Charging demand thermodynamic diagram generation method and system based on time sequence prediction

Abstract

The invention provides a charging demand thermodynamic diagram generation method and system based on time sequence prediction, and relates to the technical field of thermodynamic diagram generation, and the method comprises the steps: building a mapping relation between a network flow fluctuation rule and charging demand time sequence characteristics, and constructing a cross-domain correlation model through analyzing a data transmission quantity periodic change rule; the integrated piezoelectric vibration sensor identifies a specific vibration mode caused by the plugging action of the charging gun, and distinguishes the actual load operation state and the no-load false triggering event of the charging pile; deploying a reactive power compensation device at a power distribution side, inhibiting fluctuation of a power factor of a power grid, and generating a steady-state power supply environment; performing space-time correlation on an output result of the cross-domain correlation model and an identification result of an actual load operation state of the charging pile, and calculating a demand distribution density of the charging pile group; and synchronously fusing the characteristic parameters of the specific vibration mode and the elimination result of the no-load false triggering event, and generating a multi-dimensional thermodynamic diagram. According to the technical scheme provided by the invention, the accuracy of charging demand prediction and thermodynamic diagram generation is improved.

Description

Technical Field

[0001] The present application relates to the technical field of heat map generation, and in particular to a method and system for generating a charging demand heat map based on time series prediction. Background Art

[0002] As the penetration rate of electric vehicles within smart parks continues to increase, charging demand is becoming increasingly heterogeneous, with significant temporal and spatial variations. This is manifested in significant time-of-day concentrations of charging demand, large fluctuations in equipment utilization, and significant differences between peak and valley loads in the grid. To ensure the efficient operation of charging facilities and the stable operation of the grid, it is necessary to combine multi-dimensional data such as user behavior patterns, campus traffic flow, and weather conditions to predict the temporal and spatial distribution of charging demand in order to optimize charging resource scheduling. It is also necessary to integrate heterogeneous data sources such as charging pile operating status, user reservation data, and grid load parameters to overcome the limitations of traditional single data models and improve prediction accuracy. It is also necessary to balance charging demand with the grid's carrying capacity to avoid local overloads or resource waste, and to ensure that predictions match grid conditions.

[0003] Currently, a mainstream approach involves deep learning models based on joint spatiotemporal prediction, such as a prediction framework that integrates long-short-term memory (LSTM) networks with convolutional neural networks (CNNs). This approach analyzes historical charging data and the time series characteristics of campus traffic flow to construct a multidimensional input vector. Using LSTM to capture temporal dependencies and CNN to extract spatial distribution features, the approach then assigns prediction weights to different regions through an attention mechanism. For example, some studies have incorporated mapping rules between user charging behavior patterns and grid load periods to generate demand distribution weights, then incorporate grid load data to refine the prediction results. This approach has been piloted in some smart campuses and has partially achieved prediction of the spatiotemporal distribution of charging demand.

[0004] Although the above scheme has improved prediction accuracy, its core problem lies in the lack of a closed-loop coordination mechanism between physics, power grid and user behavior. Existing models usually treat user behavior data, power grid parameters and device status as independent inputs, and no coupling relationship is established. As a result, the prediction weights rely on static experience distribution and are difficult to adapt to scenario changes. The recognition capability of the physical layer operating status of the equipment (such as charging pile failure rate and false alarms of plug-in and unplug events) is not integrated, resulting in the prediction results containing noise, which affects the scheduling accuracy. The prediction model is not linked with the power grid regulation equipment and cannot correct the impact of power grid fluctuations on demand distribution, resulting in deviations between the prediction results and the actual carrying capacity of the power grid, which may cause local overload risks. Summary of the Invention

[0005] The present application provides a method and system for generating a charging demand heat map based on time series prediction, which is used to solve the problem of insufficient accuracy of charging demand prediction and heat map generation in the prior art.

[0006] In a first aspect, the present application provides a method for generating a charging demand heat map based on time series prediction, comprising: Establish a mapping relationship between network traffic fluctuation patterns and charging demand timing characteristics. Based on this mapping relationship, analyze the periodic variation pattern of data transmission volume of communication nodes in the area where the charging piles are located to build a cross-domain correlation model that reflects the coupling relationship between charging demand and network behavior. A piezoelectric vibration sensor is integrated into the charging pile support structure. Based on the amplitude differences of different frequency components in the charging pile's mechanical vibration signal, it identifies the specific vibration patterns caused by the plugging and unplugging of the charging gun, distinguishing between the actual load operation state of the charging pile and false triggering events of no-load. Deploy reactive power compensation devices on the distribution side. By adjusting the switching state of compensation capacitors, the fluctuation of grid power factor caused by the access of charging pile clusters is suppressed, and a steady-state power supply environment is generated that matches the charging demand prediction model. Performing spatiotemporal correlation between the output of the cross-domain correlation model and the identification result of the actual load operation status of the charging pile, and calculating the demand distribution density of the charging pile group in the target area in combination with the power parameters of the steady-state power supply environment; Based on the spatial gradient variation characteristics of the demand distribution density, the characteristic parameters of the specific vibration mode and the elimination results of the no-load false triggering event are synchronously integrated to generate a multidimensional heat map including the charging pile availability prediction and load carrying capacity. The chromaticity mapping rule of the multidimensional heat map is negatively correlated with the power factor compensation status of the power grid.

[0007] Optionally, the step of temporally and spatially correlating the output result of the cross-domain correlation model with the identification result of the actual load operation state of the charging pile, and calculating the demand distribution density of the charging pile group in the target area in combination with the power parameters of the steady-state power supply environment includes: Dividing the identification result of the actual load operation status of the charging pile into a discrete event sequence according to preset time slices, and associating the load status in each time slice in the discrete event sequence with the peak interval of the network traffic data in the corresponding time slice; weighting the load state in each time slice in the discrete event sequence according to the network traffic and charging demand coupling weight output by the cross-domain correlation model to obtain a weighted result; Establishing a constraint relationship between the power parameters of the steady-state power supply environment and the maximum instantaneous load capacity of the charging pile, performing capacity boundary correction on the weighted result based on the constraint relationship, and generating an upper limit of the available power capacity of the charging pile node; The upper limit of available power capacity is temporally and spatially superimposed with the peak interval of network traffic data, and the demand distribution density of the charging pile group in the target area is output through density field calculation. The spatial resolution of the density field is determined by the physical spacing of the charging pile nodes and the response delay time of the reactive compensation device.

[0008] Optionally, according to the spatial gradient variation characteristics of the demand distribution density, the characteristic parameters of the specific vibration mode and the elimination results of the no-load false triggering event are synchronously integrated to generate a multi-dimensional heat map including the charging pile availability prediction and load carrying capacity, including: Discretizing the spatial gradient variation characteristics of the demand distribution density into a plurality of local density variation rates according to the physical layout of the charging pile group, wherein the direction vectors of the local density variation rates are associated with the vibration propagation direction of the charging gun plugging and unplugging action; Extracting a vibration frequency distribution interval that matches the plugging and unplugging action of the charging gun from the characteristic parameters of the specific vibration mode, and generating a vibration characteristic availability probability that characterizes the availability status of the charging pile based on the energy proportion of each frequency band within the vibration frequency distribution interval; According to the result of eliminating the no-load false trigger event, the no-load false trigger frequency of the charging pile node within a preset time period is counted, and combined with the power parameters of the steady-state power supply environment, a no-load interference suppression factor reflecting the correction coefficient of the charging pile load carrying capacity is generated; The vibration feature availability probability and the no-load interference suppression factor are spatially interpolated according to the physical location of the charging pile, and the vector distribution of the local density change rate is superimposed to generate a multi-dimensional heat map including the charging pile availability prediction and load carrying capacity.

[0009] Optionally, establishing a mapping relationship between network traffic fluctuation patterns and charging demand timing characteristics, and constructing a cross-domain correlation model reflecting the coupling relationship between charging demand and network behavior by analyzing the periodic variation patterns of data transmission volume of communication nodes in the area where the charging piles are located, includes: Divide network traffic monitoring units according to the physical coverage of communication nodes in the area where the charging piles are located, perform multi-scale periodic decomposition of the data transmission volume of the communication nodes based on the communication protocol type, and extract the dominant periodic components of the hourly, daily, and weekly data transmission volume within each monitoring unit; Based on the temporal distribution characteristics of charging gun plug-in and unplug events in the historical operation data of charging piles, high-frequency demand periods, stable demand periods, and low-peak demand periods of charging demand are defined, and a mapping rule between the charging demand period label and the amplitude-frequency characteristics of the dominant periodic component is established; For each network traffic monitoring unit, the phase fluctuation interval of the dominant periodic component is matched with the duration window of the corresponding charging demand period label. By introducing the amplitude-frequency characteristic mapping rule into the matching process, a dynamic correlation weight between the network traffic fluctuation pattern and the charging demand timing characteristics is generated; Based on the association weight, a coupling relationship function between the charging demand and network behavior is constructed. The timing offset of the dominant periodic component of the network traffic in the coupling relationship function is used to correct the cross-domain association error and output a structured cross-domain correlation model.

[0010] Optionally, the step of performing spatiotemporal superposition of the available power capacity upper limit and the network traffic data peak interval, and outputting the demand distribution density of the charging pile group in the target area through density field calculation, includes: Divide the target area into spatial grid units according to the physical spacing of the charging pile nodes, discretize and sample the upper limit of the available power capacity according to the spatial grid units, and generate a local power capacity distribution; Dividing the network traffic data peak interval into multiple continuous time windows according to the time dimension, extracting the spatial distribution characteristics of the network traffic data in each time window, and generating the network traffic spatial weight corresponding to the spatial grid unit; Superimposing the local power capacity distribution and the network traffic spatial weight according to the spatial grid unit to generate a density field reflecting the coupling relationship between the power capacity of the charging pile group and the network traffic; A multi-scale convolution operation is performed on the local power capacity distribution and the network traffic spatial weight according to the spatial grid unit, the density increment of the spatial grid unit in the density field is calculated, and the demand distribution density of the charging pile group in the target area is output.

