Charging pile intelligent distribution system based on dynamic load balancing and charging pile

Abstract

The invention discloses a charging pile intelligent distribution system based on dynamic load balancing and a charging pile, and relates to the technical field of charging piles, a distribution module is used for receiving an operation feature group of each charging pile at a first response moment, generating a feature weight set, and simulating importance judgment logic in the feature weight set according to a feature priority, dynamic load distribution of the charging piles is realized; the process module is used for receiving the data of each source in real time in the dynamic load distribution process, performing multi-source analysis, checking the fault burstiness of the charging pile, analyzing the communication interruption, predicting the charging demand performance and the power supply performance based on the data of each source at a second response moment, and constructing a comprehensive risk index in combination with an evidence theory; and the updating module is used for generating a judgment signal according to the comprehensive risk index value, and the judgment signal comprises an updating instruction and an iteration result.

Description

Technical Field

[0001] The present invention relates to the technical field of charging piles, and in particular to a charging pile intelligent distribution system and a charging pile based on dynamic load balancing. Background Art

[0002] As the number of electric vehicles continues to grow, the distribution density and efficiency of charging piles directly affect the user charging experience and the sustainable development of the industry. Smart charging pile systems have emerged. With the help of the Internet of Things, big data and artificial intelligence technologies, such systems can achieve remote monitoring, fault warning and resource scheduling of charging piles. However, in actual operations, the problem of dynamic load distribution of charging piles has gradually become prominent - the charging demand in different regions and time periods varies significantly, and some charging piles may be overloaded or idle, resulting in waste of resources and reduced charging efficiency. Therefore, how to achieve uniform distribution of dynamic load of charging piles through scientific algorithms and improve overall service efficiency has become a core issue that needs to be solved in the field of smart charging piles.

[0003] At present, the traditional charging pile allocation method is mostly static allocation, which is difficult to adapt to various unexpected situations during the operation of charging piles. For example, fixed priority sorting according to the geographical location and power size of charging piles can easily lead to load imbalance. For example, when there is a large-scale centralized charging demand, the traditional system is difficult to respond quickly and dynamically adjust the allocation plan, which may cause some charging piles to fail frequently due to overload operation, while other charging piles are in a low utilization state.

[0004] The limitations of traditional load distribution methods have directly triggered a series of chain reactions. From the user's perspective, problems such as long charging queues and charging interruptions have reduced user satisfaction and even hindered the popularization of electric vehicles; from the operator's perspective, the low utilization rate of charging piles not only causes idle resources and waste, but frequent fault repairs also increase operating costs. At the risk management level, a single risk assessment method makes it difficult for the system to cope with a complex and changing operating environment. Once there is insufficient power supply or network communication interruption, the load distribution cannot be adjusted in time, which may lead to regional charging service paralysis. These problems seriously restrict the efficient operation of charging pile infrastructure, and innovative technologies are urgently needed to achieve dynamic load balancing and intelligent risk management. Summary of the invention

[0005] In view of the deficiencies in the prior art, the present invention provides a charging pile intelligent allocation system and a charging pile based on dynamic load balancing, which solve the problems in the above-mentioned background technology.

[0006] To achieve the above objectives, the present invention is implemented through the following technical solutions: a charging pile intelligent allocation system based on dynamic load balancing, comprising:

[0007] A distribution module, which is used to receive the operation feature groups of each charging pile at the first response moment, generate a feature weight set, and simulate the importance judgment logic in the feature weight set according to the feature priorities, so as to achieve dynamic load distribution of the charging piles;

[0008] A process module, which is used to receive each source data in real time during the dynamic load distribution process, based on each source data, and at the second response moment, perform multi-source analysis, check the suddenness of charging pile failures, analyze the communication interruption, predict the charging demand and power supply, and combine the evidence theory to construct a comprehensive risk index;

[0009] An update module, which is used to generate a determination signal according to the value of the comprehensive risk index, and the determination signal includes an update instruction and an iteration result.

[0010] Preferably, the distribution module includes a first receiving unit, a weight generation unit and an importance unit;

[0011] The first receiving unit is used to preset a response interval, and at the first response moment, receive the operation feature groups of each charging pile through several groups of sensors. The operation feature groups include but are not limited to voltage, current, power, connection number, charging speed, fault duration, operation duration and the number of charging guns; and preprocess the operation feature groups to remove duplicate data and fill in missing values, so as to obtain the preprocessed operation feature groups.

[0012] Preferably, the weight generation unit is used to extract the maximum and minimum values of various features in the preprocessed operation feature group, and combine their proportional relationship with the equipment health value of the corresponding charging pile to define the weights of various processed features; wherein, the equipment health value represents the ratio of the non-fault duration to the operation duration of the corresponding charging pile in the historical period.

[0013] Preferably, the importance unit is used to use the algorithm of Pearson correlation coefficient to analyze the correlation between various features and the equipment health value, so as to obtain the Pearson correlation coefficients between various features and the equipment health value respectively, and combine the Saaty1–9 scale method, and then compare the magnitude relationship between the Pearson correlation coefficients to construct a judgment matrix;

[0014] Based on the judgment matrix, use the power method to calculate the maximum eigenvalue and the corresponding eigenvector of the judgment matrix, and normalize the eigenvector corresponding to the maximum eigenvalue to obtain the weights of the corresponding features in each Pearson correlation coefficient. By arranging the weights of the corresponding features in each Pearson correlation coefficient in ascending order, determine the priorities of various features in the operation feature group, and according to the priorities and the weights of various features defined in the weight generation unit, and combine the hierarchical fusion method, calculate and obtain the final comprehensive weight of the corresponding charging pile. According to the final comprehensive weight of the corresponding charging pile, perform dynamic load distribution on the charging pile, specifically:

[0015] By calculating the ratio of the connection number of the corresponding charging pile to its corresponding final comprehensive weight, the allocation priority index of each charging pile is calculated. When there are vehicles waiting to be charged, the allocation priority indexes of each charging pile in the charging station are compared, and the charging pile corresponding to the smallest allocation priority index is selected as the preferred charging pile for the vehicle waiting to be charged this time.

