A big data-based intelligent charging pile energy efficiency operation and maintenance regulation method

Abstract

This invention relates to the field of charging pile operation and maintenance technology, and in particular to a smart charging pile energy efficiency operation and maintenance control method based on big data. It employs an edge-cloud collaborative acquisition mode to obtain multi-source operational data, completes heterogeneous data calibration through standardized preprocessing, and constructs a four-dimensional operation and maintenance feature map. A multi-dimensional coupled evaluation model with configured quantitative calculation formulas is built to normalize and solve three core factors: charging load, grid capacity, and equipment health. Relying on nonlinear cross-coupling logic, it achieves multi-dimensional collaborative judgment of operational status, completing the coordinated matching of charging pile cluster power control and equipment protection. An energy efficiency loss and equipment fault correlation rule system is constructed, enabling energy efficiency anomaly identification and accurate fault location tracing based on multi-dimensional feature matching, and tiered, refined preventative operation and maintenance work orders are pushed out. Furthermore, a full-process big data closed-loop iterative mechanism is established to adaptively update model parameters and operating condition judgment rules, continuously optimizing differentiated control strategies.

Description

Technical Field

[0001] This invention relates to the field of charging pile operation and maintenance technology, and in particular to a smart charging pile energy efficiency operation and maintenance control method based on big data. Background Technology

[0002] Currently, the energy efficiency operation and maintenance management of charging pile clusters in the industry generally adopts a static, single-dimensional management approach. That is, operating condition assessment, charging power adjustment, equipment status protection, and fault operation and maintenance are all executed independently based on single operating parameters and fixed experience thresholds. As a result, it is impossible to carry out integrated quantitative analysis and dynamic collaborative control of the three dimensions of operating information: charging pile cluster operation, distribution network load, and charging equipment status. The operation and maintenance assessments of each dimension are disconnected from each other and cannot adapt to the complex and ever-changing actual operating conditions on site.

[0003] Load regulation based solely on grid-side parameters fails to consider the aging, temperature rise, and health deterioration trends of charging pile equipment. This can easily lead to overloading of equipment with potential hidden problems, accelerating equipment aging and potentially inducing intermittent operational failures. Conversely, maintenance and protection based solely on equipment-side parameters cannot adapt to the dynamic fluctuations in the real-time capacity margin, voltage, and frequency of the distribution network. This can easily result in excessive cluster charging loads and grid fluctuations during peak hours, while also leading to the waste of idle power resources during off-peak hours.

[0004] Furthermore, the static and fixed traditional judgment logic cannot adapt to dynamic operating scenarios such as changes in user charging behavior, continuous equipment aging, and updates to grid dispatch rules. Operational judgment deviations will accumulate over time, continuously reducing the accuracy of cluster energy efficiency control. Moreover, the traditional management and control model cannot rely on multi-dimensional data correlation mining to identify hidden performance anomalies. It can only carry out passive maintenance after the equipment has obvious faults, resulting in insufficient overall operation and maintenance adaptability and management and control accuracy.

[0005] Therefore, it is urgent for technical personnel to solve the above problems. Summary of the Invention

[0006] The purpose of this invention is to provide a smart charging pile energy efficiency operation and maintenance control method based on big data, which aims to solve the problem that the existing charging pile cluster energy efficiency operation and maintenance adopts a static single-dimensional control method, and the multi-dimensional operation information is fragmented and cannot be dynamically coordinated and controlled.

