comprehensive quality score is adjusted and the priority ranking of the network paths on each floor is updated; Based on the priority sorting, the original network parameter data is stored in the corresponding database partition to establish a network environment database.

3. The charging station cloud operation and maintenance management method according to claim 1, characterized in that: When the charging pile detects that the SOC change amplitude exceeds a set threshold, the target transmission path set with the best signal quality on the current floor is screened out from the network environment database, including: The charging pile calculates the difference between the current SOC value and the last recorded SOC value in real time to obtain the SOC change range. When the SOC change range exceeds the set threshold, the path selection process is triggered; When triggering the path selection process, the WiFi, 5G and wired network path information of the corresponding floor are retrieved from the network environment database according to the floor identifier of the charging pile, and the set of available network paths on the current floor is obtained; Sorting the comprehensive quality score of each first candidate path in the set of available network paths on the current floor, and selecting the top three second candidate paths with the highest comprehensive quality scores; The first three second candidate paths are used as a target transmission path set, and the transmission parameter configuration of the target transmission path set is recorded.

4. The charging station cloud operation and maintenance management method according to claim 1, characterized in that: The method comprises: performing differentiated bit encoding on the collected charging data according to the current SOC range and charging power level of the vehicle to generate a charging data frame, and sending the charging data frame to the cloud server in a multi-path concurrent manner through the target transmission path set, including: The charging pile determines the range to which the current SOC value belongs and obtains the current SOC range of the vehicle; The charging power level is divided according to the power value detected in real time by the charging pile, and the current SOC value is encoded by bit allocation based on the current SOC range of the vehicle to form SOC encoding data; Numerically compressing the charging power level according to the quantization step size of the corresponding power segment to generate compressed power coded data; The SOC coded data and the compressed power coded data are encapsulated and combined by adding a user identifier of a first identification digit, a charging pile number of a second identification digit, and a check code of a third check digit to form a charging data frame; The charging data frame is sent to the cloud server in a multi-path concurrent manner through the target transmission path set.

5. The charging station cloud operation and maintenance management method according to claim 4, characterized in that: The sending of the charging data frame to the cloud server in a multi-path concurrent manner through the target transmission path set includes: The charging pile splits the charging data frame into several data slices according to the preset slice size, and adds a sequence identifier and a floor identifier to each data slice to form a slice data packet; Allocate the fragmented data packets to the WiFi path, the 5G path, and the wired network path according to the transmission parameter configuration of the target transmission path set; Simultaneously, multiple first network paths obtained after initialization of the second candidate path in the target transmission path set are started to execute concurrent sending processes, wherein WiFi paths among the multiple first network paths carry a first proportion of data slices, 5G paths among the multiple first network paths carry a second proportion of data slices, and wired network paths among the multiple first network paths carry a third proportion of data slices; The charging pile monitors the failure probability of each first network path in real time, and automatically switches the unsent fragments of the first network path to the second network path for continued transmission based on the failure probability; When the cloud server receives all the fragmented data packets of the same charging data frame, the charging pile receives a transmission completion confirmation signal from the cloud server and clears the locally cached fragmented data packets and path status records.

6. The charging station cloud operation and maintenance management method according to claim 5, characterized in that: The charging pile monitors the failure probability of each first network path in real time, and automatically switches the unsent fragments of the first network path to the second network path for continued transmission based on the failure probability, including: The charging pile establishes a path quality attenuation prediction model based on the vehicle moving load change between floors of the stereo garage and the network signal attenuation law, and predicts the failure probability of each first network path through the path quality attenuation prediction model; Reordering the priority of the to-be-sent fragments on each first network path according to the failure probability to obtain a high-priority switching object; Adaptively re-encoding the high-priority switching object according to the bandwidth characteristics and delay characteristics of the second network path to generate path-adapted slice data; When the actual switching is triggered, the unsent fragments of the first network path are automatically switched to the second network path for continued transmission based on the path switching decision algorithm, and the parameter weights of the path quality attenuation prediction model are updated.

7. The charging station cloud operation and maintenance management method according to claim 1, characterized in that: The cloud server receives and aggregates charging data frames from different floors of the parking garage and analyzes the charging load distribution of the entire garage, including: The cloud server spatially classifies the received charging data frames according to the floor identifiers in the charging data frames and establishes a charging state matrix; Performing an inter-floor coupling relationship analysis on the power demand data of each floor in the charging state matrix to obtain a corrected charging load value; Based on the corrected charging load value, combined with the historical traffic flow pattern and charging behavior trajectory data of the three-dimensional parking garage, the charging demand change trend of each floor is predicted; Based on the overall power capacity limit of the stereo garage and the carrying capacity of the distribution lines on each floor, the charging demand change trend is optimized with power constraints to generate the charging load distribution status of the entire garage.

8. The charging station cloud operation and maintenance management method according to claim 1, characterized in that: The method of calculating the power allocation scheme of the charging piles on each floor based on the charging load distribution of the entire warehouse and issuing the scheduling instructions to the charging piles on the corresponding floor includes: The cloud server performs vertical power flow modeling of the stereo garage based on the floor load weight and capacity margin analysis in the overall garage charging load distribution, and generates an inter-floor power coordination matrix; Performing spatiotemporal coupling on the inter-floor power coordination matrix to form a dynamic power redistribution strategy; Decomposing the dynamic power redistribution strategy according to the geographical location and electrical connection topology of the charging piles to obtain a power allocation plan for the charging piles on each floor; The power allocation scheme is differentially packaged and encoded according to the quality status of the target transmission path set on each floor to form a scheduling instruction, and the scheduling instruction is concurrently issued to the charging piles on the corresponding floor through the optimal path of each floor recorded in the network environment database.

9. A cloud-based operation and maintenance management device for a charging station, characterized in that: Used to execute the charging station cloud operation and maintenance management method according to any one of claims 1 to 8, the charging station cloud operation and maintenance management device comprises: A real-time monitoring module is used to monitor the WiFi signal strength, 5G network quality, and wired network latency parameters on each floor of the parking garage in real time, and to establish a network environment database. A screening module is configured to screen out a target transmission path set with the best signal quality on the current floor from the network environment database when the charging pile detects that the SOC change amplitude exceeds a set threshold; A concurrent sending module, configured to perform differentiated bit encoding on the collected charging data according to the current SOC range and charging power level of the vehicle, generate a charging data frame, and send the charging data frame to the cloud server in a multi-path concurrent manner through the target transmission path set; The parsing module is used by the cloud server to receive and aggregate charging data frames from different floors of the parking garage and parse the charging load distribution status of the entire garage; The calculation module is used to calculate the power distribution plan of the charging piles on each floor based on the charging load distribution status of the entire warehouse, and send the scheduling instructions to the charging piles on the corresponding floor.

10. An electronic device, characterized in that: The electronic device comprises: a memory and at least one processor, wherein instructions are stored in the memory; The at least one processor calls the instructions in the memory to enable the electronic device to execute the charging station cloud operation and maintenance management method according to any one of claims 1 to 8.

Images