A distributed photovoltaic power station intelligent operation and maintenance method based on fault tree analysis

Abstract

The application discloses a kind of distributed photovoltaic power station intelligent operation and maintenance methods based on fault tree analysis, comprising the following steps: S1, constructs distributed photovoltaic power station fault tree model;S2, real-time acquisition distributed photovoltaic power station's operating data;S3, using fault tree analysis method, according to the operating data of acquisition calculates the occurrence probability of each bottom event of fault tree;S4, according to bottom event occurrence probability identifies key failure factor;S5, according to key failure factor formulates intelligent operation and maintenance strategy.The application uses the above-mentioned one kind of distributed photovoltaic power station intelligent operation and maintenance methods based on fault tree analysis, accurately identifies key failure factor, dynamically adjusts operation and maintenance strategy, improves fault positioning and repair efficiency, reduces operation and maintenance cost, promotes distributed photovoltaic power station to active intelligent operation and maintenance upgrade.

Description

Technical Field

[0001] This invention relates to the field of operation and maintenance technology for distributed photovoltaic power plants, and in particular to an intelligent operation and maintenance method for distributed photovoltaic power plants based on fault tree analysis. Background Technology

[0002] Distributed photovoltaic (PV) power stations consist of multiple components, including PV modules, inverters, combiner boxes, monitoring systems, and transmission lines. Various malfunctions can occur in these components during operation, affecting the power station's normal power generation. Traditional operation and maintenance of distributed PV power stations mainly rely on manual inspections and simple monitoring systems, a method with numerous shortcomings.

[0003] First, manual inspections are inefficient and costly. Because distributed photovoltaic (PV) power stations are widely distributed, manual inspections require significant manpower, resources, and time, and the workload for inspectors is heavy. Furthermore, manual inspections are greatly affected by the subjective factors of the inspectors, leading to missed or incorrect inspections, making it difficult to guarantee the accuracy and comprehensiveness of the inspections. For example, in areas with complex terrain, inspectors may be unable to reach certain PV power stations, resulting in the failure to detect faults in these stations in a timely manner.

[0004] Secondly, most existing monitoring systems can only perform simple monitoring and alarm functions on the operating parameters of power plants, lacking the ability to deeply analyze and diagnose faults. When the monitoring system issues an alarm, maintenance personnel often struggle to quickly and accurately pinpoint the root cause of the fault, requiring significant time for troubleshooting, resulting in prolonged fault repair times and impacting the power plant's power generation efficiency. For example, when the power generation of a photovoltaic power plant decreases, the monitoring system may only issue an alarm indicating abnormal power, but it cannot determine whether the cause is a faulty photovoltaic module, inverter, or other equipment.

[0005] Furthermore, the faults in distributed photovoltaic (PV) power plants are diverse and complex. Different causes of faults can lead to the same fault phenomena, while the same cause can manifest different fault phenomena under different environmental conditions. In addition, as the scale of distributed PV power plants continues to expand, the amount of operational data they generate is also increasing. Traditional operation and maintenance methods cannot fully utilize this data, cannot extract valuable information from the massive amounts of data, and cannot achieve intelligent operation and maintenance of the power plants.

[0006] Therefore, there is an urgent need for a new operation and maintenance method that can effectively solve the above problems and improve the operation and maintenance efficiency and management level of distributed photovoltaic power stations. Summary of the Invention

[0007] The purpose of this invention is to provide an intelligent operation and maintenance method for distributed photovoltaic power plants based on fault tree analysis. By constructing a fault tree model and combining real-time and historical data, the method uses analysis algorithms to diagnose faults and identify key factors, thereby formulating intelligent operation and maintenance strategies to achieve efficient and intelligent operation and maintenance, improve power generation efficiency and reliability, and reduce operation and maintenance costs.

[0008] To achieve the above objectives, this invention provides an intelligent operation and maintenance method for distributed photovoltaic power plants based on fault tree analysis, comprising the following steps:

[0009] S1. Construct a fault tree model for a distributed photovoltaic power station;

[0010] S2. Real-time collection of operational data from distributed photovoltaic power stations;

[0011] S3. Using fault tree analysis, calculate the probability of occurrence of each bottom event in the fault tree based on the collected operational data;

[0012] S4. Identify key failure factors based on the probability of occurrence of basic events;

[0013] S5. Develop intelligent operation and maintenance strategies based on key failure factors.

[0014] 2. The intelligent operation and maintenance method for distributed photovoltaic power plants based on fault tree analysis according to claim 1, characterized in that S1 includes the following steps:

[0015] S11. Collect historical fault data from multiple distributed photovoltaic power stations within the target area, including basic fault data, operating parameter data, and environmental parameter data; preprocess the collected historical fault data, including data cleaning, removal of missing and outlier values, and data standardization, converting data of different dimensions to a specific range.

[0016] S12. Calculate the occurrence frequency of various faults within a certain time period, and select high-frequency fault types as important candidate events in the fault tree model:

[0017]

[0018] Among them, f F n represents the frequency of occurrence of fault type F. F This represents the number of times fault type F occurs within the statistical period, and N is the total number of times all faults occur within the statistical period.

[0019] S13. Use the decision tree algorithm to analyze the correlation between fault types and operating parameters and environmental parameters;

[0020] S14. Identify fault tree events, including top events, intermediate events, and bottom events.

[0021] Preferably, in S11, the basic fault data includes the fault occurrence timestamp t, the fault type F (including photovoltaic module fault, inverter fault, combiner box fault, and transmission line fault), and the power station number Sid where the fault occurred; the operating parameter data includes the output voltage V of the photovoltaic module. module Output current I module The input voltage V of the inverter inverter-in Input current I inverter-in Output voltage V inverter-out Output current I inverter-out The input current I of the combiner box combiner-in Output current; environmental parameter data include ambient temperature T, light intensity S, and humidity H.

[0022] Preferably, the specific steps in S13 are as follows:

[0023] S131. First, divide the historical fault data after S11 preprocessing into feature set A and label set Y; feature set A includes operating parameters and environmental parameters, each row represents a data record, and each column corresponds to a parameter; label set Y is the fault type, and each element corresponds to the fault type F of a data record.

[0024] S132. Using information gain ratio as the metric for selecting the splitting attribute, first calculate the information entropy Ent(D) of dataset D, as follows:

[0025]

[0026] Where |C| represents the number of fault types in the dataset, p i It represents the proportion of samples in the dataset that belong to the i-th type of fault.

[0027] For each attribute 'a', calculate its information gain Gain(D,a) with respect to dataset D, using the following formula:

[0028]

[0029] Where V is the number of possible values ​​for attribute a, and D n It is a subset of samples in dataset D where attribute a takes the value v, |D v | and |D| represent subsets D, respectively. v And the number of samples in dataset D;

[0030] Next, calculate the intrinsic value of attribute a, using the formula IV(a), which is:

[0031]

[0032] Finally, the information gain ratio GainRatio(D,a) of attribute a is obtained, and the formula is:

[0033]

[0034] Select the attribute with the largest information gain ratio as the partitioning attribute of the current node;

[0035] Using dataset D as the root node, the dataset is divided into multiple subsets according to the selected splitting attribute. Each subset corresponds to a child node. For each child node, the process of selecting splitting attributes and splitting the dataset is repeated to recursively construct a decision tree.