[0011] Optionally, spatially interpolating the vibration feature availability probability and the no-load interference suppression factor according to the physical location of the charging pile, and superimposing the vector distribution of the local density change rate, to generate a multidimensional heat map including charging pile availability prediction and load carrying capacity, includes: Constructing a vibration feature interpolation field based on the directional characteristics of the physical location of the charging pile, wherein the interpolation weight of the vibration feature interpolation field is adjusted by the vibration propagation path energy attenuation in the vibration feature availability probability; A no-load interference correction field is constructed based on the spatial proximity relationship of the physical locations of the charging piles, wherein the correction weight of the no-load interference correction field is controlled by the response time of the grid reactive compensation device in the no-load interference suppression factor; Visualizing the vibration characteristic interpolation field and the no-load interference correction field according to the physical position of the charging pile to generate an initial thermal distribution; The vector distribution of the local density change rate is spatially superimposed on the initial thermal distribution to generate a multi-dimensional thermal map including charging pile availability prediction and load carrying capacity.

[0012] Optionally, a piezoelectric vibration sensor is integrated into the charging pile support structure to identify specific vibration modes caused by the plugging and unplugging of the charging gun based on the amplitude differences of different frequency components in the mechanical vibration signal of the charging pile, so as to distinguish the actual load operation state of the charging pile from the no-load false trigger event, including: Place piezoelectric vibration sensors in the stress concentration area of ​​the charging pile support structure to collect multi-band vibration components in the mechanical vibration signal of the charging pile; Extracting a characteristic frequency band from the multi-band vibration components that matches the plugging and unplugging action of the charging gun, and generating a vibration energy distribution of the charging pile based on the amplitude variation law of the characteristic frequency band; Identifying a specific vibration mode caused by the plugging and unplugging action of the charging gun based on a comparison result of a peak interval of the vibration energy distribution of the charging pile and a preset vibration energy threshold; Based on the spectrum distribution characteristics and energy fluctuation trends of the specific vibration mode, the actual load operation state of the charging pile and the no-load false triggering event are distinguished.

[0013] In a second aspect, the present application provides a system for generating a charging demand heat map based on time series prediction, comprising: The analysis module is used to establish a mapping relationship between network traffic fluctuation patterns and charging demand timing characteristics. By analyzing the periodic changes in data transmission volume of communication nodes in the area where the charging piles are located, a cross-domain correlation model that reflects the coupling relationship between charging demand and network behavior is constructed; The recognition module is used to integrate a piezoelectric vibration sensor within the charging pile support structure. Based on the amplitude differences of different frequency components in the charging pile's mechanical vibration signal, it identifies the specific vibration patterns caused by the plugging and unplugging of the charging gun, and distinguishes the actual load operation state of the charging pile from false triggering events of no-load; The suppression module is used to deploy reactive power compensation devices on the distribution side. By adjusting the switching state of compensation capacitors, it suppresses the fluctuation of grid power factor caused by the access of charging pile clusters, and generates a steady-state power supply environment that matches the charging demand prediction model; A calculation module, configured to perform spatiotemporal correlation between the output of the cross-domain correlation model and the identification result of the actual load operation state of the charging pile, and calculate the demand distribution density of the charging pile group in the target area in combination with the power parameters of the steady-state power supply environment; A generation module is used to synchronously integrate the characteristic parameters of the specific vibration mode and the elimination results of the no-load false triggering event based on the spatial gradient change characteristics of the demand distribution density, to generate a multidimensional heat map including charging pile availability prediction and load carrying capacity, wherein the chromaticity mapping rule of the multidimensional heat map is negatively correlated with the power factor compensation status of the power grid.

[0014] In a third aspect, an embodiment of the present application provides a computing device comprising a processing component and a storage component; the storage component stores one or more computer instructions; the one or more computer instructions are used to be called and executed by the processing component to implement a method for generating a charging demand heat map based on time series prediction as described in the first aspect above.

[0015] In a fourth aspect, an embodiment of the present application provides a computer storage medium storing a computer program. When the computer program is executed by a computer, it implements a method for generating a charging demand heat map based on time series prediction as described in the first aspect.

[0016] In the embodiment of the present application, a mapping relationship between the fluctuation law of network traffic and the timing characteristics of charging demand is established. By analyzing the periodic variation law of the data transmission volume of the communication nodes in the area where the charging piles are located, a cross-domain correlation model that reflects the coupling relationship between charging demand and network behavior is constructed; a piezoelectric vibration sensor is integrated inside the supporting structure of the charging pile, and based on the amplitude difference of different frequency components in the mechanical vibration signal of the charging pile, the specific vibration mode caused by the plugging and unplugging action of the charging gun is identified, and the actual load operation state of the charging pile is distinguished from the no-load false triggering event; a reactive compensation device is deployed on the distribution side, and the grid is suppressed due to the access of the charging pile cluster by adjusting the switching state of the compensation capacitor. The power factor fluctuates to generate a steady-state power supply environment that matches the charging demand prediction model; the output result of the cross-domain correlation model is temporally and spatially correlated with the identification result of the actual load operation status of the charging pile, and the demand distribution density of the charging pile group in the target area is calculated in combination with the power parameters of the steady-state power supply environment; according to the spatial gradient change characteristics of the demand distribution density, the characteristic parameters of the specific vibration mode and the elimination results of the no-load false triggering event are synchronously integrated to generate a multi-dimensional heat map including the charging pile availability prediction and load carrying capacity, and the chromaticity mapping rule of the multi-dimensional heat map is negatively correlated with the power factor compensation state of the power grid.

[0017] The technical solution of this application has the following beneficial effects: By analyzing the periodic variation patterns of data transmission volume at communication nodes, a mapping relationship between network behavior and charging demand is constructed, the benchmark modeling problem of collaborative prediction of multi-source data is solved, and the accuracy of charging demand time series prediction is improved. Based on the amplitude difference of the frequency components of the vibration signal, the actual load operation of the charging pile and the no-load false triggering events are accurately distinguished, the physical layer device status perception capability is enhanced, and the prediction noise interference is reduced. By adjusting the switching state of the compensation capacitor, the power factor fluctuation of the power grid is suppressed, providing a stable electrical environment for demand forecasting that matches the grid response capability. The cross-domain model output is temporally and spatially correlated with the load state identification results, and the demand distribution density is generated in combination with the grid power parameters to achieve the accuracy and spatial resolution optimization of the charging demand prediction. Based on the spatial gradient characteristics of the demand distribution density, the vibration sensor data and the no-load exclusion results are integrated to generate a visual heat map that includes availability prediction and load carrying capacity. Its color rule is linked to the grid compensation state to intuitively reflect the grid carrying pressure.

[0018] Furthermore, by dividing the actual load state of charging piles into discrete event sequences by time slices and associating them with the peak intervals of network traffic data, the load state is weighted based on the coupling weights output by the cross-domain model. The weighted results are then modified with the grid power parameters to generate the upper limit of the available power capacity of the charging pile nodes. Finally, through density field calculation, the available power capacity and the network traffic peak are superimposed in time and space to output the demand distribution density, whose spatial resolution is jointly constrained by the physical spacing of the charging piles and the grid response delay. Through the fusion of multi-source data (network traffic, load state, grid parameters) and a weight allocation mechanism, the refined calculation of the charging demand distribution density is achieved. In combination with the collaborative constraint rules of grid response delay and physical distance between devices, the prediction results are ensured to be consistent with the grid regulation capability and adapted to the actual layout of charging piles, significantly improving the prediction accuracy and grid compatibility.

[0019] These and other aspects of the present application will become more readily apparent from the description of the following embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, a brief introduction will be given below to the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without any creative work.

[0021] Figure 1 A flowchart of a method for generating a charging demand heat map based on time series prediction provided by the present application is shown; Figure 2 A schematic diagram of the structure of a charging demand heat map generation system based on time series prediction provided by the present application is shown; Figure 3 A schematic structural diagram of a computing device provided by the present application is shown. DETAILED DESCRIPTION

[0022] In order to enable those skilled in the art to better understand the solution of the present application, the technical solution in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application.

[0023] In some of the processes described in the specification and claims of this application and the above-mentioned figures, multiple operations that appear in a specific order are included, but it should be clearly understood that these operations may not be executed in the order in which they appear in this document or may be executed in parallel. The serial numbers of the operations, such as 101, 102, etc., are only used to distinguish between different operations, and the serial numbers themselves do not represent any order of execution. In addition, these processes may include more or fewer operations, and these operations may be executed in sequence or in parallel. It should be noted that the descriptions of "first", "second", etc. in this document are used to distinguish different messages, devices, modules, etc., and do not represent a sequential order, nor do they limit "first" and "second" to being different types.

[0024] The technical solution of this application is applicable to the scenario of collaborative prediction of charging demand and network load in smart parks, and realizes charging demand prediction through multi-source data collaboration and cross-system linkage. First, a mapping model is established between the periodic fluctuation of network traffic and the time series characteristics of charging demand to quantify the driving effect of network behavior on demand; the piezoelectric vibration sensor is combined to identify the specific vibration mode of the charging gun plugging and unplugging action (main frequency energy concentration, harmonic attenuation characteristics), and filter out no-load false triggering interference; the reactive compensation device is deployed simultaneously to adjust the capacitor switching to suppress the fluctuation of the power factor of the power grid and generate steady-state power supply parameters; the network traffic weight, load state identification results and grid capacity constraints are integrated in time and space to calculate the demand distribution density; finally, a multi-dimensional heat map is generated based on the density gradient, equipment availability probability and grid status. Its color is negatively correlated with the power factor compensation value, which intuitively reflects the regional load pressure and grid health.

[0025] The following will be combined with the drawings in the embodiments of this application to clearly and completely describe the technical solutions in the embodiments of this application. Obviously, the embodiments described are only part of the embodiments of this application, not all of the embodiments. Based on the embodiments in this application, all other embodiments obtained by those skilled in the art without making creative efforts are within the scope of protection of this application.

[0026] Figure 1 A flowchart of a method for generating a charging demand heat map based on time series prediction is provided for an embodiment of the present application. Figure 1 As shown, the method includes: 101. Establish a mapping relationship between network traffic fluctuation patterns and charging demand time series characteristics. Based on the mapping relationship, analyze the periodic variation patterns of data transmission volume of communication nodes in the area where the charging piles are located, so as to construct a cross-domain correlation model that reflects the coupling relationship between charging demand and network behavior. In this step, the cross-domain correlation model is a mathematical model that reflects the coupling relationship between network behavior and charging demand. It is established by analyzing the correlation between the network traffic data of the communication nodes in the area where the charging piles are located (such as the peak data transmission volume, the proportion of protocol types) and the timing characteristics of charging demand (such as the frequency of charging gun plug-in and unplugging events, and the power consumption curve). It is used to predict the demand fluctuation trend.