[0016] Preferably, the process module includes a second receiving unit and a multi-source analysis unit;

[0017] The second receiving unit is used to receive each source data in real time on the basis of the importance unit. Each source data includes the operation characteristic group of the charging pile, the network communication characteristic group, and the power supply characteristic group; among them, the network communication characteristic group includes various characteristics during network communication; the power supply characteristic group includes the real-time remaining capacity of the power grid, the real-time power demand of the charging pile, and the upper limit of regional power distribution.

[0018] Preferably, the multi-source analysis unit is used at the second response moment to construct a fault prediction model according to the operation characteristic group of the charging pile by using the random forest algorithm, divide the operation characteristic group of the charging pile into a training set and a test set, use the training set to train the fault prediction model, adjust the parameters of the model to optimize the performance of the model, and then use the test set to verify the trained fault prediction model, evaluate the accuracy rate and recall rate of the model, input the operation characteristic group of the charging pile monitored in real time into the trained fault prediction model to make it learn the pattern of fault occurrence, and the fault prediction model outputs the probability of each charging pile having a fault in a future period of time; according to the network communication characteristic group, use the hidden Markov model to predict the network state, divide the network state into different level states, the level states include normal, unstable, and interrupted, and determine the signal strength range corresponding to each level state, and then use the received network communication characteristic group as the observation sequence and the network state as the hidden state to train the hidden Markov model through the known observation sequence and the corresponding hidden state, learn the state transition probability and observation probability to predict the probability of future network interruption; according to the operation characteristic group of the charging pile, extract the historical charging demand data, use the time series prediction algorithm to predict the charging demand in the future period, combine with the clustering analysis, and divide the high-demand areas and time periods through the clustering analysis to calculate the risk probability; according to the power supply characteristic group, extract the real-time remaining capacity of the power grid, the real-time power demand of the charging pile, and the upper limit of regional power distribution, and obtain the supply risk probability through ratio calculation.

[0019] Preferably, the multi-source analysis unit is further configured to use the Dempster-Shafer theory to take the probability of each charging pile failing within a future period of time, the probability of future network interruption, the risk probability, and the supply risk probability as independent evidence sources. For each independent evidence source, determine the basic probability assignment according to its own characteristics and data distribution to obtain the basic probability assignment function of the corresponding independent evidence source, and synthesize the basic probability assignment functions multiple times according to the D-S synthesis rule to calculate and obtain the comprehensive risk index.

[0020] Preferably, the update module includes a determination unit and an update unit;

[0021] The determination unit is configured to, when the comprehensive risk index does not exceed a pre-set risk threshold, extract the final comprehensive weight of the corresponding charging pile in the importance unit as the weight at the second response moment. Otherwise, generate an update instruction.

[0022] Preferably, the update unit is configured to determine the update instruction, and based on the final comprehensive weight of the corresponding charging pile in the importance unit, adjust the final comprehensive weight of the corresponding charging pile in the importance unit through the Sigmoid function to update the final comprehensive weight of the corresponding charging pile;

[0023] The iteration result returns the weights of the updated charging piles of each charging pile to the importance unit to recalculate the allocation priority index of each charging pile and re-obtain the preferred charging piles for the next vehicle to be charged;

[0024] And according to different response moments, generate the allocation priority index of each charging pile within the corresponding response moment in sequence, and sort the allocation priority index of each charging pile within the corresponding response moment according to the time sequence to obtain the iteration result of different response moments.

[0025] A charging pile includes a main control module, at least one detection module, and at least one power supply module provided in a cabinet;

[0026] The detection module is configured to detect the operation characteristic group of the charging pile and send the operation characteristic group to the main control module;

[0027] The main control module is configured to generate a corresponding determination signal based on the operation characteristic group and send the determination signal to the power supply module;

[0028] The power supply module is configured to supply power to the charging pile and perform dynamic load balancing based on the determination signal.

[0029] Beneficial effects:

[0030] In terms of dynamic load distribution, the distribution module can fully consider multi-dimensional factors such as voltage, current, and power by receiving the operating characteristic groups of each charging pile, generating a set of characteristic weights using scientific algorithms, and simulating the importance judgment logic based on the characteristic priorities. It can change the limitations of traditional static or empirical distribution methods, achieve dynamic and precise allocation of charging pile resources, effectively avoid the situation where some charging piles are overloaded while some are idle, improve the overall utilization efficiency of charging piles, reduce the waiting time of users, and optimize the charging experience. During the load distribution process, the process module collects multi-source data in real time, comprehensively applies technologies such as the random forest algorithm and the hidden Markov model to comprehensively analyze risks such as charging pile failures, communication interruptions, charging demands, and power supplies, and constructs a comprehensive risk index in combination with the evidence theory, enabling the system to perceive potential risks in advance. Compared with traditional single-risk assessment models, it enhances the risk warning and response capabilities. The update module flexibly generates a judgment signal based on the value of the comprehensive risk index. When the risk exceeds the threshold, it timely updates the charging pile weights and dynamically adjusts the load distribution strategy to ensure that the system always operates efficiently and stably in a complex and changeable operating environment.