[0036] Preferably, in S14, the power generation efficiency E of the distributed photovoltaic power station is lower than the normal threshold E. th Set as the top event; the formula for calculating power generation efficiency E is as follows:

[0037]

[0038] Among them, P out P represents the actual output power of the power plant. in This represents the theoretical maximum input power; when E < E th When this event occurs, a top-level failure is determined to have occurred in the distributed photovoltaic power station, indicating a decrease in power generation efficiency.

[0039] Intermediate events are determined by combining the high-frequency fault types screened in S12 with the correlation between fault types and operating and environmental parameters obtained in S13. The specific process is as follows:

[0040] First, high-frequency fault types among photovoltaic module faults, inverter faults, combiner box faults, and transmission line faults are classified as first-level intermediate events.

[0041] For photovoltaic module failures, abnormal photovoltaic module output voltage and abnormal photovoltaic module output current are considered as secondary intermediate events.

[0042] For inverter faults, abnormal inverter input voltage, abnormal inverter input current, abnormal inverter output voltage, and abnormal inverter output current are treated as secondary intermediate events.

[0043] The bottom events are determined based on equipment defects, abnormal operating parameters, and environmental factors. Furthermore, the correlation between fault types and parameters obtained from the decision tree analysis in S13 is used to further clarify the bottom events. The specific process is as follows:

[0044] Regarding equipment defects, photovoltaic module aging, photovoltaic module surface damage, inverter cooling fan failure, and inverter circuit board failure are considered as bottom events for photovoltaic modules.

[0045] Regarding abnormal operating parameters, the photovoltaic module output voltage exceeding the normal range or the inverter input current fluctuating excessively is considered a bottom-line event.

[0046] Regarding environmental factors, the bottom events are: excessively high ambient temperature causing inverter overheating, a sudden drop in light intensity affecting photovoltaic module power generation, and high humidity causing a decline in the insulation performance of transmission lines.

[0047] Preferably, S2 includes the following steps:

[0048] S21. Install voltage and current sensors at the photovoltaic module array to collect the output voltage V of the photovoltaic modules. module and output current I module ;

[0049] Voltage and current sensors are installed on the input and output sides of the inverter to collect the inverter's input voltage V. inverter-in Input current I inverter-in Output voltage V inverter-out and output current I inverter-out ;

[0050] Install current sensors at the input and output terminals of the combiner box to collect the input current I of the combiner box. combiner-in and output current I combiner-out Real-time monitoring of current transmission in the combiner box;

[0051] Environmental monitoring equipment, including temperature sensors, light intensity sensors, and humidity sensors, is installed in the power plant area to measure ambient temperature T, acquire light intensity in real time, and monitor ambient humidity H.

[0052] S22. A data transmission network is built using LoRa technology to transmit the collected data to the data processing center in real time. The data transmission network includes LoRa device deployment, channel and frequency band settings, adjustment of LoRa device spreading factor SF and transmit power, LoRa network ID and key configuration, data transmission and verification, and network monitoring and optimization.

[0053] S23. The data acquisition device collects each parameter in real time according to the set sampling frequency, and performs preliminary processing on the collected data, including analog-to-digital conversion of the collected analog signal, averaging of n data points collected continuously for each acquisition parameter to obtain filtered data, and timestamping the pre-processed data to record the time of data acquisition.

[0054] S24. Store the pre-processed and timestamped data in a local cache, and upload the data to the database of the data processing center at certain time intervals.

[0055] Preferably, S3 includes the following steps:

[0056] S31. Based on the historical fault data collected in S11, statistically analyze each bottom event B. i Combination of S under different operating states j Based on the frequency of occurrence of event B, construct a probability table for the base event and record the base event B. i Combined with operating state S j Number of co-occurring samples Total number of samples that occur in combination with motion state

[0057] S32, Transfer the real-time operating parameter vector X collected in S2 real This includes parameters such as photovoltaic module output voltage and current, inverter input and output voltage and current, ambient temperature, and illuminance, which are mapped to discretized operating states S through equal-frequency binning or decision tree partitioning rules. current ;

[0058] S33. Based on historical data, calculate the prior probability of the base event using the conditional probability formula:

[0059]

[0060] like Then, the Laplace smoothing method is used for correction:

[0061]

[0062] Where M is the total number of events at the bottom of the fault tree;

[0063] Combining the decision tree model of S13, the real-time running parameter vector X real Input this model, and the model will output the i-th bottom event B under the current parameters. i Let P be the probability of occurrence. dt (B i |X real This probability value is used to weight and correct the prior probability:

[0064] P'(B i )=α·P(B i |S current )+(1-α)P dt (B i |X real );

[0065] Where α is the weighting coefficient for historical data;

[0066] S34. Calculate the probability of intermediate events based on the logical relationships of AND and OR gates. The calculation formula is as follows:

[0067] When the intermediate event is composed of k base events B1, B2, ..., B k When connected to a logic gate, the intermediate event will only occur if the k base events occur simultaneously. The probability P(AND) of the intermediate event occurring when connected to a gate is calculated as follows:

[0068]

[0069] Among them, P(B) i ) represents the probability of the i-th base event occurring, and ∏ is the multiplication symbol;

[0070] If the intermediate events are composed of k base events B1, B2, ..., B... k By using an OR logic connection, the intermediate event will occur as long as any one of the k basic events occurs. The probability P(OR) of the intermediate event occurring through the OR gate is calculated as follows:

[0071]

[0072] Among them, 1-P(B i () represents the probability that the i-th bottom event does not occur;

[0073] S35. Calculate the probability P(E) of the top event using minimal cut sets. Suppose the fault tree has m minimal cut sets C1, C2, ..., Cn. m The probability of the occurrence of the l-th minimal cut set is P(C l If the probability of the top event P(E) is calculated using the law of total probability:

[0074]

[0075] Among them, C o Let C represent the Oth minimal cut set. l Let ∑ represent the l-th minimal cut set. 1≤l＜O≤m P(C l ∩C o () represents the summation of the probabilities of two distinct minimal cut sets occurring simultaneously;

[0076] S36. Employ a sliding time window method to continuously update historical statistical samples, optimize and adjust the probability of the bottom event, B. i With running state S j The formula for dynamically updating the number of co-occurring samples is:

[0077]

[0078] in, This represents the original number of co-occurring samples. These are expired samples. For new samples, This is the updated sample.

[0079] Preferably, S4 includes the following steps:

[0080] S41. Based on the actual operation and maintenance experience of the power plant and historical data, set a probability threshold β for the criticality of the basic events, which is used to initially screen out basic events with a high probability of occurrence.

[0081] Assign weight coefficients ω to different types of bottom events i This reflects the varying degrees of impact of each bottom event on the operation of the power plant;

[0082] S42. For each bottom event B i Combined with its occurrence probability P'(B) i ) and weighting coefficient ω″ i Calculate its comprehensive impact value I i The calculation formula is as follows:

[0083] I i =ω″ i ×P'(B i );

[0084] Wherein, P'(B i ) represents the probability of occurrence of the bottom event after correction by S34. This formula is used to quantify the overall impact of each bottom event on the operation of the power plant.