[0027] In an embodiment of the present application, based on the established mapping relationship between the network traffic fluctuation pattern and the charging demand timing characteristics, the network traffic data of the communication nodes in the area where the charging pile is located is subjected to multi-scale periodic decomposition to identify the periodic fluctuation patterns at the hourly, daily, and weekly levels. For example, the occurrence period and duration of the traffic peak are extracted through spectrum analysis. Then, combined with the time series distribution of the charging gun plug-in and unplugging events in the historical operation data of the charging pile, the high-frequency demand period (such as morning and evening peaks), the stable demand period, and the valley demand period are divided, and time period labels are established. The network traffic periodic component is matched with the charging demand period through a time alignment algorithm, and the similarity weights of the two at different time scales are calculated. In order to solve the problem of the timing offset between network traffic fluctuations and charging demand changes, based on the phase correlation characteristics of the mapping relationship, the phase synchronization technology is used to compensate for the time difference between the two, and finally the periodic component matching weights and the timing compensation parameters are integrated to generate a cross-domain correlation model.

[0028] Consider a smart campus where charging stations within the communication node coverage area experience peak network traffic (approximately 120 Gbps) between 6:00 PM and 8:00 PM daily, corresponding to a surge in charging demand during off-hours (charging plug-in and unplug frequency increases to 5 times per minute). Based on the mapping relationship between network traffic and charging demand time series, periodic decomposition reveals that the daily periodic component of network traffic remains high from 6:00 PM to 7:30 PM, while charging demand peaks between 6:30 PM and 8:00 PM. A time alignment algorithm calculates a similarity weight of 0.88 between the two. Based on the phase offset pattern in the mapping relationship, phase synchronization compensation detects that network traffic reflects charging demand changes 30 minutes in advance. The resulting cross-domain correlation model output shows a reduction in prediction error from 8% to 2.5%.

[0029] 102. Integrate a piezoelectric vibration sensor inside the charging pile support structure. Based on the amplitude differences of different frequency components in the mechanical vibration signal of the charging pile, it can identify the specific vibration mode caused by the plugging and unplugging of the charging gun, and distinguish the actual load operation state of the charging pile from the no-load false trigger event. In this step, the specific vibration pattern is the amplitude difference characteristics of different frequency components in the vibration signal caused by the plugging and unplugging of the charging gun, such as the main frequency concentration range and harmonic attenuation slope. The no-load false trigger event occurs when the charging pile is mistakenly identified as being in a loaded state due to non-genuine charging operations (such as human error or environmental interference).

[0030] In an embodiment of the present application, a piezoelectric vibration sensor is deployed in the stress concentration area of ​​the charging pile support structure to collect mechanical vibration signals. The amplitude spectrum of the 20-150Hz frequency band is extracted through frequency domain conversion to identify the main frequency range (such as 45-60Hz) and harmonic characteristics corresponding to the plugging and unplugging action of the charging gun. Based on a preset energy attenuation threshold (such as the frequency point when the fundamental frequency energy decays to 50%), a distinction is made between real charging events (smooth energy decay) and no-load false triggering events (rapid energy decay). Combined with the resonance characteristics of the mechanical structure of the charging pile (such as the fixed frequency response of the metal bracket), the judgment threshold is adjusted. For example, when the bracket screws are loose, the main frequency may offset by 5Hz, and the frequency matching range needs to be corrected accordingly. Finally, the classification results of the actual load operation status of the charging pile and the no-load false triggering event are output.

[0031] For example, continuing with the previous example, a vibration signal from a charging pile support structure was detected at 6:30 PM. Frequency domain analysis showed a dominant frequency of 50 Hz (the characteristic frequency band for charging gun plugging and unplugging), with harmonic energy attenuated to 35% of the fundamental frequency at 80 Hz, meeting the criteria for determining a true charging event (the threshold is set to attenuate to below 40% to indicate no-load). In another event, the dominant vibration frequency was 88 Hz (close to the fan vibration frequency band), and harmonic energy attenuated to 12% of the fundamental frequency at 100 Hz, triggering the elimination of a false no-load judgment. Taking into account the mechanical damping characteristics of loose bracket screws, the dominant frequency matching range was adjusted to 48-52 Hz, ultimately achieving a load state recognition accuracy of 98.5% and a reduction in the false positive rate to 1.3%.

[0032] 103. Deploy reactive power compensation devices on the distribution side to suppress grid power factor fluctuations caused by the access of charging pile clusters by adjusting the switching status of compensation capacitors, thereby generating a steady-state power supply environment that matches the charging demand prediction model. In this step, the reactive power compensation device is deployed on the distribution side and stabilizes the grid power factor by adjusting the switching state of the compensation capacitor. A steady-state power supply environment is achieved by suppressing grid power factor fluctuations and achieving a stable power supply that matches the charging demand forecast model.

[0033] In an embodiment of the present application, the power factor, voltage fluctuation rate, and reactive power demand of the power grid are monitored to predict the downward trend of the power factor caused by the access of the charging pile cluster (e.g., it may drop to 0.85 within the next 5 seconds). The reactive power to be compensated (e.g., 200kVar) is calculated based on the instantaneous load capacity of the charging pile, and the capacitor group switching combination is optimized (e.g., switching two groups of 100kVar capacitors). The stability of the compensated power grid is verified through closed-loop control to ensure that the power factor is restored to the target range (e.g., 0.95-0.98) and the voltage fluctuation rate is less than 3%. Finally, the steady-state power supply environment parameters are output, including the power factor compensation value, the voltage stability range, and the capacitor switching status.

[0034] For example, continuing with the previous example, the smart campus power distribution system detected a power factor drop to 0.84 at 6:45 PM (charging pile cluster load factor 85%), predicting a need for 220 kV reactive power compensation. The reactive power compensation device switched in two 100 kV capacitors and one 30 kV capacitor. After compensation, the power factor returned to 0.96, and the voltage fluctuation rate dropped from 5% to 1.8%. The resulting steady-state power supply parameters (power factor > 0.95, voltage fluctuation < 2%) matched the demand predicted by the cross-domain correlation model.

[0035] 104. Performing spatiotemporal correlation between the output of the cross-domain correlation model and the identification result of the actual load operation state of the charging pile, and calculating the demand distribution density of the charging pile group in the target area in combination with the power parameters of the steady-state power supply environment; In this step, the demand distribution density is a quantitative indicator of the intensity of charging demand of the charging pile group in the target area that changes over time and space, such as the number of charging requests per minute per unit area.

[0036] In this embodiment of the present application, the network traffic weight output by the cross-domain correlation model (e.g., a peak-hour weight of 0.9) is mapped to the spatial grid where the charging piles are located according to a time window (e.g., 5 minutes), forming a demand trend driven by network traffic. Combined with the identification results of the actual load status of the charging piles (e.g., 7 out of 10 charging piles in a grid are in a load state), the initial demand density (e.g., 1.4 vehicles / minute) is calculated. The power parameters of the steady-state power supply environment (e.g., the maximum instantaneous load capacity of the power grid is 90% of the rated value) are introduced, and the upper limit of the demand density is corrected by capacity constraints (e.g., from 1.5 vehicles / minute to 1.35 vehicles / minute). Finally, the spatial resolution is adjusted based on the physical spacing between the charging piles (e.g., 10 meters) and the grid response delay (e.g., 2 seconds), and the demand distribution density is output.

[0037] For example, continuing with the previous example, the network traffic weight (0.88) output by the cross-domain correlation model is correlated with the charging pile load status (70% load factor) on a 10-meter spatial grid. Combined with the grid capacity constraint factor (0.9), the initial demand density is calculated to be 1.6 vehicles / minute. After capacity correction, the output demand distribution density is 1.44 vehicles / minute. With the spatial resolution adjusted to a 10-meter grid, the peak density detected in Area A is 1.9 vehicles / minute (triggering an alarm), while the density in Area B is 0.8 vehicles / minute (idle recommendation).

[0038] 105. Based on the spatial gradient variation characteristics of the demand distribution density, the characteristic parameters of the specific vibration mode and the elimination results of the no-load false triggering event are synchronously integrated to generate a multidimensional heat map including the charging pile availability prediction and load carrying capacity. The chromaticity mapping rule of the multidimensional heat map is negatively correlated with the power factor compensation state of the power grid.

[0039] In this step, a multidimensional heat map is created, visualizing the predicted charging station availability (device status), load carrying capacity (grid constraints), and demand density (spatiotemporal distribution). The color mapping rule is that the color depth in the heat map is negatively correlated with the grid power factor compensation status. For example, dark red indicates a low power factor (<0.9), and light green indicates a high power factor (>0.95).

[0040] In this embodiment of the present application, the spatial gradient variation characteristics of the demand distribution density are mapped to the base color of the heat map (e.g., a red gradient indicates high and low density), and the probability of charging pile availability (e.g., a transparency of 30% indicates a high risk of equipment failure) and the results of no-load false trigger elimination (e.g., a gray area indicates low confidence) are superimposed. The color shift is adjusted based on the power factor compensation status of the power grid (e.g., the color shifts to purple at a power factor of 0.88), and a spatial interpolation algorithm is used to smooth the color transitions between adjacent areas. This ultimately generates an interactive multidimensional heat map, allowing operators to click to query specific parameters (e.g., the load rate of a charging pile or the switching status of the compensation capacitor).

[0041] For example, continuing with the above example, in the final generated multi-dimensional heat map, area A is displayed in dark red (demand density 1.9 vehicles / minute), with a transparency of 40% (vibration feature availability probability 60%) and a hue of purple (power factor 0.87), prompting operation and maintenance personnel to prioritize scheduling; area B is displayed in light green (demand density 0.8 vehicles / minute), with a transparency of 85% (vibration feature availability probability 95%) and a hue of positive green (power factor 0.96), recommending users to go charging; area C is marked in gray (similarity of no-load false trigger elimination results <0.6), shielding against interference from low-confidence data.

[0042] Steps 101-105 achieve global optimization of charging demand forecasting and grid operations through cross-domain data collaborative modeling (correlating network traffic with charging demand), closed-loop physical and grid sensing (vibration signature recognition and reactive power compensation linkage), and multi-dimensional visualization (heat map rendering). This overcomes the limitations of a single data source and integrates network behavior, device status, and grid parameters to significantly reduce forecast bias and misjudgment interference. It also suppresses power fluctuations caused by the integration of charging pile clusters, ensuring that forecast results match the grid's carrying capacity. Visual heat maps simultaneously reflect charging demand distribution, device health, and grid stress, providing actionable insights for resource scheduling.

