Method and device for evaluating grid-connected performance of photovoltaic power generation system
By collecting multi-dimensional data from grid-connected points, constructing a system status identification model and dynamically allocating weights, the problem of multi-dimensional data fusion for grid-connected performance evaluation of photovoltaic power generation systems was solved, enabling dynamic adjustment and optimization decisions of evaluation results and improving operation and maintenance efficiency.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- THE SECOND SHANXI PUCHENG HUADIAN POWER GENERATION CO LTD
- Filing Date
- 2025-12-12
- Publication Date
- 2026-05-08
AI Technical Summary
Existing methods for evaluating the grid-connected performance of photovoltaic power generation systems lack integrated analysis of multi-dimensional data such as topology, equipment matching, and energy efficiency losses. Furthermore, they lack dynamic weight adjustment and optimization control based on real-time operating status, resulting in one-sided evaluation results and delayed decision-making responses, which fail to provide effective support for the intelligent operation and maintenance of photovoltaic power plants.
By collecting topology data, electrical parameters, and equipment operating status data of the photovoltaic power generation system grid connection point through monitoring devices, a grid connection performance dataset is constructed. Multi-dimensional feature extraction is performed, and combined with the system status identification model and dynamic weight allocation, the evaluation results are dynamically adjusted, and the optimization mechanism is triggered to transmit the data to the management terminal for human-computer interaction.
It achieves full automation from data perception to optimization decision-making, improves operation and maintenance response speed, and provides evaluation basis with identifiable status, interpretable weights, and traceable decisions to guide the formulation of subsequent optimization strategies.
Smart Images
Figure CN121998470A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of grid-connected performance evaluation technology for photovoltaic power generation systems, specifically to a method and apparatus for evaluating the grid-connected performance of photovoltaic power generation systems. Background Technology
[0002] Distributed photovoltaic (PV) power generation refers to PV power generation facilities built near user sites, adopting a "self-generation and self-consumption, surplus power to the grid" model, completing power generation, grid connection, and use locally. With the transformation of the global energy structure and the rapid development of renewable energy, distributed PV power generation systems are being applied on a large scale. After being connected to the grid, their grid connection performance directly affects the stability of the grid, power quality, and system operating efficiency.
[0003] However, the grid connection performance of distributed photovoltaic (PV) power generation systems is affected by various factors, including but not limited to equipment selection, electrical wiring, reactive power compensation, and grid connection conditions. Existing methods for evaluating the grid connection performance of PV power generation systems typically focus only on electrical grid connection indicators (such as power factor and harmonic content). A comprehensive evaluation of the grid connection performance of PV power generation systems can be achieved by using methods such as the analytic hierarchy process (AHP) and entropy weighting, employing a fixed weight allocation system to sum the various evaluation indicators using weighted averages.
[0004] While existing technologies can achieve grid-connected performance evaluation, they still have some limitations. For example, they tend to focus on static monitoring of electrical parameters, lacking integrated analysis of multi-dimensional data such as topology, equipment matching, and energy efficiency losses. They also lack dynamic weight adjustment and decision optimization mechanisms based on real-time operating status, resulting in one-sided evaluation results and delayed decision responses. This fails to provide effective support for the intelligent operation and maintenance of photovoltaic power plants, and the delayed evaluation results cannot provide timely and accurate guidance for the online optimization and control of the system. Therefore, a comprehensive performance evaluation method is needed that can deeply perceive the operating status, dynamically adjust weights, and link with the closed-loop optimization control. Summary of the Invention
[0005] In order to overcome the above-mentioned defects of the prior art, embodiments of the present invention provide a method and apparatus for evaluating the grid-connected performance of a photovoltaic power generation system, so as to solve the problems mentioned in the background art.
[0006] To achieve the above objectives, the present invention provides the following technical solution: a method for evaluating the grid-connected performance of a photovoltaic power generation system, comprising: S1: Collect topology data, electrical parameters, equipment operating status data, and system energy efficiency loss data of the photovoltaic power generation system grid connection point through monitoring devices to obtain grid connection performance dataset; S2: Extract features from the grid-connected performance dataset to obtain a multi-dimensional feature set of grid-connected performance. The multi-dimensional feature set of grid-connected performance includes topology dimension features, equipment matching dimension features, electrical grid-connection dimension features, and energy efficiency loss dimension features. S3: Based on the multi-dimensional feature set of grid connection performance, construct a system state identification model, combine the identified system state mode and correction factor to dynamically allocate multi-dimensional feature weights, and obtain the dynamic weight allocation result of grid connection performance evaluation. S4: Based on the multi-dimensional feature set of grid connection performance and the dynamic weight allocation result of grid connection performance evaluation, calculate the comprehensive score of each dimension evaluation, and calculate the comprehensive score of system grid connection performance through the comprehensive score of each dimension evaluation. S5: Evaluate the comprehensive score of each dimension and the comprehensive score of the system grid connection performance, trigger the optimization mechanism based on the abnormal evaluation results, and transmit the optimization mechanism information to the management terminal for human-computer interaction.