[0085] S43, the combined impact value I of all bottom events i The events are sorted from largest to smallest, and the bottom events with a comprehensive impact value greater than the probability threshold β are selected. These bottom events are the key failure factors.

[0086] Based on actual needs, the top n' events are selected as key failure factors to focus on, forming a list of key failure factors n' = {N'1, N'2, ..., N'}. n' This provides a basis for formulating subsequent operation and maintenance strategies;

[0087] S44. Combining the dynamically updated probability data of the bottom events in S36, periodically recalculate the comprehensive impact value of the bottom events and update the list of key failure factors.

[0088] Therefore, the present invention employs the above-mentioned intelligent operation and maintenance method for distributed photovoltaic power plants based on fault tree analysis, and the beneficial effects are as follows:

[0089] (1) This invention constructs a fault tree model, combines historical fault data and real-time parameters, accurately identifies key fault factors such as photovoltaic module aging and inverter overheating, traces from the bottom event to the top event, solves the traditional operation and maintenance fault location problem, provides early warning of high-risk faults, reduces the probability of shutdown, ensures stable power generation of the power station, and improves reliability.

[0090] (2) This invention relies on a dynamically updated list of fault probabilities and key factors, collects and transmits data in real time through LoRa technology, and combines fault tree algorithm to quickly match maintenance work orders, shorten fault response and repair time, reduce manpower input, and reduce maintenance costs.

[0091] (3) This invention makes full use of historical fault data and real-time operation data, associates faults with parameters through decision tree algorithm, and dynamically optimizes probability model with sliding time window, breaking the dilemma of "data cannot be used" in traditional operation and maintenance, mining fault patterns and equipment performance trends from massive data, providing data support for long-term planning of power plants, and promoting the upgrade of distributed photovoltaic power plants from "passive operation and maintenance" to "active intelligent operation and maintenance".

[0092] The technical solution of the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. Attached Figure Description

[0093] Figure 1 This is an overall flowchart of an embodiment of the intelligent operation and maintenance method for distributed photovoltaic power plants based on fault tree analysis according to the present invention;

[0094] Figure 2 This is a flowchart illustrating the construction of a fault tree model in an embodiment of an intelligent operation and maintenance method for distributed photovoltaic power plants based on fault tree analysis, according to the present invention.

[0095] Figure 3 This is a schematic diagram of the process for real-time acquisition of operating data of a distributed photovoltaic power station, according to an embodiment of the present invention of an intelligent operation and maintenance method for distributed photovoltaic power stations based on fault tree analysis.

[0096] Figure 4 This is a schematic diagram illustrating the process of calculating the probability of occurrence of each bottom event in a fault tree using fault tree analysis, as an embodiment of the intelligent operation and maintenance method for distributed photovoltaic power plants based on fault tree analysis according to the present invention. Detailed Implementation

[0097] The technical solution of the present invention will be further described below with reference to the accompanying drawings and embodiments.

[0098] Unless otherwise defined, the technical or scientific terms used in this invention shall have the ordinary meaning as understood by one of ordinary skill in the art to which this invention pertains.

[0099] like Figure 1 As shown, a method for intelligent operation and maintenance of distributed photovoltaic power plants based on fault tree analysis is characterized by the following steps:

[0100] S1. Construct a fault tree model for a distributed photovoltaic power station, including the following steps:

[0101] like Figure 2 As shown, S11, Historical Fault Data Collection and Preprocessing: Collect historical fault data from multiple distributed photovoltaic power stations within the target area, including basic fault data, operating parameter data, and environmental parameter data. The basic fault data includes the timestamp t of the fault occurrence, which is used for subsequent time-dimensional analysis; the fault type F includes photovoltaic module faults, inverter faults, combiner box faults, and transmission line faults, which are used to classify and identify faults; the power station number Sid of the fault occurrence is used to distinguish the fault data of different power stations.

[0102] Operating parameter data includes the output voltage V of the photovoltaic module. module Output current I module This is used to reflect the power generation capacity of photovoltaic modules; the input voltage I of the inverter. inverter-in Input current I inverter-in Output voltage V inverter-out Output current I inverter-out This reflects the operating status of the inverter; the input current I of the combiner box combiner-in The output current is used to display the current transmission status of the combiner box. Environmental parameter data includes ambient temperature (T), illuminance (S), and humidity (H). Environmental factors have a significant impact on the operation of photovoltaic power station equipment and need to be recorded in detail.

[0103] The collected historical fault data is preprocessed, including data cleaning, removal of missing and outlier values, and data standardization, which transforms data of different dimensions to a specific range to facilitate subsequent analysis. This invention uses a minimum-maximum standardization formula:

[0104]

[0105] Where X is the original data, X min and X max Let X be the minimum and maximum values ​​of this data feature, respectively. norm This is the standardized data.

[0106] S12. Calculate the frequency of occurrence of various faults within a certain time period, using the following formula:

[0107]

[0108] Among them, f F n represents the frequency of occurrence of fault type F. F This represents the number of times fault type F occurs within the statistical period, and N is the total number of occurrences of all faults within the statistical period. This formula determines the relative frequency of different fault types, allowing for the selection of high-frequency fault types as important candidate events in the fault tree model.

[0109] S13. Using the decision tree algorithm, analyze the correlation between fault types and operating parameters and environmental parameters. The specific process is as follows:

[0110] Data preparation: First, the historical fault data after S11 preprocessing is divided into feature set A and tag set Y. Feature set A includes operating parameters (such as the output voltage Vmodule and output current Imodule of photovoltaic modules) and environmental parameters (such as ambient temperature T and light intensity S). Each row represents a data record, and each column corresponds to a parameter. Tag set Y represents the fault type, and each element corresponds to the fault type F of a data record.

[0111] Selecting the splitting attribute: Information gain ratio is used as the metric for selecting the splitting attribute.

[0112] First, calculate the information entropy Ent(D) of dataset D, using the formula:

[0113]

[0114] Where |C| represents the number of fault types in the dataset, p i It represents the proportion of samples in the dataset belonging to the i-th fault type; for each attribute a, calculate its information gain Gain(D,a) with respect to dataset D, using the formula:

[0115]

[0116] Where V is the number of possible values ​​for attribute a, and D n It is a subset of samples in dataset D where attribute a takes the value v, |D v | and |D| represent subsets D, respectively. v And the number of samples in dataset D.

[0117] Next, calculate the intrinsic value of attribute a, using the formula IV(a), which is:

[0118]

[0119] Finally, the information gain ratio GainRatio(D,a) of attribute a is obtained, and the formula is:

[0120]

[0121] The attribute with the highest information gain ratio is selected as the partitioning attribute for the current node.

[0122] Using dataset D as the root node, the dataset is divided into multiple subsets according to the selected splitting attribute. Each subset corresponds to a child node. For each child node, the process of selecting splitting attributes and splitting the dataset is repeated to recursively construct a decision tree.