[0043] In order to further improve the spatiotemporal accuracy of charging demand forecasting and ensure grid stability, this application solves the time-varying correlation problem between network behavior and charging demand based on the time slice event sequence division and weighting mechanism; introduces a grid capacity constraint correction model to break through the physical boundary limitations of traditional forecasting; generates high-resolution demand distribution through spatiotemporal superposition and density field calculation, and finally outputs actionable decision support. In some embodiments, the output results of the cross-domain correlation model are spatiotemporally correlated with the identification results of the actual load operating status of the charging pile, and combined with the power parameters of the steady-state power supply environment, the demand distribution density of the charging pile group in the target area is calculated, including: 201. Divide the identification result of the actual load operation status of the charging pile into a discrete event sequence according to preset time slices, and associate the load status in each time slice in the discrete event sequence with the peak interval of the network traffic data in the corresponding time slice; In step 201, the discrete event sequence is a discretized state sequence that divides the identification results of the actual load operating status of the charging pile into fixed time slices (e.g., 5 minutes). The charging pile is marked as loaded or unloaded within each time slice. The network traffic data peak interval is the peak data transmission range achieved by the communication node within a specific time slice.

[0044] In this embodiment of the present application, the actual load status identification results of the charging piles are divided into time slices (e.g., 5-minute windows). A sliding window is used to calculate the load status percentage within each time slice (e.g., 70% of charging piles are under load) to form a discrete event sequence. Network traffic data for the corresponding time slice is simultaneously extracted to identify its peak interval (e.g., traffic reaches 115 Gbps between 18:00 and 18:05). A wavelet transform is used to decompose the periodic components of the network traffic data. The Pearson correlation coefficient is used to calculate the correlation strength between the load status percentage and the traffic peak interval (e.g., a correlation coefficient of 0.85), forming a load-traffic correlation mapping table for the time slice dimension.

[0045] 202. Weighting the load state in each time slice in the discrete event sequence according to the network traffic and charging demand coupling weight output by the cross-domain correlation model to obtain a weighted result; In step 202 , weighting is performed to adjust the load state in the discrete event sequence according to the coupling weight of network traffic and charging demand output by the cross-domain correlation model.

[0046] In this embodiment, a time warping (DTW) algorithm is used to align the timing offsets of load event sequences and network traffic peaks based on the coupling weights output by the cross-domain correlation model (e.g., the mapping weights between the daily cycle components of network traffic and charging demand). The entropy weighting method is used to quantify the weight contributions of different time slices (e.g., the weight of peak hours is increased by 1.2 times), and a weighted correction is applied to the load state proportion (e.g., the original 70% load factor is weighted and adjusted to 78%). A sliding average filter is then introduced to eliminate short-term fluctuations to obtain a weighted result.

[0047] 203. Establish a constraint relationship between the power parameter of the steady-state power supply environment and the maximum instantaneous load capacity of the charging pile, perform capacity boundary correction on the weighted result based on the constraint relationship, and generate an upper limit of the available power capacity of the charging pile node; In step 203, the capacity boundary is modified to impose an upper limit constraint on the weighted load state according to the power parameters of the steady-state power supply environment (such as the maximum instantaneous load capacity of the power grid) to generate the upper limit of the available power capacity of the charging pile node.

[0048] In this embodiment of the present application, power parameters of a steady-state power supply environment (e.g., a power factor of 0.95 and a voltage fluctuation rate of <2%) are collected, and the maximum load capacity that the power grid can carry within the next 5 seconds (e.g., 90% of the rated capacity) is predicted using a Kalman filter. A linear constraint relationship is established between the total power demand of the charging pile group and the grid capacity, and a Lagrangian relaxation algorithm is used to perform boundary corrections on the weighted load state (e.g., constraining a 78% load rate to 75%). Finally, the upper limit of the available power capacity of each charging pile node is output (e.g., limiting the maximum power of a single pile from 100kW to 90kW).

[0049] 204. The upper limit of the available power capacity is temporally and spatially superimposed with the peak interval of the network traffic data, and the demand distribution density of the charging pile group in the target area is output through density field calculation. The spatial resolution of the density field is determined by the physical spacing of the charging pile nodes and the response delay time of the reactive compensation device.

[0050] In step 204, the density field is a charging demand distribution model generated by spatiotemporally superimposing the upper limit of available power capacity and the peak interval of network traffic data, and its spatial resolution is determined by the physical spacing of charging piles and the response delay of the power grid.

[0051] In this embodiment, the upper limit of available power capacity is mapped onto a spatial grid (e.g., 10 m x 10 m) as a power capacity distribution layer, and the peak network traffic interval is mapped onto a traffic weight layer based on time slices. The two layers of data are fused using a spatiotemporal kriging interpolation algorithm to generate an initial density field. A spatial resolution base value is defined based on the physical spacing of charging piles (e.g., 15 meters between adjacent piles). A temporal resolution compensation coefficient (e.g., 0.8) is calculated based on the response delay of the reactive power compensation device (e.g., 2 seconds). A convolutional neural network (CNN) is used to optimize the smoothness of the initial density field, ultimately outputting the demand distribution density.

[0052] Here's a specific example: Assume that a smart park faces a surge in charging demand during the evening peak period (18:00-20:00). The communication node monitors a network traffic peak of 120Gbps. Through discrete event sequence segmentation, it is identified that 70% of the charging piles are under load within a 5-minute window. The network traffic peak range is 115-118Gbps, with a correlation strength of 0.88. The cross-domain correlation model outputs a peak period coupling weight of 0.9. After weighted correction, the load rate increases to 78%. The grid parameters show that the maximum load capacity can be 85% of the rated value. The load rate is constrained to 75% through capacity boundary correction. Finally, the demand distribution density is generated based on a 10-meter grid. Area A shows a demand density of 1.8 vehicles / minute (spatial resolution 10 meters, temporal resolution 2 seconds), triggering an overload warning.

[0053] Steps 201-204 achieve deep coupling of charging demand forecasting and grid operation through multi-dimensional coordination of time, space, and grid, integrating network traffic weights and grid capacity constraints, and breaking through the traditional model's neglect of physical boundaries. Through the linkage of capacity correction and reactive power compensation, the prediction results are guaranteed to strictly match the grid's carrying capacity limit. The density field provides high-resolution spatiotemporal distribution, supporting minute-level response and resource scheduling. Combined with load state recognition and noise filtering mechanisms, it effectively resists sudden interference and data anomalies.

[0054] In order to further improve the physical perception accuracy and grid linkage capabilities of the charging demand heat map, a multi-dimensional visualization decision system is constructed by integrating equipment vibration characteristics, no-load interference suppression, and density gradient field evolution. The spatial distribution analysis capability is enhanced based on the density gradient discretization mechanism of the physical layout of the charging pile; multi-dimensional data is superimposed to generate a visualization heat map to achieve a coordinated presentation of charging demand prediction and equipment health. In some embodiments, based on the spatial gradient variation characteristics of the demand distribution density, the characteristic parameters of the specific vibration mode and the elimination results of the no-load false trigger event are synchronously integrated to generate a multi-dimensional heat map containing charging pile availability prediction and load carrying capacity, including: 301. Discretize the spatial gradient variation characteristics of the demand distribution density into a plurality of local density variation rates according to the physical layout of the charging pile group, wherein the direction vector of the local density variation rate is associated with the vibration propagation direction of the plugging and unplugging action of the charging gun; In step 301, the local density change rate is a quantitative indicator of density change in a local area (e.g., the density increase or decrease within a unit distance) based on the spatial gradient change in demand distribution density according to the physical layout of the charging piles. The direction vector is a spatial vector that represents the direction of local density change. Its direction is consistent with the propagation direction of the vibration signal caused by the charging plug's plugging and unplugging action.

[0055] In this embodiment, based on the physical layout of the charging pile cluster (e.g., a grid-like distribution or star topology), a manifold learning algorithm is used to extract the spatial gradient characteristics of the demand distribution density and identify the boundaries between high-density and low-density areas. Directional field estimation techniques are used to calculate the direction vector of the local density change rate. For example, using an energy attenuation model of the vibration signal propagation path (e.g., the attenuation rate of the main vibration frequency in a metal bracket is 2 dB / m), the direction vector is aligned with the vibration propagation direction, ultimately forming a discretized distribution map of the local density change rate, where the vector length represents the intensity of the change and the direction reflects the density diffusion trend.

[0056] 302. Extracting a vibration frequency distribution interval that matches the plugging and unplugging action of the charging gun from the characteristic parameters of the specific vibration mode, and generating a vibration characteristic availability probability that represents the availability status of the charging pile based on the energy proportion of each frequency band within the vibration frequency distribution interval; In step 302, the vibration frequency distribution range is the characteristic frequency band range corresponding to the charging gun plugging and unplugging action in the vibration signal. The vibration characteristic availability probability is a quantitative indicator of charging pile availability calculated based on the energy proportion of each frequency band within the characteristic frequency band.

[0057] In this embodiment of the present application, independent component analysis (ICA) is performed on the raw signal collected by the piezoelectric vibration sensor to isolate the characteristic frequency bands of the charging gun plugging and unplugging action (e.g., the 45Hz main frequency and its harmonics). The power spectral density is used to calculate the energy contribution of each frequency band (e.g., 65% for 45Hz energy and 20% for 60Hz harmonics), and a vibration energy distribution histogram is constructed. A support vector machine (SVM) classifier is used to train the relationship between energy distribution and device failure rate (e.g., an energy dispersion > 30% corresponds to an increased probability of failure), and the vibration feature availability probability is output (e.g., a normalized value from 0 to 1).

[0058] 303. Based on the result of eliminating the no-load false triggering event, the no-load false triggering frequency of the charging pile node within a preset time period is counted, and combined with the power parameters of the steady-state power supply environment, a no-load interference suppression factor reflecting the correction coefficient of the charging pile load carrying capacity is generated; In step 303, the no-load interference suppression factor is a load carrying capacity correction coefficient generated according to the no-load false trigger frequency and the grid power parameter.

