CN116054740A

CN116054740A - Fault detection and fault detection model processing method and device for photovoltaic power generation equipment

Info

Publication number: CN116054740A
Application number: CN202310096235.4A
Authority: CN
Inventors: 葛成; 王明
Original assignee: Alibaba Cloud Computing Ltd
Current assignee: Alibaba Cloud Computing Ltd
Priority date: 2023-01-17
Filing date: 2023-01-17
Publication date: 2023-05-02

Abstract

According to the embodiment of the application, firstly, power generation parameter values generated by the photovoltaic power generation equipment based on power generation output are obtained, then, a parameter fitting curve is constructed according to the power generation parameter values, and the parameter fitting curve is analyzed by using a fault detection model to obtain a corresponding fault detection result, wherein the fault detection model is used for extracting curve characteristics of the parameter fitting curve and carrying out equipment fault detection according to the curve characteristics. The photovoltaic power generation equipment corresponds to corresponding unique parameter fitting curves under different running states, so that fault detection results with higher accuracy can be obtained by carrying out fault detection on the photovoltaic power generation equipment based on the parameter fitting curves. Remote and large-scale fault detection of the photovoltaic power generation equipment can be realized by directly reading the real-time power generation parameter value, so that the labor cost of fault detection can be reduced, and the efficiency of fault detection is improved.

Description

Fault detection and fault detection model processing method and device for photovoltaic power generation equipment

Technical Field

The present disclosure relates to the field of photovoltaic power generation technologies, and in particular, to a fault detection method and apparatus for a photovoltaic power generation device, a fault detection model processing method and apparatus, an electronic device, and a storage medium.

Background

Photovoltaic power generation is a technology that utilizes the photovoltaic effect of semiconductor materials to directly convert light energy into electrical energy. In recent years, environmental problems are increasingly emphasized in various countries, and photovoltaic power generation technology with numerous advantages such as safety, reliability, high efficiency, no pollution and the like is becoming an important research focus on renewable clean energy sources. As photovoltaic technology advances to push photovoltaic power generation costs to continue to decrease, photovoltaic power generation formally enters a high-speed development stage that can compete with traditional energy sources.

Because the photovoltaic array (Photovoltaic Array) is exposed outdoors throughout the year, the photovoltaic equipment is easy to be influenced by the environment to generate faults, the photovoltaic array with faults cannot normally generate electricity, the service life of the photovoltaic array can be influenced, even some faults have potential safety hazards, and fire disasters can be possibly caused. In order to avoid the problems, the faults of the photovoltaic array can be detected by detecting the faults of the photovoltaic array in time, and the detected faults can be diagnosed and processed.

However, since photovoltaic power plants are typically located in remote areas, the detection of faults in photovoltaic devices typically requires significant effort and resources, and the accuracy of fault type identification is low. Therefore, how to reduce the fault detection cost of the photovoltaic equipment and improve the fault detection efficiency and accuracy has research significance.

Disclosure of Invention

The embodiments of the present application provide a method and an apparatus for detecting a fault of a photovoltaic power generation device and processing a fault detection model, so as to solve one or more of the above technical problems.

In a first aspect, an embodiment of the present application provides a method for detecting a fault of a photovoltaic power generation device, where the method includes:

acquiring a power generation parameter value generated by the photovoltaic power generation equipment based on power generation output;

constructing a parameter fitting curve according to the power generation parameter value;

and performing equipment fault detection on the parameter fitting curve based on a fault detection model to obtain a corresponding fault detection result, wherein the fault detection model is used for extracting the curve characteristics of the parameter fitting curve and performing equipment fault detection according to the curve characteristics.

In a second aspect, an embodiment of the present application provides a fault detection model processing method, where the method includes:

Acquiring power generation parameter sample values of the photovoltaic power generation equipment in various fault states;

constructing a parameter fitting curve sample according to the power generation parameter sample value;

and extracting the curve characteristics of the parameter fitting curve sample, and training a fault detection model according to the curve characteristics of the parameter fitting curve sample and fault information corresponding to the fault state.