[0007] This invention relates to a smart charging pile energy efficiency operation and maintenance control method based on big data, comprising the following steps: S1. Construct a two-level collaborative data acquisition architecture between the edge and the cloud. Use a hierarchical and differentiated sampling mode for charging pile cluster operation data, power grid operation data and environmental status data. Set differentiated acquisition frequencies for core operating parameters and auxiliary status parameters. After filtering and screening the acquired raw data, upload it to the cloud big data platform through the edge gateway in an encrypted manner. S2. The cloud-based big data platform performs standardized preprocessing on the received multi-source heterogeneous data, completes the removal of abnormal data, the completion of missing data and the temporal and spatial alignment. Based on the preprocessed standardized data, it constructs a four-dimensional operation and maintenance feature map that includes electrical operation characteristics, equipment health characteristics, power grid carrying capacity characteristics and spatiotemporal environmental characteristics, so as to realize the structured association and unified representation of multi-source heterogeneous data. S3. Establish a multi-dimensional coupled evaluation model to normalize and quantify the regional charging load factor, grid carrying capacity factor, and equipment health factor; among which, the formula for calculating the regional charging load factor is: ; In the formula: Real-time total charging power of the regional charging pile cluster; The rated maximum allowable charging power for the regional charging pile cluster; The formula for calculating the power grid carrying capacity factor is: ; In the formula: This represents the remaining available capacity of the distribution network. This refers to the rated access capacity of the distribution network. , These are the grid voltage and frequency deviation, respectively. , These are the maximum allowable voltage and frequency deviation threshold of the power grid, respectively. , , This is the weighting coefficient for the power grid's carrying capacity, and the sum of the weights is 1; The formula for calculating the equipment health factor is: ; In the formula, Real-time temperature of the charging pile power module; Based on ambient temperature; The module's maximum withstand temperature; The real-time power loss rate of the equipment; This refers to the equipment aging and degradation coefficient. , , This is the weighting coefficient for equipment health, and the sum of the weights is 1. The multidimensional coupling comprehensive evaluation value is calculated using a nonlinear cross-coupling formula. The formula is as follows: ; In the formula, , , The three types of factors are based on coupling weights, and the sum of the weights is 1. This is a cross-coupling correction coefficient; S4. Based on the differentiated operating conditions output by the multi-dimensional coupled evaluation model, match the preset energy efficiency operation and maintenance control strategy, dynamically adjust the output power of each charging pile in the cluster according to the overall operating status of the charging pile cluster corresponding to different operating conditions, and constrain the overall access load of the cluster; at the same time, determine the operating health status of the charging pile based on the equipment health factor, and perform power derating protection operation on the charging pile equipment determined to be in a sub-healthy state. S5. Pre-build a rule system for associating charging pile energy efficiency loss with equipment fault types. Compare real-time operating energy efficiency loss data with preset standard energy efficiency loss data to identify abnormal equipment energy efficiency states caused by non-human factors and non-operating condition fluctuations. Based on the constructed four-dimensional operation and maintenance feature map, associate and match electrical, equipment, power grid, and spatiotemporal multi-dimensional features to complete the source tracing of abnormal fault locations. Based on the fault location, fault type, and abnormality level obtained from the source tracing, automatically generate corresponding level of equipment preventive operation and maintenance work orders and push them to terminal equipment for operation and maintenance scheduling. S6. Real-time collection of equipment energy efficiency regulation and operation data, equipment fault operation and maintenance closed-loop data, and operating condition matching deviation data of multi-dimensional coupled evaluation model to build a full-process big data closed-loop iterative mechanism; based on changes in regional charging behavior, equipment aging status updates, and changes in power grid operation rules, adaptively update the weight parameters, coupling correction parameters, and operating condition judgment rules of the multi-dimensional coupled evaluation model, and iteratively optimize the energy efficiency operation and maintenance regulation strategies corresponding to each operating condition.

[0008] As a further improvement to the technical solution disclosed in this invention, in S1, the core operating parameters include the charging power of the charging pile, the bus voltage, the output current, and the module temperature, and the sampling period is 1 to 3 seconds; the auxiliary status parameters include the ambient temperature and humidity, the equipment standby status, and the time period indicator, and the sampling period is 30 to 60 seconds.

[0009] As a further improvement to the technical solution disclosed in this invention, in S3, , , The values ​​range from 0.2 to 0.6. , , The values ​​range from 0.2 to 0.6. , , The values ​​range from 0.25 to 0.45. The value ranges from 0.1 to 0.3; based on the comprehensive evaluation value Three operating conditions are classified: 0≤ ≤0.35 indicates a high-load pressure condition; 0.35 < ≤0.75 indicates a stable and adaptable operating condition; 0.75 < ≤1 indicates a low-end or surplus operating condition.

[0010] As a further improvement to the technical solution disclosed in this invention, in S4, under high-load pressure conditions, the charging pile cluster is subject to zoned flexible current limiting control to dynamically allocate the maximum output power of a single pile; under stable adaptation conditions, the rated power of the charging piles is maintained in steady-state operation; under low-load surplus conditions, the power limit of the charging piles is lifted; and the equipment health factor... When the value is below 0.6, the charging pile is determined to be in a sub-healthy state, and a power derating protection of 10% to 30% is implemented.