[0059] In this embodiment, the number of no-load false triggering events at charging pile nodes within a preset time period (e.g., one hour) is counted. Combined with the power factor compensation state (e.g., 0.92) and voltage fluctuation rate (e.g., 2.5%) in a steady-state power supply environment, a hidden Markov model (HMM) is used to predict the false triggering risk in future time periods. Based on the risk level (e.g., high risk, medium risk, low risk), suppression factor weights (e.g., 0.7, 0.85, 1.0) are assigned to generate a no-load interference suppression factor, which is used to correct the predicted upper limit of the load carrying capacity.

[0060] 304. Spatially interpolate the vibration feature availability probability and the no-load interference suppression factor according to the physical location of the charging pile, superimpose the vector distribution of the local density change rate, and generate a multi-dimensional heat map including the charging pile availability prediction and load carrying capacity.

[0061] In step 304 , the multi-dimensional heat map is a visualization chart integrating the probability of charging pile availability, the load carrying capacity correction coefficient, and the density gradient direction vector, and its color is negatively correlated with the power factor compensation state of the power grid.

[0062] In an embodiment of the present application, the vibration feature availability probability is mapped to a transparency channel (e.g., high transparency indicates low availability), the no-load interference suppression factor is mapped to color saturation (e.g., low saturation indicates a need for load reduction), and the local density change rate vector is mapped to a directional arrow. A spatial interpolation algorithm (e.g., inverse distance weighting) is used to fill parameter gaps in unmonitored areas, and a high-resolution thermal image is synthesized using a generative adversarial network (GAN). An attention mechanism is introduced to enhance the visual salience of overloaded areas (e.g., flashing red prompts), ultimately generating an interactive multidimensional heat map. The chromaticity mapping rule of the multidimensional heat map is negatively correlated with the power factor compensation status of the power grid. That is, the lower the power factor (e.g., 0.85), the darker the chromaticity, and the higher the power factor (e.g., 0.98), the lighter the chromaticity.

[0063] Here's a specific example: Assume that during peak weekend charging hours in a smart park, charging piles are arranged in a circular pattern. The demand density gradient direction vector in area C is detected to be consistent with the vibration propagation path, and the energy of the vibration signal at the main frequency of 50 Hz accounts for 75% (a vibration feature availability probability of 0.9). The no-load false trigger frequency in area D is 15 times per hour, and a suppression factor of 0.75 is generated based on the grid power factor of 0.88. The vibration feature availability probability and the suppression factor are spatially interpolated and superimposed on the density gradient vector. The heat map shows that area C is light green (power factor 0.95, negative correlation with color rule), with a transparency of 10% (high availability), and an arrow pointing northwest. Area D is dark red (power factor 0.88, negative correlation with color rule), with a transparency of 40% (low availability). A blurred arrow indicates risk, and the pointing arrow prompts operation and maintenance personnel to prioritize capacity expansion in area C.

[0064] Steps 301-304 achieve a multi-dimensional and refined presentation of the charging demand heat map through the collaborative analysis of physics, power grid, and spatial gradients. This integrates vibration characteristics and density gradient direction, overcoming the traditional heat map's neglect of equipment status. Load forecasts are corrected by suppression factors to ensure a strict match between the heat map and the power grid status. Multi-dimensional visualization parameters (color, transparency, and direction arrows) provide a basis for complex decision-making. No-load false trigger statistics and vibration energy analysis collaboratively filter noise data to improve reliability in complex scenarios.

[0065] To improve the accuracy of modeling the correlation between network traffic fluctuations and charging demand timing characteristics, a cross-domain causal correlation model is constructed through a multi-scale periodic decomposition and phase matching mechanism. Monitoring units are divided based on the communication protocol type and the dominant periodic components are extracted. A coupling function that can correct for timing offsets is constructed to achieve accurate prediction of network behavior and charging demand. In some embodiments, the mapping relationship between network traffic fluctuation patterns and charging demand timing characteristics is established, and by analyzing the periodic changes in the data transmission volume of communication nodes in the area where the charging piles are located, a cross-domain correlation model reflecting the coupling relationship between charging demand and network behavior is constructed, including: 401. Divide the network traffic monitoring units according to the physical coverage of the communication nodes in the area where the charging piles are located, perform multi-scale periodic decomposition on the data transmission volume of the communication nodes based on the communication protocol type, and extract the dominant periodic components of the hourly, daily, and weekly data transmission volume within each monitoring unit; In step 401, the network traffic monitoring unit divides the network traffic data collection area into zones based on the physical coverage area (e.g., a 200-meter radius) of the communication nodes in the area where the charging piles are located. The dominant cyclical component is the cyclical fluctuation component with the highest energy contribution at the hourly, daily, and weekly time scales after multi-scale decomposition of the network traffic data.

[0066] In this embodiment, network traffic monitoring units are divided according to the physical coverage radius of communication nodes (e.g., 500 meters for 4G base stations and 100 meters for Wi-Fi hotspots), ensuring a one-to-one mapping between charging stations and communication nodes within each unit. Wavelet packet decomposition is performed on the data transmission volume based on the communication protocol type (e.g., TCP, UDP, MQTT), extracting periodic components at different time scales (hours, days, weeks) within each monitoring unit. Power spectral density analysis is then used to select components with energy percentages exceeding a threshold as the dominant periodic component.

[0067] 402. Based on the temporal distribution characteristics of charging gun plug-in and unplugging events in the historical operation data of the charging pile, define the high-frequency demand period, stable demand period, and low-valley demand period of charging demand, and establish a mapping rule between the charging demand period label and the amplitude-frequency characteristic of the dominant periodic component; In step 402, the charging demand period labels are divided into high-frequency demand period, stable demand period, and low-valley demand period based on the time series distribution of charging plug plug and unplug events. The amplitude-frequency characteristic mapping rule is the correspondence between the amplitude and frequency distribution characteristics of the dominant periodic component and the charging demand period labels.

[0068] In this embodiment, the temporal distribution of charging gun plug-in and unplug events in the charging pile historical data is statistically analyzed, and a K-means clustering algorithm is used to divide the demand time period labels (e.g., cluster centers are 7:00-9:00, 12:00-14:00, and 18:00-20:00). The amplitude-frequency characteristics of the dominant periodic component (e.g., the peak amplitude and bandwidth of the daily component between 18:00-20:00) are extracted, and a mapping rule between the time period labels and the amplitude-frequency characteristics of the dominant periodic component is established through canonical correlation analysis (CCA).

[0069] 403. For each network traffic monitoring unit, the phase fluctuation interval of the dominant periodic component is matched with the duration window of the corresponding charging demand period label. By introducing the amplitude-frequency characteristic mapping rule into the matching process, a dynamic correlation weight between the network traffic fluctuation pattern and the charging demand time series characteristics is generated; In step 403, the phase fluctuation interval is the phase offset range of the dominant periodic component on the time axis. The association weight is a quantitative indicator of the degree of matching between the network traffic periodic component and the charging demand period.

[0070] In this embodiment of the present application, a sliding window analysis (e.g., a window size of 1 hour) is performed on the phase fluctuations of the dominant periodic component to detect the overlap between its peak time and the duration window of the demand period label. Based on the amplitude threshold and bandwidth range defined in the amplitude-frequency characteristic mapping rules, the phase matching tolerance is dynamically adjusted: if the amplitude exceeds the threshold and the bandwidth falls within the range (e.g., amplitude >100Gbps and bandwidth 10-20Hz in high-frequency time periods), a phase deviation of ±15 minutes is allowed, and a high weight base value (e.g., 1.0) is assigned. If the bandwidth exceeds the range (e.g., bandwidth >5Hz in low-frequency time periods), the phase tolerance is reduced to ±5 minutes, and the weight base value is reduced based on the amplitude attenuation ratio (e.g., a base value of 0.8 for an amplitude of 80Gbps). Finally, a dynamic association weight is generated based on the phase overlap (e.g., a coefficient of 1.2 is assigned for 90% overlap and 0.8 for 60% overlap).

[0071] 404. Based on the association weight, a coupling relationship function between the charging demand and the network behavior is constructed, wherein the timing offset of the dominant periodic component of the network traffic in the coupling relationship function is used to correct the cross-domain association error and output a structured cross-domain correlation model.

[0072] In step 404 , the coupling relationship function is a mathematical expression describing the association between the dominant periodic component of network traffic and the timing characteristics of charging demand, wherein the input is the timing offset of the periodic component and the output is the cross-domain correlation error correction value.

[0073] In this embodiment, a multivariate regression function with dynamic correlation weights as coefficients is constructed, with the timing offset of the dominant periodic component of network traffic (e.g., a 20-minute lag in the peak of the daily component) as the independent variable and the charging demand forecast error as the dependent variable. Function parameters are optimized using a gradient descent algorithm, and L1 regularization is introduced to mitigate overfitting. The weight distribution is dynamically adjusted based on amplitude-frequency mapping rules (e.g., broadband signals allow for greater timing offset compensation), ultimately outputting a structured cross-domain model.

[0074] Here's a specific example: Assume that the charging demand pattern of a smart park suddenly changes during holidays. The coverage area of ​​the communication nodes is divided into three monitoring units. Unit A detects a peak amplitude of 130Gbps and a bandwidth of 18Hz for the daily dominant period (which complies with the high-frequency period mapping rule). The phase fluctuation range is 18:10-18:50, with a 90% overlap with the demand period label 18:00-19:00. It is assigned a base weight of 1.0×1.2=1.2. Unit B has a dominant component amplitude of 60Gbps and a bandwidth of 8Hz (which complies with the low-peak period rule). The phase fluctuation range is 23:00-23:30, with a 70% overlap with the demand period label 22:00-24:00. The base weight is 0.6×0.9=0.54. The multivariate regression function detects a timing offset of +10 minutes for unit A and -15 minutes for unit B. After adjusting the weights based on the bandwidth rule, a structured cross-domain model is output, reducing the prediction error from 12% to 3%.

[0075] Steps 401-404 achieve accurate modeling of the relationship between network traffic and charging demand through the technical links of multi-scale period analysis, amplitude-frequency-phase coordination, and dynamic error correction. The cross-domain model integrates the characteristics of the communication protocol and the evolution of the demand period, significantly reducing the prediction deviation caused by timing offset; the dynamic weight mechanism adapts to grid load fluctuations to ensure that the prediction results match the real-time power supply capacity; the structured model supports rapid switching of multiple scenarios (holidays and weekdays), reducing reliance on manual parameter adjustment; the amplitude-frequency characteristic mapping rules and the phase fluctuation range provide a basis for causal correlation at the physical layer, improving the credibility of operation and maintenance.