[0031] By applying the Pearson correlation coefficient algorithm, quantitatively analyze the correlation between various features and the device health value, obtain the Pearson correlation coefficient, objectively measure the impact degree of features on device health based on data, avoid the one-sidedness of subjective judgment, and re-quantify the weights of various initially calculated features, shifting from vertical single-category data analysis to horizontal multi-category data analysis; construct a judgment matrix in combination with the Saaty 1–9 scale method, transform the abstract comparison of feature importance into a computable numerical relationship, and systematically present the relative importance between features in matrix form, making complex multi-factor evaluations more logical and standardized. Compared with the traditional method of simply allocating based on a single indicator, this method effectively avoids the unbalanced use of charging piles, improves the overall resource utilization efficiency, ensures the stable and efficient operation of the charging pile system, and brings a better quality and more convenient charging service experience to users.

[0032] The judgment unit and the update unit in the update module cooperate closely to dynamically adjust the charging pile weights according to the comparison between the comprehensive risk index and the preset threshold. When the comprehensive risk index does not exceed the threshold, the original weight is maintained to ensure the stable operation of the system; once the comprehensive risk index exceeds the standard, an update instruction is immediately triggered. The update unit non-linearly adjusts the charging pile weights based on the comprehensive risk index value, enabling the weight adjustment to quickly respond to high-risk situations and remain relatively stable when the risk is low, ensuring the scientificity and flexibility of the weight adjustment. Description of the Drawings

[0033] Figure 1 It is a module diagram of the intelligent distribution system for charging piles based on dynamic load balancing of the present invention. Detailed Embodiment

[0034] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

[0035] Embodiment 1

[0036] Please refer to Figure 1 , the present invention provides a smart allocation system for charging piles based on dynamic load balancing, including

[0037] An allocation module, which is used to receive the operation feature groups of each charging pile at the first response moment, generate a feature weight set, and simulate the importance judgment logic in the feature weight set according to the feature priority to achieve dynamic load allocation of the charging piles;

[0038] A process module, which is used to receive each source data in real time during the dynamic load allocation process, based on each source data, and at the second response moment, perform multi-source analysis, check the suddenness of charging pile failures, analyze communication interruptions, predict charging demands and power supplies, and combine the evidence theory (D-S theory) to construct a comprehensive risk index;

[0039] By monitoring risks in real time and adjusting weights, it is possible to avoid charging piles with potential faults or risks in a timely manner, reduce charging interruptions caused by emergencies, and improve the overall reliability of the system.

[0040] An update module, which is used to generate a determination signal according to the value of the comprehensive risk index, and the determination signal includes an update instruction and an iteration result.

[0041] In the embodiments of the present invention, the allocation module determines each feature weight and priority through scientific calculations such as the Pearson correlation coefficient and the Saaty 1-9 scale method by receiving the operation feature groups of the charging piles, and then simulates the human decision-making logic to achieve dynamic load allocation. For example, during peak hours, this module can give priority to allocating charging tasks to charging piles with low load and excellent performance according to features such as the connection number, power, and device health value of the charging piles, reducing the overload of some charging piles and the idleness of some others.

[0042] The process module collects multi-source data in real time during the load allocation process, uses advanced technologies such as the random forest algorithm and the hidden Markov model to analyze the charging pile failures, network communications, charging demands, and power supplies respectively, and constructs a comprehensive risk index with the help of the evidence theory. For example, when a large-scale event suddenly occurs in a certain area, resulting in a sharp increase in charging demand, this module can quickly predict the change in demand, and at the same time, combined with the power supply situation of the power grid, accurately evaluate the risks of each charging pile.

[0043] The update module generates a determination signal based on the comprehensive risk index. When the comprehensive risk index exceeds the threshold, it issues an update instruction in a timely manner, adjusts the weights of the charging piles through weight adjustment, and realizes dynamic optimization of the weights. For example, when the health status of a certain charging pile device deteriorates and the fault risk increases, the value of its comprehensive risk index increases. The update module will reduce the weight of this charging pile, reduce the allocation of new tasks, and ensure the stable operation of the overall system.

[0044] Embodiment 2

[0045] Please refer to Figure 1 , specifically: The allocation module includes a first receiving unit, a weight generation unit, and an importance unit;

[0046] The first receiving unit is used to preset the response interval, and at the first response moment, receive the operation feature groups of each charging pile through a plurality of groups of sensors. The operation feature groups include but are not limited to voltage, current, power, connection number, charging speed, fault duration, operation duration, and the number of charging guns; and preprocess the operation feature groups, eliminate duplicate data and fill in missing values to obtain the preprocessed operation feature groups.

[0047] The weight generation unit is used to extract the maximum and minimum values of various features in the preprocessed operation feature groups, and combine their proportional relationship with the device health value of the corresponding charging pile to define the weights of various processed features. Specifically: calculate the ratio of each feature in the corresponding charging pile to the feature with the largest value among all its features, and the calculation result is the weight of each processed feature; for better illustration, the algorithm is as follows: ; where is the feature weight of feature in the i-th charging pile, is the number of feature in the i-th charging pile, is the maximum value of feature z among all charging piles, and n is the total number of charging piles;

[0048] Extract the maximum and minimum values of various features in the preprocessed operation feature groups, and combine their proportional relationship with the device health value of the corresponding charging pile. Specifically:

[0049] The purpose of calculating the ratio of various features to the device health value of the corresponding charging pile in advance is to align with the device health value of the corresponding charging pile in subsequent weight acquisition, which is equivalent to a reference feature.

[0050] The above calculation formula shows a direct proportional relationship between the corresponding feature in the numerator and the device health value; if it is an inverse proportional relationship, the content in the numerator and denominator of this calculation formula will be interchanged.

[0051] Among them, the device health value H represents the ratio of the non-fault duration to the operation duration of the corresponding charging pile in the historical period.

[0052] In this embodiment, the allocation module and its subunits in the present invention achieve precise quantification of the operation characteristics of charging piles and scientific allocation of resources through systematic data processing and weight calculation, effectively improving the refinement degree of charging pile management and resource utilization efficiency.