[0011] As a further improvement to the technical solution disclosed in this invention, in S5, the construction steps of the association rule system include: collecting historical energy efficiency loss data, historical fault record data, and four-dimensional operation and maintenance feature data of the charging pile cluster as sample datasets; cleaning and labeling the sample datasets; establishing the corresponding association relationship between energy efficiency anomaly features and various equipment fault types; generating multiple sets of association rules through frequent itemset mining; setting rule confidence and support thresholds to complete rule filtering; binding different energy efficiency anomaly offsets, feature combinations, corresponding fault locations, fault types, and anomaly levels; and solidifying them into an energy efficiency fault association rule system that can be matched and called in real time.

[0012] As a further improvement to the technical solution disclosed in this invention, in S6, the full-process big data closed-loop iterative mechanism achieves adaptive parameter updates through a quantitative mathematical model, specifically including: Construct a model for calculating operating condition matching deviation: ; In the formula: This represents the operating condition matching deviation value. For multi-dimensional coupled comprehensive evaluation and prediction, This is a comprehensive evaluation value based on actual operation; Construct the iterative formula for the operating condition threshold: ; In the formula: The threshold is determined based on the current operating condition; The updated operating condition determination threshold; This is the threshold correction coefficient; The average deviation of the operating conditions within the cycle; Construct the model parameter iterative update formula: ; In the formula: This represents the set of weight parameters and coupling correction parameters for the multidimensional coupled evaluation model at the current moment. The set of parameters after iterative updates; The learning rate is used for iteration. This is the gradient vector of the model's loss function.

[0013] In practical applications, the smart charging pile energy efficiency operation and maintenance control method based on big data disclosed in this invention can achieve at least the following beneficial technical effects, specifically: 1) By combining data preprocessing methods to complete the standardization and calibration of multi-source heterogeneous data, and by constructing a four-dimensional operation and maintenance feature map, the structured association and unified representation of multi-dimensional data of electrical, equipment, power grid, and spatiotemporal dimensions are realized, achieving deep integration of all-dimensional operation data; at the same time, by building a multi-dimensional coupled evaluation model with complete quantitative calculation formulas, the charging load factor, power grid carrying capacity factor, and equipment health factor are normalized and quantified, and the multi-dimensional state collaborative judgment is completed through nonlinear cross-coupling logic, which can comprehensively cover three core operation scenarios: equipment operation, power grid carrying capacity, and cluster load, and realize multi-dimensional collaborative matching of charging pile cluster power regulation and equipment protection. 2) Based on the energy efficiency loss and equipment failure association rule system, energy efficiency anomaly identification and fault location accurate tracing can be completed based on multi-dimensional feature matching. Based on the hierarchical anomaly level, standardized and refined preventive operation and maintenance work order push and scheduling execution can be realized, realizing the upgrade from passive fault handling to proactive predictive operation and maintenance. At the same time, a full-process big data closed-loop iteration mechanism is constructed, which can collect full-dimensional operation and control data and model deviation data in real time. Combined with the dynamic changes of the scenario, the model weight parameters, coupling correction parameters and operating condition judgment rules are adaptively updated, and the differentiated energy efficiency control strategies corresponding to each operating condition are continuously iterated and optimized, so as to dynamically adapt to complex scenarios such as changes in user charging behavior, equipment aging and upgrading, and adjustments to power grid operation rules. Attached Figure Description

[0014] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0015] Figure 1 This is a flowchart of the smart charging pile energy efficiency operation and maintenance control method based on big data disclosed in this invention. Detailed Implementation

[0016] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present invention, and not all of them. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0017] Currently, the energy efficiency operation and maintenance management of charging pile clusters in the industry generally adopts a static, single-dimensional management approach. Operating condition assessment, charging power adjustment, equipment status protection, and fault maintenance operations are all executed independently based on single operating parameters and fixed experience thresholds. This type of management approach can only meet basic charging load adjustment and explicit fault repair needs. It has significant shortcomings in multi-dimensional operation information fusion, dynamic collaborative control, implicit performance anomaly identification, and full-process adaptive iteration. It cannot conduct integrated quantitative analysis of three types of information: charging pile cluster operation, distribution network load, and charging equipment status. The operation and maintenance assessments of each dimension are independent and incompatible with the complex and ever-changing operating conditions on site.