[0007] Preferably, a photovoltaic power generation system grid connection performance evaluation device includes: Grid connection performance data sensing module: Collects topology data, electrical data, equipment operating status data, and system energy efficiency loss data of the photovoltaic power generation system grid connection point through monitoring devices to obtain grid connection performance dataset; Grid connection performance multi-dimensional feature extraction module: Extracts features from the grid connection performance dataset to obtain a grid connection performance multi-dimensional feature set, which includes topology dimension features, equipment matching dimension features, electrical grid connection dimension features, and energy efficiency loss dimension features. Dynamic weight allocation module for grid connection performance evaluation: Based on the multi-dimensional feature set of grid connection performance, a system state identification model is constructed. The system state pattern and correction factor are combined to dynamically allocate the multi-dimensional feature weights and obtain the dynamic weight allocation result of grid connection performance evaluation. Grid connection performance multi-dimensional coupled evaluation module: Based on the multi-dimensional feature set of grid connection performance and the dynamic weight allocation result of grid connection performance evaluation, calculate the comprehensive score of each dimension evaluation, and calculate the comprehensive score of system grid connection performance through the comprehensive score of each dimension evaluation. Grid connection performance evaluation and optimization module: evaluates the comprehensive score of each dimension and the comprehensive score of the system grid connection performance, triggers the optimization mechanism based on abnormal evaluation results, and transmits the optimization mechanism information to the management terminal for human-computer interaction.
[0008] The technical effects and advantages of this invention are as follows: 1. This invention collects topology data, electrical parameters, equipment operating status data, and system energy efficiency loss data of the grid connection point of the photovoltaic power generation system through a monitoring device, constructs a grid connection performance dataset, and achieves full automation from data perception to optimization decision through a closed-loop process of evaluation, anomaly identification, optimization triggering, and human-computer interaction. After the optimization mechanism is triggered, it can be directly pushed to the management terminal to improve the operation and maintenance response speed. 2. This invention extracts features from four dimensions: topology, equipment matching, electrical grid connection, and energy efficiency loss, and constructs a system state identification mode and dynamic weight allocation mechanism. It can adaptively adjust the weights of each dimension according to the actual operating state of the system, avoiding the one-sidedness of traditional fixed weight evaluation methods under complex working conditions. 3. This invention determines the current operating state mode through a system state identification model and dynamically adjusts the weights by combining correction factors, making the evaluation process identifiable in terms of state, interpretable in terms of weights, and traceable in terms of decision-making. This makes it easier for operation and maintenance personnel to understand the evaluation basis and provides clear directional guidance for the formulation of subsequent optimization strategies. Attached Figure Description
[0009] Figure 1 This is a schematic diagram of the overall process of the present invention.
[0010] Figure 2 This is a schematic diagram of the method flow of the present invention.
[0011] Figure 3 This is a schematic diagram illustrating the feature extraction of the grid-connected performance dataset according to the present invention.
[0012] Figure 4 This is a schematic diagram of the dynamic weight allocation process of the present invention. Detailed Implementation
[0013] 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 some embodiments of the present invention, and not all embodiments. 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.
[0014] Please see Figure 1 As shown, the present invention provides a grid-connected performance evaluation device for a photovoltaic power generation system, including a grid-connected performance data sensing module, a grid-connected performance multi-dimensional feature extraction module, a grid-connected performance multi-dimensional coupled evaluation module, a grid-connected performance evaluation dynamic weight allocation module, and a grid-connected performance evaluation and optimization module.
[0015] The grid-connected performance data perception module is connected to the grid-connected performance multi-dimensional feature extraction module, the grid-connected performance multi-dimensional coupled evaluation module is connected to the grid-connected performance multi-dimensional feature extraction module and the grid-connected performance evaluation dynamic weight allocation module, and the grid-connected performance evaluation and optimization module is connected to the grid-connected performance evaluation dynamic weight allocation module.