[0123] Construction will stop when one of the following conditions is met: all samples in the subset belong to the same fault type, the attribute set is empty, or the number of samples in the subset is less than a preset threshold N. threshold .

[0124] S14. Identify fault tree events, including top events, intermediate events, and bottom events.

[0125] The apex event is the final result of fault tree analysis, representing the most undesirable fault state for a distributed photovoltaic (PV) power station. Based on the core objectives of power station power generation efficiency and normal operation, it represents the apex event where the power generation efficiency E of the distributed PV power station falls below the normal threshold E0. th Set as the top event; the formula for calculating power generation efficiency E is as follows:

[0126]

[0127] Among them, P out P represents the actual output power of the power plant. in This is the theoretical maximum input power, calculated based on factors such as illumination and equipment rated parameters. When E < E0 th When a top-level failure is triggered, the power generation efficiency decline event is determined, indicating a top-level failure in the distributed photovoltaic power station. The occurrence of a top-level failure directly affects the normal operation and power generation efficiency of the power station, and is a failure state that needs to be focused on and avoided during operation and maintenance.

[0128] Intermediate events are the direct causes of top events. Intermediate events are determined by combining the high-frequency fault types screened in S12 with the correlations between fault types and operating and environmental parameters obtained in S13. The specific process is as follows:

[0129] First, high-frequency fault types among photovoltaic module faults, inverter faults, combiner box faults, and transmission line faults are classified as first-level intermediate events; for photovoltaic module faults: based on the analysis in S13, the intermediate events are further refined, if... or Abnormal output voltage and abnormal output current of photovoltaic modules are then considered as secondary intermediate events; among them, ΔV and ΔI are the average values ​​of voltage and current when the photovoltaic module is operating normally, and ΔV and ΔI are the set voltage and current fluctuation thresholds.

[0130] For inverter faults: when

[0131]

[0132] or hour,

[0133] Inverter input voltage abnormality, inverter input current abnormality, inverter output voltage abnormality, and inverter output current abnormality are treated as secondary intermediate events. These intermediate events can more specifically reflect the possible factors that lead to inverter failure.

[0134] in, or Input voltage V during normal operation inverter-in Input current I inverter-in Output voltage V inverter-out Output current I inverter-out The average value; △V inverter-in , △I inverter-in , △V inverter-out and △I inverter-out This represents the fluctuation threshold for the corresponding parameter.

[0135] The bottom event is the most basic event in the fault tree and is the root cause of intermediate events. The bottom event is determined based on the equipment's own defects, abnormal operating parameters, and environmental factors. At the same time, the bottom event is further clarified based on the correlation between the fault type and each parameter obtained from the decision tree analysis in S13.

[0136] Regarding inherent defects in the equipment, for photovoltaic modules, aging of photovoltaic modules, surface damage of photovoltaic modules, damage to inverter cooling fans, and inverter circuit board failures are considered bottom-line events. Regarding abnormal operating parameters, photovoltaic module output voltage exceeding the normal range (i.e., V...) is considered a bottom-line event. module >V max or V module <V min , where V max V min The voltage is within the normal range (both upper and lower limits); or the inverter input current fluctuates too much. As the base event, among which... σ is the input current fluctuation. th The set fluctuation threshold.

[0137] Regarding environmental factors, excessively high ambient temperatures can cause the inverter to overheat, i.e., T>T. th T th The upper limit of the inverter's normal operating temperature and the impact of a sudden drop in light intensity on photovoltaic module power generation, i.e., S < S th S th The threshold for light intensity affecting power generation and the fact that high humidity leads to a decrease in the insulation performance of transmission lines, i.e., H>H th H th The threshold value for the effect of humidity on insulation performance is used as the base event; for example, the decision tree shows that in a high temperature and high humidity environment, i.e., T>T, the threshold value is used. th And H>H thInverters are more prone to failure, so "high temperature and high humidity environment" can be considered as a bottom event related to inverter failure.

[0138] S2. Real-time collection of operational data from distributed photovoltaic power stations, including the following steps:

[0139] like Figure 3 As shown in Figure S21, deploy corresponding data acquisition sensors and equipment at key equipment and environmental monitoring points in the distributed photovoltaic power station: install voltage sensors and current sensors at the photovoltaic module array to collect the output voltage V of the photovoltaic modules. module and output current I module Among them, the voltage sensor adopts a high-precision Hall voltage sensor, which can monitor the changes in the output voltage of the photovoltaic module in real time, and the current sensor adopts a Rogowski coil current sensor, which can accurately measure the output current.

[0140] Voltage and current sensors are installed on the input and output sides of the inverter to collect the inverter's input voltage V. inverter-in Input current I inverter-in Output voltage V inverter-out and output current I inverter-out These sensors are used to obtain the operating status parameters of the inverter. They are all highly accurate and reliable, and can adapt to complex power plant operating environments.

[0141] Install current sensors at the input and output terminals of the combiner box to collect the input current I of the combiner box. combiner-in and output current I combiner-out Real-time monitoring of the current transmission in the combiner box ensures accurate acquisition of current data.

[0142] Environmental monitoring equipment, including temperature sensors, light intensity sensors, and humidity sensors, is installed at appropriate locations within the power plant area. The temperature sensor uses a platinum resistance temperature sensor to accurately measure the ambient temperature (T); the light intensity sensor uses a silicon photovoltaic cell sensor to acquire the light intensity (S) in real time; and the humidity sensor uses a capacitive humidity sensor to accurately monitor the ambient humidity (H). These environmental parameters are crucial for analyzing the power plant's operating status.

[0143] S22. Construct a stable and reliable data transmission network to transmit the collected data to the data processing center in real time. Specifically:

[0144] A data transmission network is built using LoRa technology to transmit the collected data to the data processing center in real time. The construction of this data transmission network includes the following aspects:

[0145] First, LoRa device deployment includes terminal node deployment: LoRa terminal node devices are deployed at various data acquisition points in the distributed photovoltaic power station. These LoRa terminal nodes are connected to voltage and current sensors of the photovoltaic modules, inverter parameter sensors, and environmental monitoring sensors to ensure accurate acquisition of operational and environmental data collected by the sensors. Each LoRa terminal node is configured with a unique device ID for identification and data differentiation within the network; for example, the ID for LoRa terminal nodes in the photovoltaic module area is assigned as "PV-001," and the ID for the inverter area is assigned as "INV-001," facilitating subsequent data processing and management.

[0146] Gateway Deployment: LoRa gateways should be deployed strategically based on the size and geographical distribution of the power station. For larger power stations or those with complex terrain, multiple gateways can be configured to extend signal coverage and ensure stable connections between all LoRa terminal nodes and the gateways. Gateways are typically deployed in relatively high and open locations within the power station, such as the rooftop of the monitoring room or a dedicated signal tower, to minimize signal obstruction and improve signal transmission quality. The gateways are connected to the data processing center via wired connections such as Ethernet or fiber optic cables for further data forwarding.