In a third aspect, embodiments of the present application provide a fault detection method for a photovoltaic power generation device, where the method includes:

acquiring a power generation parameter value generated by the photovoltaic power generation equipment based on power generation output, wherein the power generation parameter value is used for determining a voltage value and a current value;

constructing a volt-ampere characteristic curve according to the voltage value and the current value;

converting the volt-ampere characteristic curve into a parameter change recursion chart, wherein the parameter change recursion chart characterizes the change trend of the power generation parameter value;

and inputting the parameter change recursion graph into a convolutional neural network for fault detection to perform equipment fault detection, and obtaining a corresponding fault detection result, wherein the fault detection result comprises whether faults exist or not and the fault type.

In a fourth aspect, embodiments of the present application provide an electronic device comprising a memory, a processor and a computer program stored on the memory, the processor implementing the method of any one of the preceding claims when the computer program is executed.

In a fifth aspect, embodiments of the present application provide a computer-readable storage medium having a computer program stored therein, which when executed by a processor, implements the method of any of the above.

Compared with the related art, the method has the following advantages:

according to the embodiment of the application, the power generation parameter value generated by the photovoltaic power generation equipment based on the power generation output is firstly obtained, then a parameter fitting curve is constructed according to the power generation parameter value, and the parameter fitting curve constructed by the power generation parameter value can be used for representing the operation state of the photovoltaic power generation equipment and can comprise a health state and various fault states. After the parameter fitting curve is constructed, the parameter fitting curve is analyzed by using a fault detection model so as to obtain a corresponding fault detection result. Because the photovoltaic power generation equipment corresponds to a corresponding unique parameter fitting curve under different running states, fault detection of the photovoltaic power generation equipment based on the curve characteristics of the parameter fitting curve can obtain a fault detection result with higher accuracy. Because the real-time power generation parameter value can be directly read from the photovoltaic power generation equipment or the device connected with the photovoltaic power generation equipment when the power generation parameter is obtained, the fault detection method provided by the embodiment of the application has the characteristic of large-scale deployment, and the remote fault detection of the photovoltaic power generation equipment can be realized, so that the labor cost of the fault detection can be reduced, and the efficiency of the fault detection is improved.

The fault detection model provided by the embodiment of the application is used for extracting the curve characteristics of the parameter fitting curve and carrying out equipment fault detection according to the curve characteristics. The model can identify various running states of the photovoltaic power generation equipment according to the input parameter fitting curve, and the fault types to be detected by the model can be flexibly added when the model is trained, so that the trained model can accurately distinguish various fault types, and the obtained fault detection result has higher accuracy. In addition, an equivalent circuit model can be built for the photovoltaic power generation equipment, and various fault states of the photovoltaic power generation equipment can be simulated by adding electrical elements or adjusting element parameters for the equivalent circuit model, so that a power generation parameter sample value for training the model can be obtained. By the method, the data volume of the power generation parameter sample value can be increased, and the power generation parameter sample value for training is provided for the fault condition which does not occur temporarily in the practical application, so that the trained fault detection model can provide a more accurate fault detection result.

In addition, in order to avoid false detection of a photovoltaic power generation apparatus affected by an external environment as a photovoltaic power generation apparatus that has failed, the acquired power generation parameter value may be corrected to a power generation parameter value corresponding to a standard environment after the power generation parameter value is acquired. And performing fault detection based on the corrected power generation parameter value, so that a more accurate fault detection result can be obtained.

The same set step length can be used for resampling the power generation parameter values, so that the resampled power generation parameter values are uniformly distributed, fault detection is performed on the photovoltaic power generation equipment based on the reconstructed parameter fitting curve, and the robustness of the fault detection can be improved.

Further, the parameter fitting curve can be converted into a parameter variation recursion map (RP). Compared with a parameter fitting curve, the parameter variation recursion chart uses the pixel values of the image pixel points to represent the variation of the power generation parameter values, so that the information represented by the image pixel points can be added on the basis of keeping the total information of the parameter fitting curve, the information represented by the parameter variation recursion chart is richer, and further, the fault detection is carried out by using the image characteristics of the parameter variation recursion chart, so that a more accurate fault detection result can be obtained.