[0076] In order to improve the spatiotemporal resolution and accuracy of the charging demand distribution density, spatial grid modeling and multi-scale data fusion are used to achieve collaborative calculation of power capacity and network traffic. This includes generating local power distribution based on grid division of the physical spacing of charging piles, building a density field based on the spatiotemporal characteristics of network traffic, and optimizing density increment calculation through multi-scale convolution to ultimately output high-precision demand distribution density. In some embodiments, the available power capacity upper limit and the peak interval of the network traffic data are spatiotemporally superimposed, and the demand distribution density of the charging pile group in the target area is output through density field calculation, including: 501. Divide the target area into spatial grid units according to the physical spacing of the charging pile nodes, discretize and sample the upper limit of the available power capacity according to the spatial grid units, and generate a local power capacity distribution; In step 501, the local power capacity distribution is a power distribution diagram obtained by discretizing and sampling the upper limit of the available power capacity of the charging pile node according to spatial grid units, reflecting the power supply capacity of each grid unit.

[0077] In this embodiment, the target area is divided into equally spaced grid cells (e.g., 10 m x 10 m) based on the physical spacing between charging pile nodes (e.g., 20 m between adjacent piles). The upper limit of the available power capacity of each charging pile (e.g., a maximum power of 100 kW per pile) is discretized and mapped to the corresponding grid cell using an inverse distance weighted interpolation algorithm. The interpolation result is then corrected based on the grid load factor (e.g., 80%) (e.g., the power capacity of overloaded grid cells is reduced to 80 kW) to generate a local power capacity distribution.

[0078] 502. Divide the network traffic data peak interval into multiple continuous time windows according to the time dimension, extract the spatial distribution characteristics of the network traffic data in each time window, and generate the network traffic spatial weight corresponding to the spatial grid unit; In step 502 , the network traffic spatial weight is the distribution weight of the network traffic data peak interval divided by the time window on the spatial grid unit, reflecting the spatiotemporal driving intensity of network behavior on charging demand.

[0079] In this embodiment, the peak period of network traffic data (e.g., 150 Gbps between 6:00 PM and 6:30 PM) is divided into time windows (e.g., 5-minute windows). The spatial distribution characteristics of traffic (e.g., TCP protocol traffic share and packet size distribution) for each communication node within each window are extracted. Principal component analysis (PCA) is then used to reduce the dimensionality and generate weight coefficients corresponding to spatial grid cells (e.g., a weight of 0.9 for the center grid and 0.6 for the edge grid during peak hours), forming the spatial weight of network traffic.

[0080] 503. Superimpose the local power capacity distribution and the network traffic spatial weight according to the spatial grid unit to generate a density field reflecting the coupling relationship between the power capacity of the charging pile group and the network traffic; In step 503 , the density field is a charging demand density spatial model generated by superimposing the local power capacity distribution and the network traffic spatial weight, which characterizes the coupling effect of power capacity and network traffic.

[0081] In this embodiment, the local power capacity distribution and the network traffic spatial weight matrix are superimposed point by point on a grid cell basis. A spatially weighted fusion algorithm (such as the entropy weight method) is used to balance the contribution ratios of power capacity and traffic weight (e.g., 60% for power capacity and 40% for traffic weight). Spatial autocorrelation analysis (Moran's I index) is introduced to detect hotspots in the density field (e.g., high-density clusters) to generate a density field.

[0082] 504. Perform a multi-scale convolution operation on the local power capacity distribution and the network traffic spatial weight according to the spatial grid unit, calculate the density increment of the spatial grid unit in the density field, and output the demand distribution density of the charging pile group in the target area.

[0083] In step 504, the density increment is the demand density change of the spatial grid unit in the density field calculated by the multi-scale convolution operation, reflecting the short-term demand fluctuation trend.

[0084] In this embodiment, a multi-scale convolution kernel (e.g., 3×3, 5×5, 7×7) is designed to perform a sliding window scan of the density field, extracting density gradient features (e.g., local mutations and regional trends) within different spatial ranges. A long short-term memory (LSTM) network is used to predict density increments for future time windows (e.g., 5 minutes). A Kalman filter is then used to correct prediction errors and output density increments. Finally, the density increments are superimposed on the current density field to generate an updated demand distribution density.

[0085] Here's a specific example: Assume that a smart park has a cluster of charging piles arranged in a star configuration, spaced 18 meters apart. Step 501 divides the target area into a 12m x 12m grid. Charging piles are densely packed in the center (10m apart), generating a local power capacity cap of 150kW (65% load factor) and 100kW in the sparser edge regions. Step 502 monitors network traffic peaks from 3:00 PM to 3:15 PM, reaching 180Gbps (live data transmission). The spatial weight of the traffic in the center grid reaches 0.98, while that in the edge grids is 0.6. Step 503 superimposes the power capacity and traffic weights to generate a density field indicating a demand density of 2.2 vehicles per minute in the center grid, triggering a yellow alert. Step 504 uses a multi-scale convolution kernel to detect the density increment in the center grid (+0.5 vehicles per minute over a 5-minute period). The LSTM predicts that the density will exceed 3.0 vehicles per minute in the next 10 minutes. This automatically initiates power allocation to adjacent idle grids, temporarily increasing the power capacity of the edge grids to 120kW to relieve the pressure.

[0086] Steps 501-504 achieve refined calculation of charging demand density through the technical path of grid power distribution, traffic weight fusion and multi-scale density evolution. Grid division is combined with multi-scale convolution to accurately capture local demand mutations. The incremental prediction and correction mechanism supports minute-level density updates. The balance between power capacity and network traffic weight ensures the physical feasibility of the prediction results. The grid model is adapted to different campus topologies to reduce deployment costs.

[0087] To enhance the heat map's ability to collaboratively represent device status and grid response, a multidimensional visualization engine is constructed through vibration propagation path energy attenuation modeling and reactive compensation response control. A vibration interpolation field is generated based on the directional characteristics of the charging pile. Density gradient vectors are superimposed to generate a multidimensional heat map, achieving three-dimensional visualization of device availability, grid status, and demand density. In some embodiments, the vibration feature availability probability and the no-load interference suppression factor are spatially interpolated according to the physical location of the charging pile, and the vector distribution of the local density change rate is superimposed to generate a multidimensional heat map containing charging pile availability predictions and load carrying capacity, including: 601. Construct a vibration feature interpolation field based on the directional characteristics of the physical location of the charging pile, where the interpolation weight of the vibration feature interpolation field is adjusted by the vibration propagation path energy attenuation in the vibration feature availability probability; In step 601, the vibration feature interpolation field is a spatial interpolation model constructed based on the directional characteristics of the physical location of the charging pile, and its interpolation weight is adjusted by the energy attenuation rate of the vibration signal propagation path.

[0088] In this embodiment, a spatial interpolation field is constructed using the Kriging interpolation algorithm based on the directional characteristics of the charging pile's physical layout (e.g., the difference in vibration propagation path attenuation between north-south charging guns). Vibration sensor data is used to fit an energy attenuation curve (e.g., an attenuation coefficient of 1.5 dB / m for metal brackets and 2.2 dB / m for concrete brackets). Interpolation weights for adjacent charging pile nodes are adjusted to ultimately generate a vibration characteristic interpolation field that reflects the spatial distribution of device availability.

[0089] 602. Construct a no-load interference correction field based on the spatial proximity of the physical locations of the charging piles, wherein the correction weight of the no-load interference correction field is controlled by the response time of the grid reactive compensation device in the no-load interference suppression factor; In step 602, the no-load interference correction field is a correction model constructed based on the spatial proximity relationship of the charging piles, and its weight is controlled by the response time of the grid reactive compensation device in the no-load interference suppression factor.

[0090] In this embodiment, the frequency of no-load false triggering at charging pile nodes (e.g., a node falsely triggers eight times per hour) is counted. Combined with the response time of the grid's reactive power compensation device (e.g., a delay of 3 seconds), a spatial autoregressive (SAR) model is used to calculate correction weights. Monte Carlo simulations are used to predict the correction effects at different response times (e.g., a weight of 0.85 for a 2-second delay and 0.7 for a 5-second delay), generating a no-load interference correction field matrix.

[0091] 603. Visualize the vibration characteristic interpolation field and the no-load interference correction field according to the physical position of the charging pile to generate an initial thermal distribution; In step 603 , the initial thermal distribution is an initial distribution map generated by superimposing the vibration characteristic interpolation field and the no-load interference correction field according to the physical location of the charging pile, reflecting the preliminary results of equipment availability and load correction.

[0092] In this embodiment, a point-by-point product operation is performed on the vibration feature interpolation field (availability probability 0-1) and the no-load interference correction field (correction coefficient 0-1). Voxel rendering technology is used to map the result to a thermal base color (for example, red indicates high availability and low interference). Bilinear interpolation is used to fill in the missing values ​​in the space, generating a smooth initial thermal distribution.

[0093] 604. Spatially superimpose the vector distribution of the local density change rate and the initial thermal distribution to generate a multi-dimensional thermal map including charging pile availability prediction and load carrying capacity.

[0094] In step 604 , the multi-dimensional heat map is a visualization chart generated by superimposing the vector distribution of the local density change rate (such as the arrow direction) and the initial heat distribution, and its color is negatively correlated with the power factor compensation state of the power grid.

[0095] In this embodiment, the vector direction of the local density change rate (e.g., southeast density diffusion) is extracted and projected onto a two-dimensional plane using a manifold learning algorithm (e.g., t-SNE). The vector arrows are superimposed on the initial thermal distribution using a transparency blending algorithm. The chromaticity channel is rendered based on the power factor (e.g., 0.88 is mapped to a deep red), ultimately generating a multidimensional heat map that includes predictions of charging pile availability and load carrying capacity.

[0096] Here's a specific example: Assume that a cluster of charging piles in a smart park is arranged radially, with the north side facing the main road (vibration propagation path attenuation of 1.2 dB / m) and the south side facing the parking lot (attenuation of 2.0 dB / m). Step 601 constructs a vibration feature interpolation field, with interpolation weights of 0.92 for the north side and 0.75 for the south side. Step 602 detects a false no-load trigger frequency of 12 times / hour in the south side, with a grid response delay of 4 seconds. A correction weight of 0.68 is obtained, and a no-load interference correction field is constructed. Step 603 integrates the generated initial thermal distribution, with the north side displaying light green (availability 0.9) and the south side displaying orange (availability 0.7). Step 604 superimposes the vector direction of the local density change rate (north points to the center area). The resulting multi-dimensional heat map prompts operations personnel to deploy additional mobile charging vehicles to the north to avoid overloading the center.