[0053] As the front-end sentry for data acquisition, the first receiving unit actively collects the operation characteristic group data of each charging pile at the first response moment by presetting the response interval, covering multi-dimensional information such as voltage, current, and power, and preprocesses the data to eliminate duplicate data and fill in missing values. For example, in a certain charging pile cluster, the first receiving unit timely discovers the duplicate recording of current data caused by sensor abnormalities in a charging pile. After processing, the accuracy of the data is ensured, providing a reliable basis for subsequent analysis.

[0054] The weight generation unit is like an intelligent analyst. Based on the operation characteristic group data processed by the first receiving unit, it extracts the maximum and minimum values of various characteristics and performs normalization processing in combination with the device health value to determine the weights of various characteristics. Taking the device health value as an example, if the proportion of the non-fault duration to the operation duration of a certain charging pile is as high as 90%, it indicates that its health condition is good. In the weight calculation, the weight of the characteristic (such as the charging speed) that is directly proportional to it will increase accordingly. Suppose a charging pile has a fast charging speed and a high device health value, and the calculated weight of its charging speed characteristic is 0.8; while another charging pile has a low device health value due to frequent failures. Even if the charging speeds are the same, the weight of its charging speed characteristic may only be 0.3. This differential weight setting enables the system to preferentially allocate charging tasks to charging piles with better performance and more stable states, avoiding resource misallocation, improving the overall service quality and equipment utilization rate, and ensuring the dynamic and efficient optimal allocation of charging pile resources in a complex and changeable operation environment.

[0055] Embodiment 3

[0056] Please refer to Figure 1 , specifically: To make the calculation of various characteristic weights more logical and understandable, a calculation method combining priority sorting and normalization is adopted to convert each individual weight into a comprehensive weight;

[0057] The importance unit is used to analyze the correlation between various characteristics and the device health value using the algorithm of the Pearson correlation coefficient, so as to respectively obtain the Pearson correlation coefficients between various characteristics and the device health value, and in combination with the Saaty 1–9 scale method, then compare the magnitude relationships between the Pearson correlation coefficients to construct a judgment matrix, where each element in the judgment matrix reflects the relative ratio between two Pearson correlation coefficients;

[0058] Further, for better illustration, the following is the expression form of the judgment matrix A: ;

[0059] where m is the number of comparison pairs, which is equal to the total number of features. is the relative importance ratio of the first Pearson correlation coefficient to the second Pearson correlation coefficient; Example: If = 3, then = 1 / 3;

[0060] The device health value is measured by the ratio of the historical non-fault duration to the operation duration, which intuitively reflects the operation stability of the charging pile.

[0061] Based on the judgment matrix, the power method is used to calculate the maximum eigenvalue and the corresponding eigenvector of the judgment matrix, and the eigenvector corresponding to the maximum eigenvalue is normalized to obtain the weights of the corresponding features in each Pearson correlation coefficient. By arranging the weights of the corresponding features in each Pearson correlation coefficient r in ascending order, the priority of various features within the operation feature group is determined, and the weights of various features defined in the priority and weight generation unit are combined. Combining with the hierarchical fusion method, the final comprehensive weight of the corresponding charging pile is calculated and obtained.

[0062] In the judgment matrix, the maximum eigenvalue is the eigenvalue with the largest absolute value among the eigenvalues of the matrix.

[0063] The power method is an iterative method for calculating the dominant eigenvalue (i.e., the eigenvalue with the largest absolute value) of a matrix and its corresponding eigenvector. Its basic idea is to continuously left-multiply an initial vector by the matrix to obtain a vector sequence. As the number of iterations increases, the vector sequence will gradually converge to the dominant eigenvector of the matrix, and the dominant eigenvalue can be obtained through certain calculations.

[0064] Further, based on the judgment matrix, the power method is used to calculate the maximum eigenvalue and the corresponding eigenvector of the judgment matrix, specifically including:

[0065] Initialization: Select a non-zero initial vector x0, which can be a vector all equal to 1, and set the number of iterations u = 0.

[0066] Iterative calculation: Calculate , find the element with the largest absolute value in y(u + 1), and calculate the vector for the next iteration.

[0067] Convergence judgment: Judge whether the convergence condition is satisfied. It can be judged whether the absolute value of the difference between the vectors obtained from two adjacent iterations is less than a pre-set threshold. If the convergence condition is satisfied, stop the iteration.

[0068] Calculate the maximum eigenvalue: When the iteration converges, the corresponding element is the maximum eigenvalue of matrix A.

[0069] Through hierarchical fusion, the importance differences of various weight factors are reflected, and each individual weight is mapped to the same value range (0 - 1), eliminating the influence of the original numerical sizes of different weights, making the weights comparable and fusible.

[0070] According to the final comprehensive weight of the corresponding charging pile, dynamic load distribution is carried out for the charging piles, specifically as follows:

[0071] By calculating the ratio of the connection number of the corresponding charging pile to its corresponding final comprehensive weight, the distribution priority index of each charging pile is calculated;

[0072] Furthermore, for better illustration, it is reflected by the formula: ; where is the distribution priority index of the i-th charging pile, is the connection number of the i-th charging pile, is the final comprehensive weight of the i-th charging pile;

[0073] When there are vehicles waiting to be charged, compare the distribution priority indexes of each charging pile in the charging station, and select the charging pile corresponding to the smallest distribution priority index as the preferred charging pile for the vehicle waiting to be charged this time.

[0074] Among them, the corresponding eigenvector is the vector corresponding to the maximum eigenvalue, and its components represent the relative importance weights of the Pearson correlation coefficients in the whole;

[0075] Calculate the correlation between each performance index through the Pearson correlation coefficient.