[0147] The second step, network parameter configuration, includes channel and frequency band settings: Based on local radio management regulations and actual application requirements, select appropriate LoRa channels and frequency bands. Common LoRa frequency bands include 433MHz, 868MHz (Europe), and 915MHz (North America). If multiple LoRa networks are operating simultaneously in the same area, channels must be planned appropriately to avoid channel conflicts. Through the LoRa gateway's management interface, configure the channel and frequency band parameters used for communication between terminal nodes and the gateway to ensure that all devices transmit data on the same wireless frequency.

[0148] Third, adjusting the spreading factor and transmit power: Adjust the spreading factor (SF) and transmit power of the LoRa device according to the data transmission distance and network load. The spreading factor (SF) typically ranges from 6 to 12. A larger spreading factor improves signal interference resistance and transmission distance but reduces data transmission rate; a smaller spreading factor has the opposite effect. For terminal nodes far from the gateway or with severe signal obstruction, increase the spreading factor (e.g., set to 10 or 12) and transmit power to ensure reliable data transmission. For nodes closer to the gateway, reduce the spreading factor and transmit power to reduce power consumption and interference. The spreading factor and transmit power of each terminal node can be customized using the gateway's configuration tool.

[0149] Fourth, Network ID and Key Configuration: To ensure network security, a unique network ID, namely NetID, is set for the LoRa network. All terminal nodes and gateways must use the same NetID to join the network. At the same time, network communication keys are configured, including session keys (used for data encryption) and application keys (used for application layer data decryption and verification). The correct NetID, session key, and application key are entered into the terminal node and gateway devices respectively to ensure that the data is encrypted during transmission and to prevent the data from being stolen or tampered with.

[0150] Fifth, data transmission and verification include data encapsulation and transmission: The LoRa terminal node encapsulates the collected operational and environmental data, adds device ID, timestamp, and other information, and generates data packets according to the format specified by the LoRa protocol. A CRC (Cyclic Redundancy Check) algorithm is used to calculate the checksum for the data packets, and the calculated CRC checksum is appended to the end of the data packet. After encapsulation and checksum calculation, the terminal node transmits the data packets to the gateway via the LoRa wireless channel according to a set transmission period (e.g., once every 10 seconds).

[0151] Sixth, Data Reception and Forwarding: The LoRa gateway monitors the wireless channel in real time and receives data packets from the terminal nodes. The gateway parses the received data packets, extracting the device ID, data content, and CRC checksum. It recalculates the CRC checksum of the data packet and compares it with the received checksum. If the checksums match, the data transmission is considered correct. The gateway extracts the valid data (such as photovoltaic module voltage, inverter current, etc.) from the data packet and forwards the data to the data processing center via a wired network (such as Ethernet) according to a pre-set protocol format. If the checksums do not match, the gateway sends a retransmission request to the terminal node, requesting the terminal node to retransmit the data packet until the data transmission is correct.

[0152] Seventh, Network Status Monitoring and Optimization: During data transmission, the LoRa gateway monitors the network status in real time, including indicators such as the connection status of terminal nodes, signal strength, data transmission rate, and packet loss rate. The network status data is visualized through the gateway's management interface or supporting monitoring software. When a terminal node is found to have excessively low signal strength or excessively high packet loss rate, the gateway can remotely adjust parameters such as the spreading factor and transmit power of that node, or notify maintenance personnel to inspect and maintain the node device to ensure the stable operation of the LoRa data transmission network.

[0153] S23. Real-time Data Acquisition and Preliminary Processing: The data acquisition device acquires various parameters in real time according to the set sampling frequency (e.g., once per second) and performs preliminary processing on the acquired data, including analog-to-digital conversion of acquired signals such as voltage, current, and temperature into digital signals for computer processing. Then, data filtering is performed to remove noise interference from the acquired data; a moving average filtering algorithm is used to average n consecutively acquired data points for each parameter to obtain the filtered data.

[0154] For example, the output voltage V of a photovoltaic module module Its filtered value V filter The calculation formula is:

[0155]

[0156] Among them, V module (ti) represents the output voltage value of the photovoltaic module collected at time ti, and n is the size of the filter window, which can be adjusted according to the actual situation. The pre-processed data is timestamped to accurately record the data acquisition time, ensuring the time accuracy of the data and providing accurate time dimension information for subsequent data analysis.

[0157] S24. Data Storage and Upload: The pre-processed and timestamped data is stored in a local cache, and in this embodiment, it is uploaded to the database of the data processing center every minute at regular intervals. During data storage, efficient data storage formats such as CSV and JSON are used to facilitate data storage and retrieval. Simultaneously, to prevent data loss, a local data backup mechanism is set up to periodically back up and store the collected data, ensuring data security and integrity.

[0158] S3. Using fault tree analysis, calculate the probability of occurrence of each bottom event in the fault tree based on the collected operational data, including the following steps:

[0159] like Figure 4 As shown, S31, establish a bottom event probability database: based on the historical fault data collected in S11, statistically analyze each bottom event B. i Combination of S under different operating states j Based on the frequency of occurrence of event B, construct a probability table for the base event and record the base event B. i Combined with operating state S j Number of co-occurring samples Total number of samples that occur in combination with motion state Among them, B i This represents the i-th underlying event (e.g., ambient temperature is too high). S jThis indicates that the j-th combination of operating states includes the discretized range of operating parameters and environmental parameters, such as "ambient temperature T≥35℃" and light intensity S>800lx.

[0160] S32, Real-time Status Mapping: Mapping the real-time operating parameter vector X collected by S2. real This includes parameters such as photovoltaic module output voltage and current, inverter input and output voltage and current, ambient temperature, and illuminance, which are mapped to discretized operating states S through equal-frequency binning or decision tree partitioning rules. current .

[0161] S33. Prior Probability Calculation and Correction: Based on historical data, the prior probability of the base event is calculated using the conditional probability formula.

[0162]

[0163] like Then, the Laplace smoothing method is used for correction:

[0164]

[0165] Where M is the total number of events at the bottom of the fault tree.

[0166] Combining the decision tree model of S13, the real-time running parameter vector X real Input this model, and the model will output the i-th bottom event B under the current parameters. i Let P be the probability of occurrence. dt (B i |X real The prior probability is then weighted and corrected using this result:

[0167] P'(B i )=α·P(B i |S current )+(1-α)P dt (B i |X real );

[0168] Where α is the weighting coefficient for historical data, which is usually taken as 0.7.

[0169] S34. Logic Gate Probability Calculation: In a fault tree, intermediate events are formed by connecting base events through logical relationships such as AND gates and OR gates. This invention calculates the probability of intermediate events occurring based on these logical relationships. The calculation formula is as follows:

[0170] AND gate probability calculation: When the intermediate event is composed of k base events B1, B2, ..., B... kWhen connected logically, it means that the intermediate event will only occur if the k base events occur simultaneously; according to the multiplication principle of probability, its probability P(AND) is calculated as follows:

[0171]

[0172] Among them, P(B) i ) represents the probability of the i-th bottom event occurring, and ∏ is the multiplication symbol. Multiplying the probabilities of the k bottom events together gives the probability of the intermediate events connected to the gate.