[0016] Grid connection performance data sensing module: Collects topology data, electrical data, equipment operating status data, and system energy efficiency loss data of the photovoltaic power generation system grid connection point through monitoring devices to obtain grid connection performance dataset; Grid connection performance multi-dimensional feature extraction module: Extracts features from the grid connection performance dataset to obtain a grid connection performance multi-dimensional feature set, which includes topology dimension features, equipment matching dimension features, electrical grid connection dimension features, and energy efficiency loss dimension features. Dynamic weight allocation module for grid connection performance evaluation: Based on the multi-dimensional feature set of grid connection performance, a system state identification model is constructed. The system state pattern and correction factor are combined to dynamically allocate the multi-dimensional feature weights and obtain the dynamic weight allocation result of grid connection performance evaluation. Grid connection performance multi-dimensional coupled evaluation module: Based on the multi-dimensional feature set of grid connection performance and the dynamic weight allocation result of grid connection performance evaluation, calculate the comprehensive score of each dimension evaluation, and calculate the comprehensive score of system grid connection performance through the comprehensive score of each dimension evaluation. Grid connection performance evaluation and optimization module: evaluates the comprehensive score of each dimension and the comprehensive score of the system grid connection performance, triggers the optimization mechanism based on abnormal evaluation results, and transmits the optimization mechanism information to the management terminal for human-computer interaction.
[0017] Please see Figure 2 As shown, a method for evaluating the grid-connected performance of a photovoltaic power generation system includes: S1: collecting topology data, electrical parameters, equipment operating status data, and system energy efficiency loss data of the grid connection point of the photovoltaic power generation system through a monitoring device to obtain a grid-connected performance dataset; S2: extracting features from the grid-connected performance dataset to obtain a multi-dimensional feature set of grid-connected performance, which includes topology dimension features, equipment matching dimension features, electrical grid-connected dimension features, and energy efficiency loss dimension features; S3: constructing a system state identification model based on the multi-dimensional feature set of grid-connected performance, and dynamically allocating multi-dimensional feature weights by combining the identified system state patterns and correction factors to obtain a dynamic weight allocation result for grid-connected performance evaluation; S4: calculating the comprehensive score of each dimension based on the multi-dimensional feature set of grid-connected performance and the dynamic weight allocation result of grid-connected performance evaluation, and calculating the comprehensive score of the system grid-connected performance based on the comprehensive scores of each dimension; S5: evaluating the comprehensive scores of each dimension and the comprehensive score of the system grid-connected performance, triggering an optimization mechanism based on abnormal evaluation results, and transmitting the optimization mechanism information to the management terminal for human-computer interaction.
[0018] S1: Collect topology data, electrical parameters, equipment operating status data, and system energy efficiency loss data of the photovoltaic power generation system grid connection point through monitoring devices to obtain grid connection performance dataset; In this embodiment, it should be specifically noted that the system topology can be obtained from the design diagram of the distributed photovoltaic power generation system. The monitoring devices include, but are not limited to, the power / voltage sensors built into the string inverters, the monitoring modules of the smart combiner boxes, the smart energy meters deployed at the grid connection point, the grid-side dispatch communication terminal, and the bidirectional energy meters, power transmitters, power quality analyzers, photovoltaic inclined total radiation meters, current transformers, etc. installed at the point of common coupling (PCC).
[0019] In this embodiment, it is necessary to specifically explain that the topology data includes preset evaluation electrical connection points, number of power generation units, total number of circuits, number of reference nodes and reference circuits, actual maximum total active power generation, and rated total power of the system; electrical data includes the voltage of the point of common coupling, the grid rated voltage, and the effective value of the h-th harmonic current; equipment operating status data includes the nominal DC-side total power and AC-side rated power of all photovoltaic modules, and the number of inverters; system energy efficiency loss data includes AC grid-connected power, solar irradiance and total area of the photovoltaic array, preset evaluation line average loss power, and system average output power; In this embodiment, it should be specifically noted that the monitoring device includes all the aforementioned smart meters, environmental monitoring instruments, equipment communication gateways, etc.; at the same time, industrial Ethernet, fiber optic ring networks, or wireless private networks (such as 4G / 5G, LoRa) are used to ensure real-time and reliable data transmission.