[0076] In this embodiment, the importance unit uses the Pearson correlation coefficient algorithm to deeply analyze the correlation between various operation characteristics of the charging pile and the equipment health value, quantitatively obtain the Pearson correlation coefficient, so as to objectively reflect the influence degree of each characteristic on the equipment health. For example, if the Pearson correlation coefficient between the current characteristic and the equipment health value is close to 1, it indicates that the current magnitude is highly positively correlated with the equipment health status and needs to be considered key in load distribution. Combining with the Saaty1–9 scale method, compare the Pearson correlation coefficients to construct a judgment matrix, then use the power method to calculate the maximum eigenvalue and eigenvector, and perform normalization processing to obtain the weights of each characteristic, and then determine the characteristic priority. Based on this, combined with the characteristic weights defined by the weight generation unit, the final comprehensive weight of the charging pile is calculated by the hierarchical fusion method.

[0077] In actual allocation, the importance unit calculates the allocation priority index to associate the number of charging pile connections with the final comprehensive weight, presenting the load allocation priority of each charging pile in a numerical form. For example, in a certain charging station, one charging pile has a small number of connections but a high comprehensive weight, and the calculated allocation priority index value is small. Another charging pile has a large number of connections but a low comprehensive weight, and the allocation priority index value is large. The system will preferentially allocate the vehicles to be charged to the first charging pile to avoid overloading the second charging pile and achieve load balancing.

[0078] In this process, the first receiving unit is responsible for collecting and preprocessing the operation data of the charging piles; the weight generation unit calculates the feature weights of the data; and the importance unit further determines the feature priorities and comprehensive weights. The three work together to form a complete link from data processing to strategy execution for the dynamic load allocation of charging piles, effectively solving the problem of resource imbalance easily caused by traditional allocation methods.

[0079] Embodiment 4

[0080] Please refer to Figure 1 , specifically: The process module includes a second receiving unit and a multi-source analysis unit;

[0081] The second receiving unit is used to receive each source of data in real time based on the importance unit. Each source of data includes the operation feature group, network communication feature group, and power supply feature group of the charging pile; among them, the network communication feature group includes various features during network communication; the power supply feature group includes the real-time remaining capacity of the power grid, the real-time power demand of the charging pile, and the upper limit of regional power distribution.

[0082] The multi-source analysis unit is used to construct a fault prediction model using the random forest algorithm according to the operation feature group of the charging pile at the second response moment. The operation feature group of the charging pile is divided into a training set and a test set. The training set is used to train the fault prediction model, adjust the parameters of the model to optimize the performance of the model, and then the test set is used to verify the trained fault prediction model, evaluate the accuracy rate and recall rate of the model, and input the real-time monitored operation feature group of the charging pile into the trained fault prediction model to enable it to learn the pattern of fault occurrence. The fault prediction model outputs the probability of each charging pile having a fault in a future period of time;

[0083] According to the network communication feature group, the Hidden Markov Model (HMM) is used to predict the network state. The network state is divided into different hierarchical states, including normal, unstable, and interrupted. The signal strength range corresponding to each hierarchical state is determined. Then, the received network communication feature group is used as the observation sequence, and the network state is used as the hidden state. The HMM is trained through the known observation sequence and the corresponding hidden state to learn the state transition probability and the observation probability. The forward-backward algorithm is used to calculate the probability distribution of the hidden state given the model parameters and the observation sequence, and then the probability of future network interruption is obtained.

[0084] The commonly used training algorithm is the Baum-Welch algorithm, which is an iterative algorithm based on maximum likelihood estimation for estimating the parameters of the Hidden Markov Model.

[0085] According to the operation feature group of the charging pile, historical charging demand data is extracted. The time series prediction algorithm (such as the Prophet algorithm) is used to predict the charging demand in the future period. Combining with cluster analysis, high-demand regions and time periods are divided to calculate the risk probability. Specifically, the predicted charging demand is subtracted from the maximum carrying capacity of the corresponding charging station, and according to the numerical state of the calculation result, a ratio calculation is performed with the maximum value of the predicted demand in all regions, so as to finally obtain the risk probability.

[0086] Furthermore, for better illustration, it is further demonstrated through an algorithm: ; where is the risk probability of the i-th charging pile facing the charging demand, is the predicted charging demand in the region where the i-th charging pile is located at time t, is the maximum carrying capacity of the i-th charging pile, that is, the maximum load that the charging pile can provide charging services for vehicles simultaneously; is the maximum value of the predicted demand in all regions; this formula represents the risk probability of large-scale centralized charging demand by calculating the difference between the predicted demand and the maximum carrying capacity of the charging pile and normalizing it to the range of 0 to 1.

[0087] The logic for obtaining the risk probability is: First, calculate , obtain the difference between the predicted demand and the maximum carrying capacity of charging pile i at time t. If the difference is greater than 0, it indicates that the predicted demand exceeds the maximum carrying capacity of the charging pile, and there is an overload risk. At this time, take this difference as the numerator; if the difference is less than or equal to 0, it means that the predicted demand is within the carrying range of the charging pile, and there is no overload risk. At this time, the numerator is taken as 0, and then divide the numerator by the maximum value of the predicted demand in all regions to normalize the risk probability to the interval of 0 to 1, so as to facilitate the comparison and analysis of the risk levels of different charging piles facing large-scale centralized charging demands.

[0088] Cluster analysis: Divide the charging areas into different clusters according to geographical location and charging pile distribution density factors. For each cluster, analyze the distribution of its historical charging demands to determine high-demand areas and time periods. For example, the K-means clustering algorithm can be used to divide the areas into K clusters according to the similarity of charging demands.