[0097] In order to improve the accuracy and anti-interference ability of charging pile load state identification, a physical layer perception closed loop is constructed through multi-band analysis of vibration signals and energy evolution modeling. Vibration signal collection based on stress concentration areas enhances effective feature extraction; spectral characteristics and energy trends are combined to distinguish between load and no-load interference, achieving highly robust state identification. In some embodiments, the piezoelectric vibration sensor is integrated inside the charging pile support structure. Based on the amplitude difference of different frequency components in the mechanical vibration signal of the charging pile, the specific vibration mode caused by the plugging and unplugging of the charging gun is identified to distinguish the actual load operation state of the charging pile from the no-load false trigger event, including: 701. Arrange a piezoelectric vibration sensor in the stress concentration area of ​​the charging pile support structure to collect multi-band vibration components in the mechanical vibration signal of the charging pile; In step 701, the stress concentration area is the area in the charging pile support structure that is prone to significant vibration due to mechanical stress concentration. The multi-band vibration component is the vibration components in multiple frequency intervals obtained by frequency domain decomposition of the mechanical vibration signal collected by the piezoelectric vibration sensor.

[0098] In this embodiment, piezoelectric vibration sensors are deployed at the bolted joints and bracket welds of the charging pile support structure. Raw vibration signals are collected at a high sampling rate (e.g., 10 kHz). Wavelet packet decomposition is used to decompose the signal into multiple frequency bands (e.g., sub-bands spaced 10 Hz apart), and the time-domain amplitude sequence of each frequency band is extracted. Key frequency bands are selected based on their energy percentage (e.g., frequency bands with an energy percentage greater than 5% are retained), generating multi-band vibration components.

[0099] 702. Extract a characteristic frequency band from the multi-band vibration components that matches the plugging and unplugging action of the charging gun, and generate a vibration energy distribution of the charging pile based on an amplitude variation pattern of the characteristic frequency band; In step 702 , the vibration energy distribution is a curve showing the evolution of the square integral of the amplitude of each sub-band within the characteristic frequency band over time.

[0100] In this embodiment, independent component analysis (ICA) is performed on multi-band vibration components to isolate the characteristic frequency bands of the charging gun plugging and unplugging actions (e.g., the 45Hz main frequency and its 90Hz second harmonic). The short-term energy of each sub-band within the characteristic frequency band is calculated (e.g., the sum of the squared amplitudes of a 50ms window) to generate the charging pile vibration energy distribution. The energy curve is then smoothed using a sliding average filter to eliminate transient impact noise (e.g., sudden vibrations caused by passing vehicles).

[0101] 703. Identify a specific vibration mode caused by the plugging and unplugging action of the charging gun based on a comparison result of a peak interval of the vibration energy distribution of the charging pile and a preset vibration energy threshold; In step 703, the preset vibration energy threshold is an energy critical value set according to historical data statistics. The specific vibration pattern is a vibration signal pattern that meets the energy threshold condition and has specific spectrum characteristics.

[0102] In this embodiment, statistics are collected to determine the peak energy distribution in characteristic frequency bands of historical charging events (e.g., fundamental frequency energy peaks within the range of 70-100dB), and thresholds are set (e.g., fundamental frequency energy >75dB is considered a valid event). A support vector machine (SVM) is used to classify energy distribution curves (e.g., single-peak, double-peak, multi-peak), screening for single-peak, concentrated energy curves consistent with charging gun plugging and unplugging. The spectral flatness index (e.g., <0.3 is considered valid) is used to verify pattern consistency and identify specific vibration patterns.

[0103] 704. Distinguish between the actual load operation state of the charging pile and the no-load false triggering event based on the spectrum distribution characteristics and energy fluctuation trend of the specific vibration mode.

[0104] In step 704, the spectrum distribution characteristic is the energy concentration and harmonic attenuation law of the vibration signal in the frequency domain. The energy fluctuation trend is the stability or sudden change of the vibration signal energy over time.

[0105] In this embodiment of the present application, principal component analysis (PCA) is performed on the spectrum of a specific vibration mode to extract key features such as the fundamental frequency energy percentage and harmonic attenuation slope. A hidden Markov model (HMM) is used to model energy fluctuation trends (e.g., the slow rise and fall of energy in a real charging event, and the rapid spike in a no-load event). A joint discrimination rule is then established to distinguish between the actual load operation state of the charging pile and no-load false triggering events: load operation is determined when the fundamental frequency energy percentage is greater than 65% and the energy fluctuation trend is stable; no-load false triggering is determined when the fundamental frequency energy is dispersed (percentage <50%) or the fluctuation suddenly changes.

[0106] Here's a specific example: A charging pile was temporarily added to a smart park, and vibration signals were detected at the bolts of its metal bracket. Wavelet packet decomposition was performed in step 701 to extract the 45Hz, 90Hz, and 135Hz frequency bands, forming a multi-band vibration component dataset. In step 702, ICA was used to isolate the 45Hz main frequency energy, which accounts for 70%. The short-term energy of each sub-band was calculated to distribute the charging pile's vibration energy. In step 703, the SVM classified it as a single-peak concentrated vibration, with a fundamental frequency energy peak of 82dB (exceeding the 75dB threshold), indicating a specific vibration mode. In step 704, PCA showed that the fundamental frequency energy accounted for 72%, and the HMM determined that the energy trend was stable, confirming it was a real load. Another vibration event, with a dispersed 60Hz main frequency energy (accounting for 48%), was detected by the HMM as a sudden energy change and marked as a no-load false trigger.

[0107] Figure 2The present invention provides a schematic diagram of a system for generating a charging demand heat map based on time series prediction. Figure 2 As shown, the system includes: Analysis module 21 is used to establish a mapping relationship between network traffic fluctuation patterns and charging demand time series characteristics. By analyzing the periodic changes in data transmission volume of communication nodes in the area where the charging piles are located, a cross-domain correlation model reflecting the coupling relationship between charging demand and network behavior is constructed; Identification module 22, which is used to integrate a piezoelectric vibration sensor within the charging pile support structure. Based on the amplitude differences of different frequency components in the charging pile mechanical vibration signal, it identifies the specific vibration mode caused by the plugging and unplugging of the charging gun, and distinguishes the actual load operation state of the charging pile from the no-load false trigger event; Suppression module 23 is used to deploy reactive power compensation devices on the distribution side. By adjusting the switching state of compensation capacitors, it suppresses the fluctuation of power factor of the power grid caused by the access of charging pile clusters and generates a steady-state power supply environment that matches the charging demand prediction model. A calculation module 24 is configured to perform spatiotemporal correlation between the output of the cross-domain correlation model and the identification result of the actual load operation state of the charging pile, and calculate the demand distribution density of the charging pile group in the target area in combination with the power parameters of the steady-state power supply environment; A generation module 25 is used to synchronously integrate the characteristic parameters of the specific vibration mode and the elimination results of the no-load false triggering event according to the spatial gradient change characteristics of the demand distribution density, to generate a multidimensional heat map including the charging pile availability prediction and load carrying capacity, wherein the chromaticity mapping rule of the multidimensional heat map is negatively correlated with the power factor compensation state of the power grid.

[0108] Figure 2 The charging demand heat map generation system based on time series prediction can be executed Figure 1 The implementation principles and technical effects of the method for generating a charging demand heat map based on time series prediction described in the illustrated embodiment are not further elaborated. The specific manner in which each module and unit performs operations in the system for generating a charging demand heat map based on time series prediction in the above embodiment has been described in detail in the relevant embodiments of the method and will not be elaborated on here.

[0109] In one possible design, Figure 2 The charging demand heat map generation system based on time series prediction of the embodiment shown can be implemented as a computing device, such as Figure 3 As shown, the computing device may include a storage component 31 and a processing component 32; The storage component 31 stores one or more computer instructions, wherein the one or more computer instructions are called and executed by the processing component 32 .

[0110] The processing component 32 is used for the above Figure 1 The embodiment provides a method for generating a charging demand heat map based on time series prediction.

[0111] The processing component 32 may include one or more processors to execute computer instructions to complete all or part of the steps in the above method. Of course, the processing component may also be implemented as one or more application-specific integrated circuits (ASICs), digital signal processors (DSPs), digital signal processing devices (DSPDs), programmable logic devices (PLDs), field programmable gate arrays (FPGAs), controllers, microcontrollers, microprocessors, or other electronic components to perform the above method.

[0112] The storage component 31 is configured to store various types of data to support operations at the terminal. The storage component can be implemented by any type of volatile or non-volatile memory device, or a combination thereof, such as static random access memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic memory, flash memory, magnetic disk, or optical disk.

[0113] Of course, a computing device may also include other components, such as input / output interfaces, display components, communication components, etc.

[0114] The input / output interface provides an interface between the processing component and the peripheral interface module, which can be an output device, an input device, etc.

[0115] The communication component is configured to facilitate, among other things, wired or wireless communications between the computing device and other devices.

[0116] Among them, the computing device can be a physical device or an elastic computing host provided by a cloud computing platform, etc. In this case, the computing device can refer to a cloud server, and the above-mentioned processing components, storage components, etc. can be basic server resources rented or purchased from the cloud computing platform.

[0117] The present application also provides a computer storage medium storing a computer program, wherein the computer program can achieve the above-mentioned Figure 1 The illustrated embodiment provides a method for generating a charging demand heat map based on time series prediction.

[0118] 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.

[0119] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate, and the components shown as units may or may not be physical units, i.e., they may be located in one location or distributed across multiple network units. Some or all of the modules may be selected based on actual needs to achieve the objectives of the present embodiment. Persons of ordinary skill in the art will be able to understand and implement the present invention without inventive effort.

[0120] Through the above description of the embodiments, those skilled in the art will clearly understand that each embodiment can be implemented using software plus a necessary general-purpose hardware platform, or of course, hardware. Based on this understanding, the essence of the above technical solution, or the portion that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, a magnetic disk, or an optical disk, and includes a number of instructions for causing a computer device (such as a personal computer, server, or network device) to execute the methods described in each embodiment or certain portions of the embodiments.