[0020] Please see Figure 3 As shown, S2: Feature extraction is performed on the grid-connected performance dataset to obtain a multi-dimensional feature set of grid-connected performance. The multi-dimensional feature set of grid-connected performance includes topology dimension features, equipment matching dimension features, electrical grid connection dimension features, and energy efficiency loss dimension features, including the following steps: S2.1: Topology Dimension Features: Based on the topology data in the grid connection performance dataset, topology dimension features are extracted, and the electrical wiring rationality index I is calculated respectively. to and N-1 redundancy index I re ; S2.1.1: The electrical wiring rationality index I to The redundancy score f is obtained by evaluating the backup switch configuration (e.g., backup circuit breaker, tie switch, etc.) of the electrical connection points through a preset assessment. sw Power outage maintenance isolation score f is and wiring complexity penalty factor f co We obtain I by weighting. to =a1×f sw +a2×fis -a3×f co a1, a2, and a3 are the corresponding weight coefficients, satisfying a1 + a2 = 1. a3 is an independent penalty weight, obtained based on historical data regression analysis, for example, a1 = 0.4, a2 = 0.6, and a3 = 0.5. , where n is the preset total number of electrical connection points for evaluation, and s i Configure an identifier for the redundant switch of the i-th critical node. If the node is configured with a spare switch, then s i If 1 is configured, then s is not configured. i w is 0 i w is the weight of the i-th electrical connection point. i =P af / P to P af The loss of power to node i will cause the total rated power of node i and all its downstream load nodes to decrease, P. to The total rated power of the system; f is =1-(dy y / dy t ), dy y and dy t These represent the number of units that were forced to stop generating power during power outage maintenance and isolation, and the total number of generating units, respectively. N no and N lo N represents the total number of nodes and the total number of loops in the system wiring diagram. no,0 and N lo,0 To preset the baseline number of nodes and the baseline number of loops for a photovoltaic power generation system of the same capacity, α and β are the corresponding weights used to adjust the contribution ratio of the number of nodes and the number of loops to the complexity, satisfying α+β=1, for example α=0.6 and β=0.4; This embodiment requires specific explanation of the preset evaluation electrical connection points: these refer to the electrical connection points in the photovoltaic power generation system electrical wiring diagram used to achieve power collection, voltage transformation, and grid connection, including but not limited to photovoltaic string combiner points, inverter DC / AC side connection points, grid connection point switch interfaces, etc.; power generation unit refers to the smallest functional module that constitutes the photovoltaic power generation system and can independently generate and control power. A power generation unit typically includes: DC side, AC side, and supporting equipment, etc.; the total number of nodes in the system wiring diagram is the number of all electrical connection points in the wiring diagram (including critical nodes and non-critical branch nodes); the total number of circuits in the system wiring diagram is the number of paths in the wiring diagram that form closed electrical paths (including main circuits, branch circuits, and backup circuits).
[0021] S2.1.2: The N-1 redundancy index I re The pass rate R of N-1 verification of n1 preset evaluation devices in a photovoltaic power generation system was statistically analyzed. jThe average value of R is calculated; j Given a piecewise function, if the evaluated device j fails and exits operation, the actual maximum total active power P of all remaining devices in the system is... af j ≥System rated total power P ra The threshold η is passed by the N-1 check. th The product of R is then j =1, otherwise R j =P af j / (P ra ×η th ),Right now The N-1 check passes the threshold η th It is usually set to 0.7 or 0.8; This embodiment requires specific explanation of the preset evaluation equipment selection rules in the photovoltaic power generation system: First, the functional principle: equipment whose failure would cause a complete interruption of the power transmission path or failure of the system voltage level transformation, specifically including: power collection and conversion equipment: step-up transformers, centralized inverters; critical power transmission equipment: DC combiner boxes and AC combiner cabinets used to collect the power of multiple power generation units; backbone connection equipment: backbone DC cables or AC collector lines connecting the photovoltaic array area and the grid connection point (which can be equivalently regarded as a "device" for verification); Second, the no-redundancy principle: equipment that has been clearly configured as dual-circuit power supply and hot standby (such as switch cabinets with dual-circuit power supply) is regarded as a logical unit and counted only once.
[0022] In this embodiment, it is necessary to specifically explain the total active power P. af j It can be obtained through power flow calculation. First, define the decision variable vector p, p=[P1,P2,...,P n1-1 ] T P n1-1 Let P be the active power output value of the (n1-1)th device; then, under the premise of satisfying all constraints, let P be the actual maximum total active power generated by all remaining devices in the system. af j The constraints include topological connectivity constraints: power is transmitted only through networks that remain connected after a fault; equipment capacity constraints: the output power of all remaining inverters must not exceed their current maximum allowable power; line current constraints: the effective value of the current in each branch must not exceed its maximum current limit; and voltage safety constraints: the voltage of each node must be within the allowable range. Finally, the problem is solved using linear programming. Let I be a row vector containing all 1s, and p be a vector of decision variables. The optimal objective function value Z is the total active power P. af j .