[0089] According to the power supply characteristic group, extract the real-time remaining capacity Er of the power grid, the real-time power demand Ed of the charging pile, and the upper limit Ea of regional power distribution. Through ratio calculation, obtain the supply risk probability EF, specifically: ; where is the supply risk probability of the i-th charging pile, is the real-time power demand of the i-th charging pile, that is, the electric power actually consumed by this charging pile at the current moment. The upper limit Ea of regional power distribution refers to the maximum power capacity that can be allocated to the charging pile in the corresponding charging pile area; the real-time remaining capacity Er of the power grid is the remaining power capacity available for distribution in the power grid at the current moment;

[0090] The logic for obtaining the supply risk probability is as follows: First calculate , that is, the difference between the sum of the real-time power demands of all charging piles and the upper limit of regional power distribution, to measure whether the total power demand of the charging piles in this area exceeds the allocable power upper limit. If the difference is greater than 0, it indicates insufficient power supply and there is a risk. At this time, take this difference as the numerator; if the difference is less than or equal to 0, it means that the power supply can meet the demand and there is no risk, and the numerator is taken as 0. Finally, divide the numerator by the real-time remaining capacity of the power grid to normalize the supply risk probability to the range of 0 to 1, so as to obtain the risk probability of power supply shortage in the area where the i-th charging pile is located, which is used to evaluate the tightness of the power supply in this area.

[0091] In the embodiment of the present invention, the second receiving unit is responsible for collecting multi-source data such as the charging pile operation characteristic group, the network communication characteristic group, and the power supply characteristic group in real time, providing rich and timely data support for the system operation. Just like the information collector of the system, it ensures the comprehensiveness and real-time nature of the data.

[0092] The multi-source analysis unit is like the intelligent brain of the system, deeply mining the collected data. It uses the random forest algorithm to build a fault prediction model, optimizes the model performance through training and validation, and can accurately predict the future fault probability of the charging pile. For example, by analyzing features such as voltage fluctuations and operation duration, it can predict in advance that a certain charging pile may malfunction within 3 days due to component aging, facilitating the maintenance personnel to conduct pre-maintenance and reducing the impact of sudden failures on users.

[0093] The hidden Markov model is used to predict the network state, which can accurately estimate the network interruption risk. For example, during large-scale events, it can judge in advance the probability of the surrounding charging pile network being interrupted due to the surge in users, and make preparations for network optimization or alternative plans in advance. The time series prediction algorithm combined with clustering analysis is used to predict the charging demand, which can effectively identify high-demand areas and time periods. For example, around scenic spots during holidays, it can accurately predict the charging peak to avoid overcrowding of charging piles.

[0094] Calculating the supply risk probability based on the power supply characteristic group can monitor the power supply and demand balance in real time. For example, when the remaining capacity of a certain regional power grid is insufficient, the system can timely adjust the power distribution of the charging piles to prevent power overload. Through the comprehensive analysis of multi-source data, the system can comprehensively grasp the operation risks of the charging piles, provide a scientific basis for subsequent dynamic load distribution and weight adjustment, effectively improve the stability and reliability of the charging pile system, and ensure the charging experience of users.

[0095] Embodiment 5

[0096] Please refer to Figure 1 , specifically: for better illustration, here, the probability of each charging pile malfunctioning in a future period of time, the probability of future network interruption, the risk probability, and the supply risk probability are added symbol tags, which are F, G, DF, and EF respectively;

[0097] The multi-source analysis unit is also used to, according to the evidence theory, take the probability F of each charging pile malfunctioning in a future period of time, the probability G of future network interruption, the risk probability DF, and the supply risk probability EF as independent evidence sources. For each independent evidence source, determine the basic probability assignment according to its own characteristics and data distribution to obtain the basic probability assignment function Y of the corresponding independent evidence source, and synthesize the basic probability assignment functions Y multiple times according to the D-S synthesis rule to calculate and obtain the comprehensive risk index CRI. Specifically: ; where is the comprehensive risk index of the i-th charging pile; , , and are the basic probability assignment functions of the corresponding independent evidence sources in the i-th charging pile respectively, is the synthesis operator in the D-S synthesis rule.

[0098] For each independent evidence source, the basic probability assignment is determined according to its own characteristics and data distribution. For example, for the probability F of a charging pile malfunctioning, based on the statistical distribution of historical failure data and the current operating state of the charging pile, it can be mapped to a basic probability assignment function. Assuming the failure risk is divided into three levels: low, medium, and high, the basic probability is assigned according to the probability that the value of the malfunction probability F falls into different levels. For example, if the malfunction probability F is in the range of (0 - 0.3), the probability of belonging to the low-risk level is 0.7, the probability of belonging to the medium-risk level is 0.2, and the probability of belonging to the high-risk level is 0.1.

[0099] The D-S combination rule is a method for fusing the basic probability assignments of multiple evidence sources. It can combine the basic probability assignment functions of different evidence sources to obtain a comprehensive basic probability assignment function, thereby achieving information fusion. Through the operator, , , and are gradually combined to finally obtain the basic probability assignment function of the comprehensive risk index , thus comprehensively considering various risk factors and integrating the information of multiple independent evidence sources.

[0100] In the Naive Bayes algorithm, the posterior probability of each category can be calculated based on the training data to determine the value of the basic probability assignment function.

[0101] The update module includes a determination unit and an update unit;

[0102] The determination unit is used to extract the final comprehensive weight of the corresponding charging pile in the importance unit as the weight at the second response moment when the comprehensive risk index does not exceed the pre-set risk threshold. Otherwise, an update instruction is generated.