[0018] To address the aforementioned industry pain points, this invention proposes a smart charging pile energy efficiency operation and maintenance control method based on big data. This method is implemented using a two-tiered edge-cloud collaborative architecture and a cloud-based big data platform. The edge gateway is responsible for the hierarchical collection, filtering, and encrypted uploading of multi-source data from the site. The cloud-based big data platform integrates core functions such as data standardization preprocessing, four-dimensional operation and maintenance feature map construction, multi-dimensional coupled evaluation calculation, energy efficiency strategy control, fault tracing and location, and model closed-loop iteration. It serves as the core carrier for the entire process of computation and execution of this method.

[0019] like Figure 1 As shown, the smart charging pile energy efficiency operation and maintenance control method based on big data specifically includes the following steps: S1. Construct a two-level collaborative data acquisition architecture between the edge and the cloud. Use a hierarchical and differentiated sampling mode for charging pile cluster operation data, power grid operation data and environmental status data. Set differentiated acquisition frequencies for core operating parameters and auxiliary status parameters. After filtering and screening the acquired raw data, upload it to the cloud big data platform through the edge gateway in an encrypted manner. S2. The cloud-based big data platform performs standardized preprocessing on the received multi-source heterogeneous data, completes the removal of abnormal data, the completion of missing data and the temporal and spatial alignment. Based on the preprocessed standardized data, it constructs a four-dimensional operation and maintenance feature map that includes electrical operation characteristics, equipment health characteristics, power grid carrying capacity characteristics and spatiotemporal environmental characteristics, so as to realize the structured association and unified representation of multi-source heterogeneous data. S3. Build a multi-dimensional coupling evaluation model, normalize and quantify the regional charging load factor, power grid carrying capacity factor, and equipment health factor, and use a nonlinear cross-coupling formula to calculate the multi-dimensional coupling comprehensive evaluation value. S4. Based on the differentiated operating conditions output by the multi-dimensional coupling evaluation model, match the preset energy efficiency operation and maintenance control strategy, dynamically adjust the output power of each charging pile in the cluster according to the overall operating status of the charging pile cluster corresponding to different operating conditions, and constrain the overall access load of the cluster; at the same time, determine the operating health status of the charging pile based on the equipment health factor, and perform power derating protection operation on the charging pile equipment in the sub-healthy state. S5. Pre-build a rule system for the association between charging pile energy efficiency loss and equipment fault type. Compare the real-time operating energy efficiency loss data of the equipment with the preset standard energy efficiency loss data to identify abnormal energy efficiency status of the equipment caused by non-human factors and non-operating condition fluctuations. Based on the four-dimensional operation and maintenance feature map, match multi-dimensional features to complete the source tracing of abnormal fault points. Automatically generate corresponding level of preventive operation and maintenance work orders according to the fault location, type and level, and push them to the terminal equipment for operation and maintenance scheduling. S6. Real-time collection of three types of data: equipment energy efficiency control, fault operation and maintenance closed loop, and operating condition matching deviation, to build a full-process big data closed-loop iterative mechanism; combined with changes in regional charging behavior, equipment aging status, and power grid operation rules, adaptively update the weight parameters, coupling correction parameters, and operating condition judgment rules of the multi-dimensional coupled evaluation model, and iteratively optimize the energy efficiency operation and maintenance control strategies corresponding to each operating condition.

[0020] To balance data acquisition accuracy and transmission efficiency, S1 performs differentiated sampling on core operating parameters and auxiliary status parameters. The core operating parameters include charging pile charging power, bus voltage, output current, and module temperature, with a sampling period of 1 to 3 seconds; the auxiliary status parameters include ambient temperature and humidity, equipment standby status, and time period identifier, with a sampling period of 30 to 60 seconds.