[0173] For example, if an intermediate event is formed by connecting two basic events, "dust accumulation on the surface of photovoltaic modules" and "light intensity below the threshold," through an AND gate, and the probabilities of these two basic events are known to be P(B1) = 0.1 and P(B2) = 0.2, then the probability of the intermediate event is P(AND) = 0.1 × 0.2 = 0.02.

[0174] OR gate probability calculation: If the intermediate event is composed of k base events B1, B2, ..., B... k By using OR logic connections, the intermediate event will occur as long as any one of the k basic events occurs. This invention first calculates the probability that none of the basic events occur, then subtracts this probability from 1 to obtain the probability P(OR) of the intermediate event occurring via the OR gate. The calculation formula is as follows:

[0175]

[0176] Among them, (1-P(B) i Let P(B1) represent the probability that the i-th basic event does not occur. Multiplying the probabilities of all basic events not occurring and subtracting this product from 1 gives the probability of the intermediate event connected by the OR gate. For example, if an intermediate event is caused by two basic events, namely, inverter cooling fan failure and inverter circuit board failure, connected by an OR gate, and their probabilities of not occurring are (1-P(B1) = 0.9 and (1-P(B2) = 0.8), then the probability of the intermediate event occurring is P(OR) = 1 - 0.9 × 0.8 = 0.28.

[0177] S35. Top Event Verification: The top event represents the least desirable fault state in a distributed photovoltaic power station. The probability P(E) of the top event is calculated using minimal cut sets. Assume the fault tree has m minimal cut sets C1, C2, ..., C m The probability of each cut set is P(C l The probability P(E) of the top event is calculated using the law of total probability:

[0178]

[0179] Among them, C oLet C represent the Oth minimal cut set. l Let ∑ represent the l-th minimal cut set. 1≤l＜O≤m P(C l ∩C o This represents the summation of the probabilities of two distinct minimal cut sets occurring simultaneously, because when the probabilities of the cut sets were accumulated separately earlier, the case of two cut sets occurring simultaneously was calculated repeatedly (in P(C)). l ) and P(C O Each value is calculated once, so these duplicate values ​​need to be subtracted to correct the probability calculation results.

[0180] S36. Dynamic Update: Because the operation and fault patterns of the power plant change over time, a sliding time window method is used to update historical statistical samples and optimize probability calculation. In this embodiment, the window size Z is set to 30 days. (Bottom Event B) i With running state S j The formula for updating the number of co-occurring samples is:

[0181]

[0182] in, This represents the original number of co-occurring samples. These are expired samples. For new samples, This is the updated sample; by continuously updating the sample and adjusting the probability of the bottom event in real time, the fault tree analysis can reflect the operating status of the power plant in real time.

[0183] S4. Identify key failure factors based on the probability of occurrence of basic events, specifically including the following steps:

[0184] S41. Setting Probability Thresholds and Weighting Coefficients: Based on actual power plant operation and maintenance experience and historical data, set a probability threshold β for the criticality of low-level events to initially screen out low-level events with a high probability of occurrence. Set weighting coefficients ω for different types of low-level events. i This reflects the varying degrees of impact of each event on the operation of the power plant.

[0185] Specifically: events related to equipment defects have a significant impact on the long-term stable operation of the power plant and can be assigned a higher weight; events related to environmental factors have a relatively lower weight. Where 0 ≤ ω i ≤1, and M represents the total number of events at the bottom of the fault tree.

[0186] S42. Calculate the overall impact value of the bottom events: For each bottom event B i Combined with its occurrence probability P'(B) i ) and weighting coefficient ω″ i Calculate the comprehensive impact value I i The calculation formula is as follows:

[0187] I i =ω″ i ×P'(B i );

[0188] Wherein, P'(B i ) represents the probability of occurrence of the bottom event after correction by S34. This formula quantifies the overall impact of each bottom event on the operation of the power plant.

[0189] S43. Sorting and Filtering Key Failure Factors: The combined impact value of all bottom events I i The events are sorted from largest to smallest, and those with a combined impact value greater than the probability threshold β are identified as critical failure factors. Furthermore, based on actual needs, the top n' events can be selected as key failure factors for focused attention, forming a critical failure factor list N' = {N'1, N'2, ..., N'}. n' This provides a basis for formulating subsequent operation and maintenance strategies. For example, if the comprehensive impact value of multiple basic events is calculated and sorted, it is found that the comprehensive impact value of 7 basic events is greater than β = 0.05, then these 7 basic events are identified as critical failure factors; if it is stipulated that the top 5 should be selected, then the top 5 basic events with the highest comprehensive impact value are selected as core critical failure factors.

[0190] S44. Dynamically Update the List of Critical Failure Factors: As time progresses and the power plant's operating status changes, the probability and impact of critical events will also change. Therefore, based on the dynamically updated probability data of critical events in S37, the comprehensive impact value of critical events should be recalculated periodically, such as monthly, and the list of critical failure factors should be updated to ensure that the operation and maintenance strategy always targets the currently existing critical failure factors, thereby improving the accuracy and effectiveness of operation and maintenance.

[0191] S5. Develop intelligent operation and maintenance strategies based on key failure factors: Establish an operation and maintenance strategy knowledge base and match corresponding operation and maintenance measures according to key failure factors; when the key failure factor is "dust accumulation on the surface of photovoltaic modules", trigger a cleaning work order and arrange manual cleaning; when the key failure factor is "inverter overheating", trigger a heat dissipation system inspection work order to adjust the cooling fan speed or clean the heat sink; when the key failure factor is "poor line contact", trigger a line inspection work order to check and repair line connections.

[0192] Example 1

[0193] I. Scenario and Objectives There are multiple distributed photovoltaic power stations in an industrial park. The components, inverters, and other equipment have been operating for a long time, and frequent failures affect power generation. The method proposed in this invention can accurately diagnose faults and perform intelligent operation and maintenance, thereby improving power generation efficiency and reliability.

[0194] II. Specific steps:

[0195] S1. Constructing a fault tree model includes:

[0196] S11. Data Collection and Preprocessing: Collect one year's worth of historical fault data from five power stations within the park, including fault timestamps, fault types such as components and inverters, power station numbers, and component voltage and current, inverter electrical parameters, and environmental temperature and humidity. Clean the data, removing missing and outlier values, and standardize it using minimum-maximum units.

[0197] S12. Screening high-frequency faults: Calculate the fault frequency and screen out high-frequency faults with a component fault rate of 35% and an inverter fault rate of 30% as model candidates.

[0198] S13. Decision tree analysis association: Operation and environmental parameters are divided into feature sets and fault types are labeled. Information gain ratio is used to select the partitioning attributes and construct a decision tree. It is found that "high ambient temperature + large inverter current fluctuation" is prone to triggering inverter faults.

[0199] S14. Determine the fault tree events: The top event is set as "power generation efficiency <80% or fault shutdown"; the first-level intermediate events are module failure and inverter failure; the second-level intermediate events are such as abnormal module voltage; the bottom events include module aging, inverter fan damage, and inverter overheating caused by high ambient temperature.