[0103] The update unit is used to determine the update instruction and adjust the final comprehensive weight of the corresponding charging pile in the importance unit by means of the Sigmoid function based on the final comprehensive weight of the corresponding charging pile in the importance unit, so as to update the final comprehensive weight of the corresponding charging pile;

[0104] Furthermore, for better illustration, through the algorithm, the final comprehensive weight of the corresponding charging pile is obtained for update; is the weight of the i-th charging pile after update, is the comprehensive risk index of the i-th charging pile, is the slope control parameter used to control the steepness of the curve. The larger k is, the steeper the function is at the inflection point, and the more drastic the weight change is; is a position control parameter used to control the position of the inflection point, that is, the CRI value when the weight starts to change significantly; when the comprehensive risk index is larger, it indicates that the comprehensive risk of the charging pile is higher. Accordingly, it is necessary to reduce its weight to reduce the possibility of assigning new tasks to this charging pile.

[0105] When the comprehensive risk index tends to 1, tends to 1, and at this time tends to 0 (the weight is significantly reduced);

[0106] When the comprehensive risk index approaches 0, tends to 0, and at this time tends to the final comprehensive weight Wz (the weight is basically unchanged).

[0107] By collecting historical data, including the usage situation, fault records, user feedback, etc. of the charging piles under different comprehensive risk indexes, analyze the relationship between the comprehensive risk index and the actual business effect. By observing the data distribution and change trend, find the comprehensive risk index value at which the weight starts to change significantly as a reference for the position control parameter. For the determination of the slope control parameter, different slope control parameter values can be tried, draw the curve of the adjustment coefficient changing with the comprehensive risk index, and observe which k value can make the curve better conform to the business expectation, that is, the weight change matches the actual risk situation.

[0108] Iteration result: Return the weights of the updated charging piles of each charging pile to the importance unit to recalculate the allocation priority index of each charging pile and re-obtain the preferred charging piles for the next vehicle to be charged;

[0109] And according to different response times, generate the allocation priority indexes of each charging pile within the corresponding response time in turn, and sort the allocation priority indexes of each charging pile within the corresponding response time according to the time sequence to obtain the iteration results of different response times.

[0110] In the embodiment of the present invention, the multi-source analysis unit uses the evidence theory (D-S theory), takes the charging pile failure probability, network interruption probability, charging demand risk probability, and power supply risk probability as independent evidence sources, and through the basic probability assignment function and the D-S synthesis rule, integrates the multi-dimensional risk information into a comprehensive risk index. This method avoids the one-sidedness of single risk assessment and can more comprehensively and accurately reflect the actual risk situation of the charging pile. For example, when the failure probability of a certain charging pile is low, but the power supply risk probability is high, the comprehensive risk index will consider both of them comprehensively to avoid underestimating the potential risk of this charging pile.

[0111] The determination unit decides whether to trigger a weight update instruction based on the comparison between the comprehensive risk index and the preset threshold. When the comprehensive risk index does not exceed the threshold, the original weight is continued; otherwise, the update process is initiated to ensure that the system can respond to risk changes in a timely manner.

[0112] Based on the update instruction, the update unit dynamically adjusts the charging pile weight through a formula. Among them, the slope control parameter and the position control parameter can flexibly adjust the sensitivity and inflection point of the weight change. For example, during the peak charging period, due to insufficient power supply, the comprehensive risk index of a certain charging pile increases, and the system reduces its weight through the update unit to reduce the allocation of new tasks; when the risk decreases, the weight can gradually recover to achieve dynamic balance.

[0113] In summary, the multi-source analysis unit is responsible for integrating multi-dimensional risk information to generate a comprehensive evaluation index; the determination unit is used to judge the risk situation and decide whether to initiate weight update; the update unit dynamically adjusts the charging pile weight according to risk changes, so as to realize the intelligent and dynamic allocation of the charging pile load and ensure the efficient and stable operation of the system.

[0114] The foregoing parameters and the calculated parameters have all been normalized (such as Z-score standardization) to eliminate the influence of different original numerical sizes, making each data comparable and fusible.

[0115] Embodiment 6

[0116] Please refer to Figure 1 , specifically: A charging pile includes a main control module, at least one detection module, and at least one power supply module disposed in a cabinet;

[0117] The detection module is used to detect the operation characteristic group of the charging pile and send the operation characteristic group to the main control module;

[0118] The main control module is used to generate a corresponding determination signal based on the operation characteristic group and send the determination signal to the power supply module;

[0119] The power supply module is used to supply power to the charging pile and perform dynamic load balancing allocation based on the determination signal.

[0120] Although the embodiments of the present invention have been shown and described, for those of ordinary skill in the art, it can be understood that various changes, modifications, substitutions, and variations can be made to these embodiments without departing from the principles and spirits of the present invention. The scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. The intelligent charging pile allocation system based on dynamic load balancing is characterized by: include, The allocation module is used to receive the operation feature group of each charging pile at the first response time, generate a feature weight set, and simulate the importance judgment logic in the feature weight set according to the feature priority to realize dynamic load allocation of the charging pile; The process module is used to receive data from various sources in real time during the dynamic load distribution process. Based on the data from various sources, it performs multi-source analysis at the second response time to check the suddenness of charging pile failures, analyze communication interruptions, predict charging demand and power supply, and build a comprehensive risk indicator based on evidence theory. The update module is used to generate a determination signal according to the value of the comprehensive risk indicator, and the determination signal includes an update instruction and an iteration result.

2. The charging pile intelligent allocation system based on dynamic load balancing according to claim 1 is characterized in that: The allocation module includes a number one receiving unit, a weight generating unit and an importance unit; The first receiving unit is used to pre-set the response interval and receive the operation characteristic group of each charging pile through a plurality of groups of sensors at the first response time. The operation characteristic group includes voltage, current, power, number of connections, charging speed, fault duration, operation duration and number of charging guns; The running feature group is preprocessed to remove duplicate data and fill in missing values ​​to obtain the preprocessed running feature group.