[0021] To achieve accurate quantitative assessment of multi-dimensional operational status, S3 uses a standardized mathematical model to calculate each factor and the overall assessment value. The specific formula is as follows: Formula for calculating the regional charging load factor: ; In the formula: Real-time total charging power of the regional charging pile cluster; The rated maximum allowable charging power for the regional charging pile cluster; The formula for calculating the power grid carrying capacity factor is: ; In the formula: This represents the remaining available capacity of the distribution network. This refers to the rated access capacity of the distribution network. , These are the grid voltage and frequency deviation, respectively. , These are the maximum allowable voltage and frequency deviation threshold of the power grid, respectively. , , This is the weighting coefficient for the power grid's carrying capacity, and the sum of the weights is 1; The formula for calculating the equipment health factor is: ; In the formula, Real-time temperature of the charging pile power module; Based on ambient temperature; The module's maximum withstand temperature; The real-time power loss rate of the equipment; This refers to the equipment aging and degradation coefficient. , , This is the weighting coefficient for equipment health, and the sum of the weights is 1. The multidimensional coupling comprehensive evaluation value is calculated using a nonlinear cross-coupling formula. The formula is as follows: ; In the formula, , , The three types of factors are based on coupling weights, and the sum of the weights is 1. This is a cross-coupling correction coefficient; Based on this, S3 clarifies the value range of each parameter and the rules for classifying operating conditions: , , The values ​​range from 0.2 to 0.6. , , The values ​​range from 0.2 to 0.6. , , The values ​​range from 0.25 to 0.45. The value ranges from 0.1 to 0.3; based on the comprehensive evaluation value Three operating conditions are classified: 0≤ ≤0.35 indicates a high-load pressure condition; 0.35 < ≤0.75 indicates a stable and adaptable operating condition; 0.75 < ≤1 indicates a low-end or surplus operating condition.

[0022] Based on the above operating condition classification results, S4 performs differentiated control actions for different operating conditions and equipment states. Under high-load pressure conditions, it implements zoned flexible current limiting control for the charging pile cluster, dynamically allocating the maximum output power of a single pile; under stable adaptation conditions, it maintains the steady-state operation of the charging pile's rated power; under low-off-peak surplus conditions, it lifts the power limit on the charging pile; when the equipment health factor H is below 0.6, the charging pile is judged to be in a sub-healthy state, and a 10% to 30% power derating protection is implemented.

[0023] To achieve accurate correlation between energy efficiency anomalies and equipment failures, S5 constructs a correlation rule system according to a specific process. First, it collects historical energy efficiency loss data, historical failure record data, and four-dimensional operation and maintenance feature data of the charging pile cluster as a sample dataset. The sample dataset is cleaned and labeled to establish the correspondence between energy efficiency anomaly features and failure types. Then, multiple sets of correlation rules are generated through frequent itemset mining. Confidence and support thresholds are set to complete rule filtering. Different energy efficiency anomaly offsets and feature combinations are bound to the corresponding failure locations, types, and levels. Finally, a real-time energy efficiency failure correlation rule system is solidified.

[0024] To achieve adaptive iteration of the model and control strategy, S6 uses a quantitative mathematical model to complete parameter updates for the entire closed-loop iteration process. First, a working condition matching deviation calculation model is constructed: Construct a model for calculating operating condition matching deviation: ; In the formula: This represents the operating condition matching deviation value. For multi-dimensional coupled comprehensive evaluation and prediction, This is a comprehensive evaluation value based on actual operation; Construct the iterative formula for the operating condition threshold: ; In the formula: The threshold is determined based on the current operating condition; The updated operating condition determination threshold; This is the threshold correction coefficient; The average deviation of the operating conditions within the cycle; Construct the model parameter iterative update formula: ; In the formula: This represents the set of weight parameters and coupling correction parameters for the multidimensional coupled evaluation model at the current moment. The set of parameters after iterative updates; The learning rate is used for iteration. This is the gradient vector of the model's loss function.

[0025] In practical applications, this method forms a complete closed loop for the energy efficiency operation and maintenance management of charging pile clusters through hierarchical data acquisition, multi-source data preprocessing, four-dimensional feature map association, multi-dimensional coupling evaluation, differentiated energy efficiency control, proactive fault tracing, and full-process closed-loop iteration. The data acquisition stage relies on a two-level architecture and differentiated sampling to balance acquisition accuracy and transmission efficiency; the preprocessing stage completes the normalization of multi-source data and achieves unified association of multi-dimensional data through four-dimensional feature maps; the evaluation stage completes the collaborative judgment of load, grid, and equipment status through a multi-dimensional coupling model; the control stage performs flexible power adjustment based on operating conditions and equipment status; the operation and maintenance stage relies on association rules to achieve anomaly identification and fault tracing; and the iteration stage continuously optimizes the model and strategy through deviation correction to adapt to the dynamic changes of complex scenarios.

[0026] By combining data preprocessing methods to complete the standardization and calibration of multi-source heterogeneous data, and by constructing a four-dimensional operation and maintenance feature map, the structured association and unified representation of multi-dimensional data of electrical, equipment, power grid and spatiotemporal dimensions are realized, and the deep integration of all-dimensional operation data is achieved.