[0200] S2: Real-time data collection, the steps are as follows:

[0201] S21. Deploy data acquisition equipment: Hall voltage and Rogowski coil current sensors are installed in the component array; high-precision electrical parameter sensors are installed on the input / output side of the inverter; current sensors are installed in the combiner box; and platinum resistance temperature, silicon photovoltaic cell illumination, and capacitive humidity sensors are installed in the power station.

[0202] S22. Establish a LoRa network: Deploy terminal nodes connected to sensors, assigning them the ID PV-002; install the gateway on the roof of the power station monitoring building, connecting it to the data center via Ethernet. Configure the channel to 433MHz, spreading factor SF=8, and transmit power, and set the network ID and encryption key. The terminal transmits data every 10 seconds, which is verified by the gateway before being forwarded, monitoring the network status in real time.

[0203] S23. Data Acquisition and Processing: Acquire data per second, perform analog-to-digital conversion, apply moving average filtering, set a window size of 5, and add timestamps.

[0204] S24. Data storage and upload: Locally cached data, uploaded to the data center database every minute in CSV format, and locally backed up to prevent data loss.

[0205] S3. Calculate the probability of the bottom event occurring;

[0206] S31. Establish a probability database: Statistically analyze the co-occurrence frequency of low-level events in historical data, such as combinations of component aging and operating status (e.g., "temperature 35℃ + light intensity 800lx"), and construct a probability table.

[0207] S32. Real-time state mapping: Map the real-time component voltage (380V), temperature (32℃), etc., to discretized states using decision tree rules.

[0208] S33. Prior probability calculation correction: Calculate the prior probability using the conditional probability formula. If it is 0, then perform Laplace smoothing correction. Combine the real-time probability output of the decision tree with weighted correction, with a historical weight of 0.7.

[0209] S34. Logic gate probability calculation: For example, "component aging" and "voltage over-limit" are intermediate events connected through an AND gate, and the probability is the product of the probabilities of the two events; "inverter fan failure" or "circuit board failure" are intermediate events connected through an OR gate, and the probability is 1 minus the product of the probabilities of the two events not occurring.

[0210] S35. Calculation of the probability of the top event: Calculate the probability of the top event using the law of total probability through the minimum cut set.

[0211] S36. Dynamically update probability: Set a sliding time window of 30 days and update the sample to adjust the probability of the bottom event and adapt to changes in power plant operation.

[0212] S4. Identify key failure factors;

[0213] S41. Set thresholds and weights: Based on experience, set the probability threshold to 0.2, the weight of equipment defect type events to 0.6, and the weight of environmental factor type events to 0.4.

[0214] S42. Calculate the overall impact value: For example, if the probability of "component aging" after correction is 0.3 and the weight is 0.6, the overall impact value is 0.3 × 0.6 = 0.18.

[0215] S43. Screening key factors: Sort by comprehensive impact value and screen out 5 key failure factors such as "component aging" and "inverter fan failure" to form a list.

[0216] Table 1 List of Key Failure Factors

[0217]

[0218] S44. Dynamically Updated List: The overall impact value is recalculated monthly, and the list of key factors is updated to ensure the accuracy of the strategy.

[0219] S5. Develop intelligent operation and maintenance strategies: Match key failure factors with operation and maintenance measures: "Component aging" triggers a component replacement work order, arranging for professional personnel to replace it; "Inverter fan failure" triggers a fan repair work order, inspecting and replacing the fan, quickly responding to failures and improving operation and maintenance efficiency. After implementation, fault location time is reduced by 60%, fault repair time is reduced by 50%, power plant power generation efficiency is increased by 8%, and operation and maintenance costs are reduced by 30%, achieving intelligent and efficient operation and maintenance of distributed photovoltaic power plants.

[0220] Therefore, the present invention adopts the above-mentioned intelligent operation and maintenance method for distributed photovoltaic power plants based on fault tree analysis, which effectively breaks through the bottleneck of traditional operation and maintenance, adapts to the complex operation and maintenance needs of distributed photovoltaic power plants, and provides technical support for intelligent operation and maintenance in the industry.

[0221] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit them. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can still be made to the technical solutions of the present invention, and these modifications or equivalent substitutions cannot cause the modified technical solutions to deviate from the spirit and scope of the technical solutions of the present invention.

Claims

1. A method for intelligent operation and maintenance of distributed photovoltaic power plants based on fault tree analysis, characterized in that, Includes the following steps: S1. Construct a fault tree model for a distributed photovoltaic power station; S2. Real-time collection of operational data from distributed photovoltaic power stations; S3. Using fault tree analysis, calculate the probability of occurrence of each bottom event in the fault tree based on the collected operational data; S4. Identify key failure factors based on the probability of occurrence of basic events; S5. Develop intelligent operation and maintenance strategies based on key failure factors; S3 includes the following steps: S31. Based on the historical fault data collected in S11, statistical analysis of each basic event. Combination of different operating states Based on the frequency of occurrence of the events, construct a probability table for the underlying events and record the underlying events. Combination with operating status Number of co-occurring samples Total number of samples that occur in combination with motion state ; S32, Transfer the real-time operating parameter vector collected by S2 This includes the output voltage and current of photovoltaic modules, the input and output voltage and current of inverters, ambient temperature, and irradiance, which are mapped to discretized operating states through equal-frequency binning or decision tree partitioning rules. ; S33. Based on historical data, calculate the prior probability of the base event using the conditional probability formula: ； like Then, the Laplace smoothing method is used for correction: ； in, This represents the total number of events at the bottom of the fault tree. Combining the decision tree model of S13, the real-time running parameter vector Input this model, and the model will output the first [number]th [item] under the current parameters. The bottom line event The probability of occurrence is denoted as This probability value is used to weight and correct the prior probability: in, Historical data weighting coefficients; S34. Calculate the probability of intermediate events based on the logical relationships of AND and OR gates. The calculation formula is as follows: When intermediate events are The bottom line event , ,..., When connected logically, only if this The intermediate event will only occur if all the underlying events occur simultaneously. The probability of the intermediate event connected to the gate is... The calculation formula is: ; in, Indicates the first The probability of a base event occurring, ∏ is the multiplication sign; If the intermediate event is The bottom line event , ,..., Through or logical connection, as long as this If any one of the base events occurs, the intermediate event will occur; the probability of the intermediate event occurring in the OR gate connection is... The calculation formula is: ； in, Indicates the first The probability that a fundamental event will not occur; S35. Calculate the probability of the top event using minimal cut sets. Assume that the fault tree exists. m Minimum cut set , ,..., , No. The probability of the occurrence of a minimum cut set is The probability of the top event occurring is... The calculation uses the law of total probability: ； in, Indicates the first Minimal cut set Indicates the first Minimal cut set This represents the summation of the probabilities of two distinct minimal cut sets occurring simultaneously. S36. Employ a sliding time window method to continuously update historical statistical samples and optimize the probability of bottom events. running status The formula for dynamically updating the number of co-occurring samples is: ； in, This represents the original number of co-occurring samples. These are expired samples. For new samples, This is the updated sample.