3. The charging pile intelligent allocation system based on dynamic load balancing according to claim 2 is characterized in that: The weight generation unit is used to extract the maximum value of each type of feature in the pre-processed operating feature group, and combine it with the proportional relationship between it and the equipment health value of the corresponding charging pile to define the weights of each type of feature after processing; wherein the equipment health value represents the ratio of the non-fault time and the operating time of the corresponding charging pile in the historical period.

4. The charging pile intelligent allocation system based on dynamic load balancing according to claim 3 is characterized in that: The importance unit is used to use the Pearson correlation coefficient algorithm to analyze the correlation between various features and equipment health values, so as to obtain the Pearson correlation coefficients between various features and equipment health values ​​respectively, and then compare the size relationship between the Pearson correlation coefficients in combination with the Saaty1-9 scaling method to construct a judgment matrix; Based on the judgment matrix, the power method is used to calculate the maximum eigenvalue and the corresponding eigenvector of the judgment matrix, and the eigenvector corresponding to the maximum eigenvalue is normalized to obtain the weight of the corresponding feature in each Pearson correlation coefficient. The weights of the corresponding features in each Pearson correlation coefficient are arranged in ascending order to determine the priority of each feature in the operation feature group. According to the weights of each feature defined in the priority and weight generation unit, combined with the hierarchical fusion method, the final comprehensive weight of the corresponding charging pile is calculated. According to the final comprehensive weight of the corresponding charging pile, the dynamic load distribution of the charging pile is carried out, specifically: The allocation priority index of each charging pile is calculated by calculating the ratio of the number of connections of the corresponding charging piles to their corresponding final comprehensive weights. When there is a vehicle to be charged, the allocation priority indexes of each charging pile in the charging station are compared, and the charging pile corresponding to the smallest allocation priority index is selected as the preferred charging pile for the vehicle to be charged this time.

5. The charging pile intelligent allocation system based on dynamic load balancing according to claim 4 is characterized in that: The process module includes a No. 2 receiving unit and a multi-source analysis unit; The second receiving unit is used to receive various source data in real time based on the importance unit. The source data includes the operation feature group of the charging pile, the network communication feature group and the power supply feature group; among them, the network communication feature group includes various features during network communication; the power supply feature group includes the real-time remaining capacity of the power grid, the real-time power demand of the charging pile and the upper limit of regional power distribution.

6. The charging pile intelligent allocation system based on dynamic load balancing according to claim 5 is characterized in that: The multi-source analysis unit is used to build a fault prediction model using a random forest algorithm according to the operation feature group of the charging pile at the second response time, divide the operation feature group of the charging pile into a training set and a test set, use the training set to train the fault prediction model, adjust the parameters of the model to optimize the performance of the model, and then use the test set to verify the trained fault prediction model, evaluate the accuracy and recall rate of the model, input the operation feature group of the charging pile monitored in real time into the trained fault prediction model to make it learn the mode of fault occurrence, and the fault prediction model outputs the probability of each charging pile failing in the future; according to the network communication feature group, a hidden Markov model is used to predict the network state, and the network state is divided into different level states, and the level state Including normal, unstable and interrupted, and determine the signal strength range corresponding to each level state, then use the received network communication feature group as the observation sequence, and the network state as the hidden state, train the hidden Markov model through the known observation sequence and the corresponding hidden state, learn the state transition probability and observation probability, and predict the probability of future network interruption; according to the operation feature group of the charging pile, extract the historical charging demand data, use the time series prediction algorithm to predict the charging demand in the future time period, and combine cluster analysis to divide the high demand areas and time periods to calculate the risk probability; according to the power supply feature group, extract the real-time remaining capacity of the power grid, the real-time power demand of the charging pile and the upper limit of the regional power distribution, and obtain the supply risk probability through ratio calculation.

7. The charging pile intelligent allocation system based on dynamic load balancing according to claim 6 is characterized in that: The multi-source analysis unit is also used to take the probability of failure of each charging pile in the future, the probability of future network interruption, the risk probability and the supply risk probability as independent evidence sources according to the evidence theory. For each independent evidence source, the basic probability distribution is determined according to its own characteristics and data distribution to obtain the basic probability distribution function of the corresponding independent evidence source, and each basic probability distribution function is synthesized multiple times according to the DS synthesis rule to calculate the comprehensive risk index.

8. The charging pile intelligent allocation system based on dynamic load balancing according to claim 7 is characterized in that: The updating module includes a determining unit and an updating unit; The determination unit is used to extract the final comprehensive weight of the corresponding charging pile in the importance unit as the weight of the second response moment when the comprehensive risk index does not exceed the preset risk threshold, and otherwise generate an update instruction.

9. The charging pile intelligent allocation system based on dynamic load balancing according to claim 8 is characterized in that: An updating unit, used for determining an updating instruction, and adjusting the final comprehensive weight of the corresponding charging pile in the importance unit by a Sigmoid function based on the final comprehensive weight of the corresponding charging pile in the importance unit, so as to update the final comprehensive weight of the corresponding charging pile; As a result of the iteration, the updated weight of each charging pile is returned to the importance unit to recalculate the allocation priority index of each charging pile and re-obtain the preferred charging pile for the vehicle to be charged next time; According to different response times, the allocation priority index of each charging pile in the corresponding response time is generated in sequence, and the allocation priority index of each charging pile in the corresponding response time is sorted in time sequence to obtain iterative results of different response times.

10. A charging pile, using the system according to any one of claims 1 to 9, characterized in that: It includes a main control module, at least one detection module and at least one power supply module arranged in the cabinet; The detection module is used to detect the operation feature group of the charging pile and send the operation feature group to the main control module; The main control module is used to generate a corresponding determination signal based on the operation feature group and send the determination signal to the power supply module; The power supply module is used to supply power to the charging pile and to evenly distribute the dynamic load based on the determination signal.

Images