[0023] S2.2: Equipment Matching Dimension Features: Based on the centralized equipment operating status data in the grid-connected performance dataset, equipment matching dimension features are extracted, and the inverter-module capacity matching index I is calculated respectively. mis And reactive power compensation equipment response index I q The I mis The mismatch deviation is obtained by subtracting the average mismatch deviation of all N inverters from 1. The mismatch deviation is calculated from the nominal DC-side total power P of all photovoltaic modules connected to the J-th inverter. dc J The ratio P to the rated power on the AC side ac J The absolute difference obtained by subtracting 1 is... ; T is the evaluation time window, Q ref (t) and Q act (t) represents the reference value of reactive power required by the system at time t and the actual reactive power output of reactive power compensation, respectively. λ is the attenuation constant (e.g., 0.3), with the unit being one-tenth of the reactive power unit, used to adjust the sensitivity of the error. S2.3: Electrical Grid Connection Dimension Features: Based on the centralized electrical data in the grid connection performance dataset, electrical grid connection dimension features are extracted, and the voltage safety operation index I is calculated respectively. vs Harmonic-Grid Adaptability Index I har The I vs As a piecewise function, if the voltage V at the point of common coupling between the photovoltaic power generation system and the upstream public power grid is within the evaluation time window T... pcc ≥ Rated voltage of the power grid V nom I vs If the voltage safety lower limit threshold V is 1, low <V pcc <V nom I vs =(V pcc -V low ) / (V nom -V low If V pcc ≤V low I vs =0; the I har By using a preset maximum harmonic order H (e.g., 40, 50, etc.) combined with the effective value I of the h-th harmonic current at the point of common coupling... h and the corresponding upper limit value I of the h-th harmonic current h,max Calculated, i.e. b hThe preset h-th harmonic weighting coefficient can be set according to the degree of harmonic hazard. For example, odd harmonics have a greater weight than even harmonics, and lower harmonics have a greater weight than higher harmonics. b5=0.3, b7=0.25, etc. β1 is a scaling factor used to adjust the result to the range of 0 to 1. For example, β1=0.5. In this embodiment, it should be specifically noted that the common connection point refers to the unique and legally defined electrical connection point between the distributed photovoltaic power generation system and the upper-level public power grid for power exchange, power metering, and grid connection agreement. Its location is clearly marked in the grid connection agreement and electrical design drawings.
[0024] S2.4: Energy Efficiency Loss Dimension Features: Based on the centralized system energy efficiency loss data in the grid-connected performance dataset, energy efficiency loss dimension features are extracted, and the system comprehensive efficiency index I is calculated respectively. eff And the preset evaluation line transmission efficiency index I loss The I eff By evaluating the total AC power P of the system output at time t within the evaluation time window T. aac The integral of (t), the integral of solar irradiance G(t) at time t, and the total area A of the photovoltaic array are used to calculate the result. η stc The conversion efficiency of the photovoltaic module under standard test conditions (e.g., 0.24); the I loss The average power loss P of the N2 preset evaluation lines within the evaluation time window T is subtracted from 1. lo-ke With the system's average output power P tot The ratio is obtained, that is, I loss =1-P lo-ke / P tot , (Exchange route) (DC line), N1 = T / Δt, where Δt is a fixed sampling interval. p tot The total active power output of the system is measured by the smart meter at the grid connection point, k is the kth preset evaluation line, and K is the Kth sampling. This embodiment requires specific explanation of the preset evaluation line selection rules: Capacity ratio rule: Any line whose designed current carrying capacity or rated transmission capacity accounts for more than a preset threshold of the total rated capacity of the system is included in the set; Topology uniqueness rule: Line segments located between the power generation unit and the grid connection point that are unique and have no backup parallel paths; Length and resistance rule: Lines with a length exceeding a certain value (such as 100 meters) and a relatively small conductor cross-sectional area (large resistance) should be considered for inclusion even if the capacity ratio is not large, because the absolute value of loss may be significant; The total area A of the photovoltaic array is obtained by multiplying the total number of photovoltaic array blocks, the length and width of a single photovoltaic module.
[0025] Please see Figure 4 As shown, S3: Based on the multi-dimensional feature set of grid-connected performance, a system state identification model is constructed. The identified system state patterns and correction factors are combined to dynamically allocate multi-dimensional feature weights, resulting in a dynamic weight allocation result for grid-connected performance evaluation. This includes the following steps: S3.1: N-1 Redundancy Index I based on multi-dimensional feature set of grid connection performance re Voltage safety operation index I vs System overall efficiency index I eff and inverter health rate I nb A system steady-state model M1, a system fluctuation state model M2, a system fault state model M3, and a system grid weakness model M4 are constructed respectively; the inverter health rate I... nb The value is obtained by the ratio of the number of inverters generating electricity normally in the photovoltaic power generation system to the total number of inverters. S3.2: If I re ≥ the corresponding threshold η re (e.g., 0.9) and I nb ≥ the corresponding threshold η nb (e.g., 0.95) and I vs ≥ the corresponding threshold η vs (e.g., 0.97) and I eff ≥ the corresponding threshold η eff (For example, 0.85), then the system is in steady-state mode M1 (indicating that the system topology is good, the equipment is good, the voltage is operating well, and the overall efficiency has no significant loss); if I nb ≥η nb And I vs ≥η vs And I eff <η eff If the system is in fluctuating state mode M2 (indicating that the system equipment is in good condition and the voltage is operating well, but there is a loss in overall efficiency, characterized by efficiency loss caused by power fluctuations); if I nb <η nb Then the system is in fault state mode M3 (the only key judgment in this mode is the decline in equipment health rate, which has the highest priority); if I vs <η vs If so, the system is in the weak grid mode M4 (a decrease in the electrical grid connection voltage safety operation index is the core indicator of a weak grid). S3.3: N-1 Redundancy Index I Based on Multi-Dimensional Feature Set of Grid-Connected Performance re Inverter-Module Capacity Matching Index I mis Voltage safety operation index I vs And the preset evaluation line transmission efficiency index I loss Calculate the correction factor C for the topology dimension respectively.t Correction factor C for equipment matching dimension e Correction factor C for electrical grid connection dimension g And the correction factor C for energy efficiency loss dimension L C t =2-I re C e =2-I mis C g =1+max(0,1-I vs ), C L =2-I loss ; In this embodiment, it should be specifically noted that the correction factor is designed to be negatively correlated with the corresponding index. When the performance index of a certain dimension decreases, its correction factor increases, thereby increasing the weight ratio of that dimension in the dynamic weight allocation, so as to focus on the weak links of the current system.