[0027] A multi-dimensional coupled evaluation model with complete quantitative calculation formulas is built to normalize and quantify the charging load factor, grid carrying capacity factor, and equipment health factor. Through nonlinear cross-coupling logic, a multi-dimensional state collaborative judgment is completed, covering three core scenarios: equipment operation, grid carrying capacity, and cluster load. This enables multi-dimensional collaborative matching of charging pile cluster power regulation and equipment protection.

[0028] Based on the rule system that associates energy efficiency loss with equipment failure, the system completes the identification of energy efficiency anomalies and the accurate tracing of fault locations by matching multi-dimensional features. Based on the hierarchical anomaly level, the system completes the standardized and refined push and scheduling of preventive operation and maintenance work orders, realizing the transformation of the operation and maintenance mode from passive fault handling to proactive predictive operation and maintenance.

[0029] We construct a closed-loop iterative mechanism for big data throughout the entire process, collect all-dimensional operation and control data and model deviation data in real time, and adaptively update model parameters and operating condition judgment rules in combination with dynamic changes in scenarios. We iteratively optimize the energy efficiency control strategies corresponding to each operating condition to adapt to complex scenarios such as user charging behavior, equipment aging, and grid rule adjustments.

[0030] Finally, it should be noted that the smart charging pile energy efficiency operation and maintenance control method based on big data described in the above implementation method can be implemented by physical hardware such as computer chips, edge gateways, and cloud servers combined with computer programs, or by software systems that integrate corresponding functions.

Claims

1. A smart charging pile energy efficiency operation and maintenance control method based on big data, characterized in that, Includes the following steps: S1. Construct a two-level collaborative data acquisition architecture between the edge and the cloud. Use a hierarchical and differentiated sampling mode for charging pile cluster operation data, power grid operation data and environmental status data. Set differentiated acquisition frequencies for core operating parameters and auxiliary status parameters. After filtering and screening the collected raw data, upload it to the cloud big data platform through the edge gateway in an encrypted manner. S2. The cloud-based big data platform performs standardized preprocessing on the received multi-source heterogeneous data, completes the removal of abnormal data, the completion of missing data and the temporal and spatial alignment. Based on the preprocessed standardized data, it constructs a four-dimensional operation and maintenance feature map that includes electrical operation characteristics, equipment health characteristics, power grid carrying capacity characteristics and spatiotemporal environmental characteristics, so as to realize the structured association and unified representation of multi-source heterogeneous data. S3. Establish a multi-dimensional coupled evaluation model to normalize and quantify the regional charging load factor, grid carrying capacity factor, and equipment health factor; among which, the formula for calculating the regional charging load factor is: ; In the formula: Real-time total charging power of the regional charging pile cluster; The rated maximum allowable charging power for the regional charging pile cluster; The formula for calculating the power grid carrying capacity factor is: ; In the formula: This represents the remaining available capacity of the distribution network. This refers to the rated access capacity of the distribution network. , These are the grid voltage and frequency deviation, respectively. , These are the maximum allowable voltage and frequency deviation threshold of the power grid, respectively. , , This is the weighting coefficient for the power grid's carrying capacity, and the sum of the weights is 1; The formula for calculating the equipment health factor is: ; In the formula: Real-time temperature of the charging pile power module; Based on ambient temperature; The module's maximum withstand temperature; The real-time power loss rate of the equipment; This refers to the equipment aging and degradation coefficient. , , This is the weighting coefficient for equipment health, and the sum of the weights is 1. The multidimensional coupling comprehensive evaluation value is calculated using a nonlinear cross-coupling formula. The formula is as follows: ; In the formula: , , The three types of factors are based on coupling weights, and the sum of the weights is 1. This is a cross-coupling correction coefficient; S4. Based on the differentiated operating conditions output by the multi-dimensional coupled evaluation model, match the preset energy efficiency operation and maintenance control strategy, dynamically adjust the output power of each charging pile in the cluster according to the overall operating status of the charging pile cluster corresponding to different operating conditions, and constrain the overall access load of the cluster; at the same time, determine the operating health status of the charging pile based on the equipment health factor, and perform power derating protection operation on the charging pile equipment determined to be in a sub-healthy state. S5. Pre-build a rule system for associating charging pile energy efficiency loss with equipment fault types. Compare real-time operating energy efficiency loss data with preset standard energy efficiency loss data to identify abnormal equipment energy efficiency states caused by non-human factors and non-operating condition fluctuations. Based on the constructed four-dimensional operation and maintenance feature map, associate and match electrical, equipment, power grid, and spatiotemporal multi-dimensional features to complete the source tracing of abnormal fault locations. Based on the fault location, fault type, and abnormality level obtained from the source tracing, automatically generate corresponding level of equipment preventive operation and maintenance work orders and push them to terminal equipment for operation and maintenance scheduling. S6. Real-time collection of equipment energy efficiency regulation and operation data, equipment fault operation and maintenance closed-loop data, and operating condition matching deviation data of multi-dimensional coupled evaluation model to build a full-process big data closed-loop iterative mechanism; based on changes in regional charging behavior, equipment aging status updates, and changes in power grid operation rules, adaptively update the weight parameters, coupling correction parameters, and operating condition judgment rules of the multi-dimensional coupled evaluation model, and iteratively optimize the energy efficiency operation and maintenance regulation strategies corresponding to each operating condition.