[0121] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present application, rather than to limit them. Although the present application has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. 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 application.

Claims

1. A method for generating a charging demand heat map based on time series prediction, characterized in that: include: Establish a mapping relationship between network traffic fluctuations and charging demand timing characteristics, and analyze the periodic variation of data transmission volume of communication nodes in the area where the charging piles are located based on the mapping relationship, so as to build a cross-domain correlation model that reflects the coupling relationship between charging demand and network behavior; A piezoelectric vibration sensor is integrated inside the charging pile support structure. Based on the amplitude difference of different frequency components in the mechanical vibration signal of the charging pile, the specific vibration mode caused by the plugging and unplugging of the charging gun is identified to distinguish the actual load operation state of the charging pile from the no-load false trigger event. Deploy reactive power compensation devices on the distribution side to suppress the fluctuation of power factor of the power grid caused by the access of charging pile clusters by adjusting the switching state of compensation capacitors, and generate a steady-state power supply environment that matches the charging demand prediction model; The output result of the cross-domain correlation model is temporally and spatially correlated with the identification result of the actual load operation state of the charging pile, and the demand distribution density of the charging pile group in the target area is calculated in combination with the power parameters of the steady-state power supply environment; According to the spatial gradient variation characteristics of the demand distribution density, the characteristic parameters of the specific vibration mode and the elimination results of the no-load false triggering event are synchronously integrated to generate a multi-dimensional heat map including the charging pile availability prediction and load carrying capacity.

2. The method according to claim 1, characterized in that The step of temporally and spatially associating the output result of the cross-domain correlation model with the identification result of the actual load operation state of the charging pile, and calculating the demand distribution density of the charging pile group in the target area in combination with the power parameter of the steady-state power supply environment includes: The identification result of the actual load operation state of the charging pile is divided into a discrete event sequence according to a preset time slice, and the load state in each time slice in the discrete event sequence is associated with the peak interval of the network traffic data of the corresponding time slice; According to the coupling weight of network traffic and charging demand output by the cross-domain correlation model, weighting the load state in each time slice in the discrete event sequence to obtain a weighted result; Establishing a constraint relationship between the power parameter of the steady-state power supply environment and the maximum instantaneous load capacity of the charging pile, performing capacity boundary correction on the weighted result based on the constraint relationship, and generating an upper limit of the available power capacity of the charging pile node; The upper limit of available power capacity is superimposed on the peak interval of network traffic data in time and space, and the demand distribution density of the charging pile group in the target area is output through density field calculation. The spatial resolution of the density field is determined by the physical spacing of the charging pile nodes and the response delay time of the reactive compensation device.

3. The method according to claim 1, characterized in that According to the spatial gradient variation characteristics of the demand distribution density, the characteristic parameters of the specific vibration mode and the elimination results of the no-load false triggering event are synchronously integrated to generate a multi-dimensional heat map including the charging pile availability prediction and load carrying capacity, including: Discretizing the spatial gradient variation characteristics of the demand distribution density into a plurality of local density variation rates according to the physical layout of the charging pile group, wherein the direction vector of the local density variation rate is associated with the vibration propagation direction of the plugging and unplugging action of the charging gun; Extracting a vibration frequency distribution interval that matches the plugging and unplugging action of the charging gun from the characteristic parameters of the specific vibration mode, and generating a vibration characteristic availability probability that characterizes the availability state of the charging pile based on the energy proportion of each frequency band in the vibration frequency distribution interval; According to the elimination result of the no-load false triggering event, the no-load false triggering frequency of the charging pile node within a preset time period is counted, and the no-load interference suppression factor reflecting the correction coefficient of the charging pile load carrying capacity is generated in combination with the power parameters of the steady-state power supply environment; The vibration feature availability probability and the no-load interference suppression factor are spatially interpolated according to the physical position of the charging pile, and the vector distribution of the local density change rate is superimposed to generate a multi-dimensional heat map including the charging pile availability prediction and load carrying capacity.

4. The method according to claim 1, characterized in that: The mapping relationship between the fluctuation law of network traffic and the time series characteristics of charging demand is established, and based on the mapping relationship, the periodic variation law of data transmission volume of communication nodes in the area where the charging pile is located is analyzed to construct a cross-domain correlation model reflecting the coupling relationship between charging demand and network behavior, including: Divide the network traffic monitoring units according to the physical coverage of the communication nodes in the area where the charging piles are located, perform multi-scale periodic decomposition of the data transmission volume of the communication nodes based on the communication protocol type, and extract the dominant periodic components of the hourly, daily, and weekly data transmission volume in each monitoring unit; According to the time series distribution characteristics of charging gun plugging and unplugging events in the historical operation data of the charging pile, the high-frequency demand period, stable demand period and valley demand period of charging demand are defined, and the mapping rules between the charging demand period label and the amplitude-frequency characteristics of the dominant periodic component are established; For each network traffic monitoring unit, the phase fluctuation interval of the dominant periodic component is matched with the duration window of the corresponding charging demand period label, and the dynamic correlation weight between the network traffic fluctuation law and the charging demand timing characteristics is generated by introducing the amplitude-frequency characteristic mapping rule in the matching process; Based on the association weight, a coupling relationship function between the charging demand and the network behavior is constructed, and the timing offset of the dominant periodic component of the network traffic in the coupling relationship function is used to correct the cross-domain association error, and a structured cross-domain correlation model is output.

5. The method according to claim 2, characterized in that: The step of performing spatiotemporal superposition of the upper limit of the available power capacity and the peak interval of the network traffic data, and outputting the demand distribution density of the charging pile group in the target area through density field calculation, comprises: Divide the target area into spatial grid units according to the physical spacing of the charging pile nodes, discretize and sample the upper limit of the available power capacity according to the spatial grid units, and generate a local power capacity distribution; Divide the network traffic data peak interval into multiple continuous time windows according to the time dimension, extract the spatial distribution characteristics of the network traffic data in each time window, and generate the network traffic spatial weight corresponding to the spatial grid unit; The local power capacity distribution and the network traffic spatial weight are superimposed according to the spatial grid unit to generate a density field reflecting the coupling relationship between the power capacity of the charging pile group and the network traffic; The local power capacity distribution and the network traffic spatial weight are subjected to multi-scale convolution operation according to the spatial grid unit, the density increment of the spatial grid unit in the density field is calculated, and the demand distribution density of the charging pile group in the target area is output.

6. The method according to claim 3, characterized in that The vibration feature availability probability and the no-load interference suppression factor are spatially interpolated according to the physical position of the charging pile, and the vector distribution of the local density change rate is superimposed to generate a multi-dimensional heat map including the charging pile availability prediction and load carrying capacity, including: Constructing a vibration feature interpolation field based on the directional characteristics of the physical position of the charging pile, wherein the interpolation weight of the vibration feature interpolation field is adjusted by the energy attenuation of the vibration propagation path in the vibration feature availability probability; A no-load interference correction field is constructed based on the spatial proximity relationship of the physical positions of the charging piles, wherein the correction weight of the no-load interference correction field is controlled by the response time of the grid reactive compensation device in the no-load interference suppression factor; Visualizing the vibration characteristic interpolation field and the no-load interference correction field according to the physical position of the charging pile to generate an initial thermal distribution; The vector distribution of the local density change rate is spatially superimposed on the initial thermal distribution to generate a multi-dimensional thermal map including charging pile availability prediction and load carrying capacity.

7. The method according to claim 1, characterized in that The piezoelectric vibration sensor is integrated inside the charging pile support structure, and based on the amplitude difference of different frequency components in the mechanical vibration signal of the charging pile, the specific vibration mode caused by the plugging and unplugging action of the charging gun is identified to distinguish the actual load operation state of the charging pile from the no-load false trigger event, including: Arrange piezoelectric vibration sensors in the stress concentration area of ​​the charging pile support structure to collect multi-band vibration components in the mechanical vibration signal of the charging pile; Extracting a characteristic frequency band matching the plugging and unplugging action of the charging gun from the multi-band vibration components, and generating a vibration energy distribution of the charging pile based on the amplitude variation law of the characteristic frequency band; According to the comparison result of the peak interval of the vibration energy distribution of the charging pile and the preset vibration energy threshold, the specific vibration mode caused by the plugging and unplugging action of the charging gun is identified; Based on the spectrum distribution characteristics and energy fluctuation trends of the specific vibration mode, the actual load operation state of the charging pile and the no-load false triggering event are distinguished.

8. A charging demand heat map generation system based on time series prediction, characterized in that: include: The analysis module is used to establish a mapping relationship between the fluctuation law of network traffic and the timing characteristics of charging demand. By analyzing the periodic variation law of the data transmission volume of the communication nodes in the area where the charging piles are located, a cross-domain correlation model that reflects the coupling relationship between charging demand and network behavior is constructed; The recognition module is used to integrate a piezoelectric vibration sensor inside the charging pile support structure. Based on the amplitude difference of different frequency components in the mechanical vibration signal of the charging pile, it can identify the specific vibration mode caused by the plugging and unplugging of the charging gun and distinguish the actual load operation state of the charging pile from the no-load false trigger event. The suppression module is used to deploy reactive power compensation devices on the distribution side. By adjusting the switching state of the compensation capacitor, it suppresses the fluctuation of the power factor of the power grid caused by the access of the charging pile cluster and generates a steady-state power supply environment that matches the charging demand prediction model. A calculation module, used to perform spatiotemporal correlation between the output result of the cross-domain correlation model and the identification result of the actual load operation state of the charging pile, and calculate the demand distribution density of the charging pile group in the target area in combination with the power parameters of the steady-state power supply environment; A generation module is used to synchronously fuse the characteristic parameters of the specific vibration mode and the elimination results of the no-load false triggering event according to the spatial gradient change characteristics of the demand distribution density, to generate a multidimensional heat map including the charging pile availability prediction and load carrying capacity, wherein the chromaticity mapping rule of the multidimensional heat map is negatively correlated with the power factor compensation state of the power grid.

9. A computing device, characterized in that It comprises a processing component and a storage component; the storage component stores one or more computer instructions; the one or more computer instructions are used to be called and executed by the processing component to implement a method for generating a charging demand heat map based on time series prediction as described in any one of claims 1 to 7.

10. A computer storage medium, characterized in that: A computer program is stored, and when the computer program is executed by a computer, a method for generating a charging demand heat map based on time series prediction as described in any one of claims 1 to 7 is implemented.

Images