[0026] S3.4: Combine the identified system state patterns and correction factors to dynamically allocate multi-dimensional feature weights, obtaining the weights ω for each dimension. m This yields the dynamic weight allocation result for grid connection performance evaluation, where m∈{t,e,g,L}, representing the topology dimension, equipment matching dimension, electrical grid connection dimension, and energy efficiency loss dimension, respectively. ω m M The preset basic weights for the m-th dimension under the current system state mode M (can be obtained through historical data using the analytic hierarchy process, for example, the basic weights for each dimension under the stable state mode M1 are 0.15, 0.15, 0.30, 0.40). S4: Based on the multi-dimensional feature set of grid connection performance and the dynamic weight allocation results of grid connection performance evaluation, calculate the comprehensive score of each dimension evaluation, and then calculate the comprehensive score of system grid connection performance through the comprehensive scores of each dimension evaluation; firstly, the comprehensive score SC of each dimension evaluation is obtained by weighted summing of each index calculated from each dimension feature of grid connection performance and its corresponding weight. m Then, the overall score S for each dimension is evaluated. m and the weights ω of each dimension m The weighted summation yields the overall system grid connection performance score SC. n² is the number of exponents calculated in the m-th dimension, and γ i The weights of the i-th index satisfy Σγ i =1, I i Let i be the i-th index; ω m Let m be the weight of the m-th dimension; S5: Evaluate the comprehensive score of each dimension and the comprehensive score of system grid connection performance. Trigger the optimization mechanism based on abnormal evaluation results and transmit the optimization mechanism information to the management terminal for human-computer interaction; if the comprehensive score of system grid connection performance SC ≥ the corresponding threshold SC th And the overall score of each dimension is SC m ≥ Corresponding exclusive threshold SC m,th This indicates that the photovoltaic power generation system has good grid connection performance; if SC≥SC th SC exists m <SC m,th If the number of abnormal dimensions is ≤ N3 (e.g., N3 = 1), it is determined to be a local anomaly, triggering the corresponding dimension optimization mechanism. Optimization mechanism information is generated and transmitted to the management terminal for human-computer interaction. This optimization mechanism information includes the system's overall grid performance score, the overall score for anomaly dimension evaluation, the current system status mode, and optimization suggestions for the corresponding dimension anomaly. Otherwise, it proceeds to SC < SC. th The process; if SC < SC th Then calculate SC m,th Subtract SC m The difference ΔSC m Filter out the differences ΔSC that are greater than or equal to 0. m The optimization mechanism information is generated and transmitted to the management terminal for human-computer interaction. The optimization mechanism information includes the comprehensive score of the system's grid connection performance, the priority ranking results of the comprehensive score of the anomaly dimension evaluation, the current system status mode, and anomaly optimization suggestions for each dimension. This embodiment requires specific explanation, for example, SC≥SC. th Furthermore, there are anomalies in the comprehensive score of the electrical grid connection dimension assessment, generating optimization suggestions for the electrical grid connection dimension, including optimization suggestions for the voltage safety operation index and harmonic-grid adaptability index. For example, if the voltage support is insufficient, it is recommended to report to the grid dispatching department and request an inspection of the operation mode of the upstream substation or adjustment of the transformer taps.
[0027] Secondly: The accompanying drawings of the embodiments disclosed in this invention only involve the structures involved in the embodiments disclosed in this invention. Other structures can refer to the general design. In the absence of conflict, the same embodiment and different embodiments of this invention can be combined with each other. In conclusion, the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.