2. The intelligent operation and maintenance method for distributed photovoltaic power plants based on fault tree analysis according to claim 1, characterized in that, S1 includes the following steps: S11. Collect historical fault data of multiple distributed photovoltaic power stations in the target area, including basic fault data, operating parameter data and environmental parameter data; The collected historical fault data is preprocessed, including data cleaning, removal of missing and outlier values, and data standardization, which converts data of different dimensions to a specific range. S12. Calculate the occurrence frequency of various faults within a certain time period, and select high-frequency fault types as important candidate events in the fault tree model: ； in, Indicates the fault type The frequency of occurrence, Indicates the fault type The number of times it occurs within the statistical period. This represents the total number of all faults occurring within the statistical period. S13. Use the decision tree algorithm to analyze the correlation between fault types and operating parameters and environmental parameters; S14. Identify fault tree events, including top events, intermediate events, and bottom events.

3. The intelligent operation and maintenance method for distributed photovoltaic power plants based on fault tree analysis according to claim 2, characterized in that, In S11, the basic fault data includes the timestamp of the fault occurrence. Fault type This includes photovoltaic module faults, inverter faults, combiner box faults, and transmission line faults, along with the power station numbers where the faults occurred. Operating parameter data includes the output voltage of the photovoltaic module. Output current The input voltage of the inverter Input current Output voltage Output current Input current of the combiner box Output current; environmental parameter data including ambient temperature. Light intensity ,humidity .

4. The intelligent operation and maintenance method for distributed photovoltaic power plants based on fault tree analysis according to claim 3, characterized in that, The specific steps for S13 are as follows: S131. First, divide the historical fault data after preprocessing in S11 into feature sets. and tag set ; among which feature set This includes runtime parameters and environment parameters; each row represents a data record, and each column corresponds to a parameter; tag set. For each fault type, each element corresponds to a fault type in a data record. ; S132. Using information gain ratio as the metric for selecting the splitting attribute, first calculate the dataset... Information entropy The formula is: ； in, This indicates the number of fault types in the dataset. It belongs to the first in the dataset. i The proportion of samples of each type of fault; For each attribute Calculate its effect on the dataset Information gain The formula is: ； in, It is an attribute The number of possible values, It is a dataset Medium attributes Values a subset of samples and Representing subsets and dataset The number of samples; Recalculated attributes The inherent value is given by the formula: The formula is: ； Finally, the attributes are obtained. Information gain rate The formula is: ； Select the attribute with the largest information gain ratio as the partitioning attribute of the current node; With dataset As the root node, the dataset is divided into multiple subsets based on the selected splitting attribute. Each subset corresponds to a child node. For each child node, the process of selecting splitting attributes and splitting the dataset is repeated to recursively build the decision tree.

5. The intelligent operation and maintenance method for distributed photovoltaic power plants based on fault tree analysis according to claim 3, characterized in that, In S14, the power generation efficiency of distributed photovoltaic power stations will be... Below the normal threshold Set as top event; power generation efficiency The calculation formula is as follows: ； in, This represents the actual output power of the power station. This is the theoretical maximum input power; when When this event occurs, a top-level failure is determined to have occurred in the distributed photovoltaic power station, indicating a decrease in power generation efficiency. Intermediate events are determined by combining the high-frequency fault types screened in S12 with the correlation between fault types and operating and environmental parameters obtained in S13. The specific process is as follows: First, high-frequency fault types among photovoltaic module faults, inverter faults, combiner box faults, and transmission line faults are classified as first-level intermediate events. For photovoltaic module failures, abnormal photovoltaic module output voltage and abnormal photovoltaic module output current are considered as secondary intermediate events. For inverter faults, abnormal inverter input voltage, abnormal inverter input current, abnormal inverter output voltage, and abnormal inverter output current are treated as secondary intermediate events. The bottom events are determined based on equipment defects, abnormal operating parameters, and environmental factors. Furthermore, the correlation between fault types and parameters obtained from the decision tree analysis in S13 is used to further clarify the bottom events. The specific process is as follows: Regarding equipment defects, photovoltaic module aging, photovoltaic module surface damage, inverter cooling fan failure, and inverter circuit board failure are considered as bottom events for photovoltaic modules. Regarding abnormal operating parameters, the photovoltaic module output voltage exceeding the normal range or the inverter input current fluctuating excessively is considered a bottom-line event. Regarding environmental factors, the bottom events are: excessively high ambient temperature causing inverter overheating, a sudden drop in light intensity affecting photovoltaic module power generation, and high humidity causing a decline in the insulation performance of transmission lines.

6. The intelligent operation and maintenance method for distributed photovoltaic power plants based on fault tree analysis according to claim 5, characterized in that, S2 includes the following steps: S21. Install voltage and current sensors at the photovoltaic module array to collect the output voltage of the photovoltaic modules. and output current ; Voltage and current sensors are installed on the input and output sides of the inverter to collect the inverter's input voltage. Input current Output voltage and output current ; Install current sensors at the input and output terminals of the combiner box to collect the input current of the combiner box. and output current Real-time monitoring of current transmission in the combiner box; Environmental monitoring equipment, including temperature sensors, light intensity sensors, and humidity sensors, will be installed within the power plant area to measure ambient temperature. Real-time acquisition of light intensity and monitoring of ambient humidity ; S22. A data transmission network is built using LoRa technology to transmit the collected data to the data processing center in real time. The data transmission network includes LoRa device deployment, channel and frequency band settings, adjustment of LoRa device spreading factor SF and transmit power, LoRa network ID and key configuration, data transmission and verification, and network monitoring and optimization. S23. The data acquisition device acquires each parameter in real time according to the set sampling frequency, and performs preliminary processing on the acquired data, including analog-to-digital conversion of the acquired analog signals and continuous acquisition of each parameter. The data is averaged to obtain filtered data. The data is then timestamped to record the time of data collection. S24. Store the pre-processed and timestamped data in a local cache, and upload the data to the database of the data processing center at certain time intervals.

7. The intelligent operation and maintenance method for distributed photovoltaic power plants based on fault tree analysis according to claim 6, characterized in that, S4 includes the following steps: S41. Based on actual power plant operation and maintenance experience and historical data, set a probability threshold for the criticality of basic events. This is used to initially screen out low-probability events; Set weighting coefficients for different types of bottom events This reflects the varying degrees of impact of each bottom event on the operation of the power plant; S42, For each bottom event Combined with its probability of occurrence and weighting coefficients Calculate its comprehensive impact value The calculation formula is as follows: ； in, The probability of occurrence of the bottom event after S34 correction is used to quantify the overall impact of each bottom event on the operation of the power plant. S43, Combined impact value of all bottom events Sort the data in descending order and filter out those with a comprehensive impact value greater than the probability threshold. These are the bottom-level events, which are the key failure factors; Based on actual needs, select the top rankings. The bottom-line event is identified as a key failure factor, and a list of key failure factors is formed. This provides a basis for formulating subsequent operation and maintenance strategies; S44. Combining the dynamically updated probability data of the bottom events in S36, periodically recalculate the comprehensive impact value of the bottom events and update the list of key failure factors.

Images