2. The method for energy efficiency operation and maintenance control of smart charging piles based on big data as described in claim 1, characterized in that, In S1, the core operating parameters include charging pile charging power, bus voltage, output current, and module temperature, with a sampling period of 1 to 3 seconds; the auxiliary status parameters include ambient temperature and humidity, equipment standby status, and time period identifier, with a sampling period of 30 to 60 seconds.

3. The method for energy efficiency operation and maintenance control of smart charging piles based on big data as described in claim 1, characterized in that, In S3 , , The values ​​range from 0.2 to 0.

6. , , The values ​​range from 0.2 to 0.

6. , , The values ​​range from 0.25 to 0.

45. The value ranges from 0.1 to 0.3; based on the comprehensive evaluation value Three operating conditions are classified: 0≤ ≤0.35 indicates a high-load pressure condition; 0.35 < ≤0.75 indicates a stable and adaptable operating condition; 0.75 < ≤1 indicates a low-end or surplus operating condition.

4. The method for energy efficiency operation and maintenance control of smart charging piles based on big data according to claim 3, characterized in that, In S4, under high-load and high-pressure conditions, the charging pile cluster is subject to zoned flexible current limiting control, dynamically allocating the maximum output power of a single pile; under stable and adaptable conditions, the rated power of the charging piles is maintained in steady state; under off-peak and surplus conditions, the power limit of the charging piles is lifted; and the equipment health factor... When the value is below 0.6, the charging pile is determined to be in a sub-healthy state, and a power derating protection of 10% to 30% is implemented.

5. The method for energy efficiency operation and maintenance control of smart charging piles based on big data as described in claim 1, characterized in that, In S5, the steps for constructing the association rule system include: collecting historical energy efficiency loss data, historical fault record data, and four-dimensional operation and maintenance feature data of the charging pile cluster as sample datasets; cleaning and labeling the sample datasets; establishing the corresponding association relationship between energy efficiency anomaly features and various equipment fault types; generating multiple sets of association rules through frequent itemset mining; setting rule confidence and support thresholds to complete rule filtering; binding different energy efficiency anomaly offsets, feature combinations, corresponding fault locations, fault types, and anomaly levels; and solidifying them into an energy efficiency fault association rule system that can be matched and called in real time.

6. The method for energy efficiency operation and maintenance control of smart charging piles based on big data as described in claim 1, characterized in that, In S6, the end-to-end big data closed-loop iterative mechanism achieves adaptive parameter updates through a quantitative mathematical model, specifically including: Construct a model for calculating operating condition matching deviation: ; In the formula: This represents the operating condition matching deviation value. For multi-dimensional coupled comprehensive evaluation and prediction, This is a comprehensive evaluation value based on actual operation; Construct the iterative formula for the operating condition threshold: ; In the formula: The threshold is determined based on the current operating condition; The updated operating condition determination threshold; This is the threshold correction coefficient; The average deviation of the operating conditions within the cycle; Construct the model parameter iterative update formula: ; In the formula: This represents the set of weight parameters and coupling correction parameters for the multidimensional coupled evaluation model at the current moment. The set of parameters after iterative updates; The learning rate is used for iteration. This is the gradient vector of the model's loss function.

Images