Claims
1. A method for evaluating the grid-connected performance of a photovoltaic power generation system, characterized in that: include: S1: Collect topology data, electrical parameters, equipment operating status data, and system energy efficiency loss data of the photovoltaic power generation system grid connection point through monitoring devices to obtain grid connection performance dataset; S2: Extract features from the grid-connected performance dataset to obtain a multi-dimensional feature set of grid-connected performance. The multi-dimensional feature set of grid-connected performance includes topology dimension features, equipment matching dimension features, electrical grid-connection dimension features, and energy efficiency loss dimension features. S3: Based on the multi-dimensional feature set of grid connection performance, construct a system state identification model, combine the identified system state mode and correction factor to dynamically allocate multi-dimensional feature weights, and obtain the dynamic weight allocation result of grid connection performance evaluation. S4: Based on the multi-dimensional feature set of grid connection performance and the dynamic weight allocation result of grid connection performance evaluation, calculate the comprehensive score of each dimension evaluation, and calculate the comprehensive score of system grid connection performance through the comprehensive score of each dimension evaluation. S5: Evaluate the comprehensive score of each dimension and the comprehensive score of the system grid connection performance, trigger the optimization mechanism based on the abnormal evaluation results, and transmit the optimization mechanism information to the management terminal for human-computer interaction.
2. The method for evaluating the grid-connected performance of a photovoltaic power generation system according to claim 1, characterized in that: The topology dimension features in S2 are extracted based on the topology data in the grid connection performance dataset, and the electrical wiring rationality index I is calculated accordingly. to and N-1 redundancy index I re ; Electrical wiring rationality index I to The redundancy score f of the backup switch configuration at the preset evaluation electrical wiring points is used to assess the redundancy. sw Power outage maintenance isolation score f is and wiring complexity penalty factor f co We obtain I by weighting. to =a1×f sw +a2×f is -a3×f co a1, a2 and a3 are the corresponding weight coefficients, satisfying a1+a2=1, and a3 is an independent penalty weight; N-1 Redundancy Index I re The pass rate R of N-1 verification of n1 preset evaluation devices in a photovoltaic power generation system was statistically analyzed. j The average value of R is calculated; j Given a piecewise function, if critical equipment j fails and goes out of service, the actual maximum total active power P of all remaining equipment in the system is... af j ≥System rated total power P ra The threshold η is passed by the N-1 check. th The product of R is then j =1, otherwise R j =P af j / (P ra ×η th ).
3. The method for evaluating the grid-connected performance of a photovoltaic power generation system according to claim 1, characterized in that: The equipment matching dimension features in S2 are extracted based on the centralized equipment operating status data in the grid-connected performance dataset. The inverter-module capacity matching index I is calculated for each feature. mis And reactive power compensation equipment response index I q The I mis The mismatch deviation is obtained by subtracting the average mismatch deviation of all N inverters from 1. The mismatch deviation is obtained by taking the nominal DC-side total power P of all photovoltaic modules connected to the J-th inverter. dc J The ratio P to the rated power on the AC side ac J The absolute difference obtained by subtracting 1; T is the evaluation time window, Q ref (t) and Q act (t) represents the reference value of reactive power required by the system at time t and the actual reactive power output of reactive power compensation, respectively, and λ is the attenuation constant, with the unit being one-unit of reactive power.
4. The method for evaluating the grid-connected performance of a photovoltaic power generation system according to claim 1, characterized in that: The electrical grid connection dimension characteristics in S2 are as follows: Based on the centralized electrical data of the grid connection performance dataset, electrical grid connection dimension features are extracted, and the voltage safety operation index I is calculated respectively. vs Harmonic-Grid Adaptability Index I har The I vs As a piecewise function, if the voltage V at the point of common coupling between the photovoltaic power generation system and the upstream public power grid is within the evaluation time window T... pcc ≥ Rated voltage of the power grid V nom I vs If the voltage safety lower limit threshold V is 1, low <V pcc <V nom I vs =(V pcc -V low ) / (V nom -V low If V pcc ≤V low I vs =0; the I har By using the preset highest harmonic order H, combined with the effective value I of the h-th harmonic current at the point of common coupling... h and the corresponding upper limit value I of the h-th harmonic current h,max Calculated.
5. The method for evaluating the grid-connected performance of a photovoltaic power generation system according to claim 1, characterized in that: Energy efficiency loss dimension characteristics in S2: Based on the grid-connected performance dataset, energy efficiency loss dimension features are extracted from the centralized system energy efficiency loss data, and the system comprehensive efficiency index I is calculated accordingly. eff And the preset evaluation line transmission efficiency index I loss The I eff By evaluating the total AC power P of the system output at time t within the time window T. aac The integral of (t), the integral of solar irradiance G(t) at time t, the total area A of the photovoltaic array, and the conversion efficiency η of the photovoltaic module under standard test conditions. stc Calculated; the I loss The average power loss P of the N2 preset evaluation lines within the evaluation time window T is subtracted from 1. lo-ke With the system's average output power P tot The ratio is obtained, that is, I loss =1-P lo-ke / P tot .
6. The method for evaluating the grid-connected performance of a photovoltaic power generation system according to claim 1, characterized in that: The dynamic weight allocation result for grid connection performance evaluation in S3 is based on the N-1 redundancy index I in the multi-dimensional feature set of grid connection performance. re Voltage safety operation index I vs System overall efficiency index I eff and inverter health rate I nb A system steady-state model M1, a system fluctuation state model M2, a system fault state model M3, and a system grid weakness model M4 are constructed respectively; the inverter health rate I... nb The value is obtained by the ratio of the number of inverters generating electricity normally in the photovoltaic power generation system to the total number of inverters. The N-1 redundancy index I based on the multi-dimensional feature set of grid connection performance. re Inverter-Module Capacity Matching Index I mis Voltage safety operation index I vs And the preset evaluation line transmission efficiency index I loss Calculate the correction factor C for the topology dimension respectively. t Correction factor C for equipment matching dimension e Correction factor C for electrical grid connection dimension g And the correction factor C for energy efficiency loss dimension L C t =2-I re C e =2-I mis C g =1+max(0,1-I vs ), C L =2-I loss ; By combining the identified system state patterns and correction factors, multi-dimensional feature weights are dynamically assigned to obtain the weights ω for each dimension. m This yields the dynamic weight allocation results for grid connection performance evaluation.
7. The method for evaluating the grid-connected performance of a photovoltaic power generation system according to claim 1, characterized in that: The comprehensive score for system grid connection performance in S4 is obtained by first weighting and summing the calculated indices and their corresponding weights from each dimension of grid connection performance to obtain the comprehensive score SC for each dimension. m Then, the overall score S for each dimension is evaluated. m and the weights ω of each dimension m The weighted summation yields the overall grid-connected performance score SC of the system.
8. The method for evaluating the grid-connected performance of a photovoltaic power generation system according to claim 1, characterized in that: The optimization mechanism information in S5: If the overall grid-connected performance score SC of the system is greater than or equal to the corresponding threshold SC th And the overall score of each dimension is SC m ≥ Corresponding exclusive threshold SC m,th This indicates that the photovoltaic power generation system has good grid connection performance; if SC≥SC th SC exists m <SC m,th If the number of abnormal dimensions is ≤ N3, it is determined to be a local anomaly, triggering the corresponding dimension optimization mechanism. Optimization mechanism information is generated and transmitted to the management terminal for human-computer interaction. This optimization mechanism information includes the system's overall grid performance score, the overall score for anomaly dimension evaluation, the current system status mode, and optimization suggestions for the corresponding dimension anomaly. Otherwise, it proceeds to SC < SC. th The process; if SC < SC th Then calculate SC m,th Subtract SC m The difference ΔSC m Filter out differences ΔSC that are greater than or equal to 0. m The optimization mechanism information is generated and transmitted to the management terminal for human-computer interaction. The optimization mechanism information includes the comprehensive score of the system's grid connection performance, the priority ranking results of the comprehensive score of the anomaly dimension evaluation, the current system status mode, and the corresponding anomaly optimization suggestions.
9. A photovoltaic power generation system grid connection performance evaluation device, used for applying the photovoltaic power generation system grid connection performance evaluation method according to any one of claims 1-8, comprising: Grid connection performance data sensing module: Collects topology data, electrical data, equipment operating status data, and system energy efficiency loss data of the photovoltaic power generation system grid connection point through monitoring devices to obtain grid connection performance dataset; Grid connection performance multi-dimensional feature extraction module: Extracts features from the grid connection performance dataset to obtain a grid connection performance multi-dimensional feature set, which includes topology dimension features, equipment matching dimension features, electrical grid connection dimension features, and energy efficiency loss dimension features. Dynamic weight allocation module for grid connection performance evaluation: Based on the multi-dimensional feature set of grid connection performance, a system state identification model is constructed. The system state pattern and correction factor are combined to dynamically allocate the multi-dimensional feature weights and obtain the dynamic weight allocation result of grid connection performance evaluation. Grid connection performance multi-dimensional coupled evaluation module: Based on the multi-dimensional feature set of grid connection performance and the dynamic weight allocation result of grid connection performance evaluation, calculate the comprehensive score of each dimension evaluation, and calculate the comprehensive score of system grid connection performance through the comprehensive score of each dimension evaluation. Grid connection performance evaluation and optimization module: evaluates the comprehensive score of each dimension and the comprehensive score of the system grid connection performance, triggers the optimization mechanism based on abnormal evaluation results, and transmits the optimization mechanism information to the management terminal for human-computer interaction.