The foregoing description is merely an overview of the technical solutions of the present application, and in order to make the technical means of the present application more clearly understood, it is possible to implement the present application according to the content of the present specification, and in order to make the above and other objects, features and advantages of the present application more clearly understood, the following detailed description of the present application will be given.

Drawings

In the drawings, the same reference numerals refer to the same or similar parts or elements throughout the several views unless otherwise specified. The figures are not necessarily drawn to scale. It is appreciated that these drawings depict only some embodiments according to the application and are not to be considered limiting of its scope.

FIG. 1 shows a schematic IV curve of a photovoltaic array in a healthy operating state;

FIG. 2 shows a schematic IV graph of a photovoltaic array in different states;

fig. 3 shows a schematic view of a scenario of a fault detection scheme of a photovoltaic power generation device provided in an embodiment of the present application;

fig. 4 shows a flowchart of a fault detection method of a photovoltaic power generation device provided in an embodiment of the present application;

FIG. 5A shows a set of IV graphs before and after data correction provided in an embodiment of the present application;

FIG. 5B illustrates another set of IV graphs before and after data correction provided in an embodiment of the present application;

FIG. 6 shows schematic diagrams of IV curves and respective corresponding RP-charts for three different states provided in the embodiments of the present application;

FIG. 7 shows a schematic structural diagram of a fault detection model provided in an embodiment of the present application;

FIG. 8 is a flowchart of a fault detection model processing method provided in an embodiment of the present application;

fig. 9 is a schematic circuit diagram of an analog circuit according to an embodiment of the present application;

FIG. 10 illustrates a flow chart of another method of fault detection for a photovoltaic power plant provided in an embodiment of the present application;

fig. 11 shows a block diagram of a fault detection device of a photovoltaic power generation apparatus provided in an embodiment of the present application;

FIG. 12 is a block diagram showing a fault detection model processing apparatus provided in an embodiment of the present application; and

fig. 13 shows a block diagram of a fault detection device of another photovoltaic power generation apparatus provided in an embodiment of the present application;

fig. 14 shows a block diagram of an electronic device used to implement an embodiment of the present application.

Detailed Description

Hereinafter, only certain exemplary embodiments are briefly described. As will be recognized by those of skill in the pertinent art, the described embodiments may be modified in various different ways without departing from the spirit or scope of the present application. Accordingly, the drawings and description are to be regarded as illustrative in nature, and not as restrictive.

In order to facilitate understanding of the technical solutions of the embodiments of the present application, the following describes related technologies of the embodiments of the present application. The following related technologies may be optionally combined with the technical solutions of the embodiments of the present application, which all belong to the protection scope of the embodiments of the present application.

In one related art prior to the present application, a shadowing or short circuit fault of a photovoltaic array may be detected by identifying a significant bright spot in infrared thermal imaging. However, since the modules in the photovoltaic array do not continue to warm up due to faults at lower irradiance levels, possible faults cannot be detected by way of identifying significant bright spots in infrared thermal imaging. In addition, the method also needs operation and maintenance management personnel to use a specific instrument to perform fault detection, so that the labor cost is high, and the detection efficiency is low.

In another related art prior to the present application, fault detection may be performed by estimating theoretical output voltage, current, and power of the photovoltaic array, and calculating a difference between the theoretical value and the actual value, through the difference. Because the realization of the technology is very dependent on the accuracy of theoretical value calculation, in the use process of the photovoltaic array, the accuracy theoretical value cannot be estimated due to the aging of electronic elements, so that the accuracy of fault detection is poor.

In view of the foregoing, embodiments of the present application provide a new fault detection scheme for a photovoltaic power generation device, so as to solve the above technical problems in whole or in part.

First, referring to fig. 1, the following description is given to technical terms related to the present application:

photovoltaic array: in the context of photovoltaic power generation, photovoltaic powerThe cell unit is the minimum unit for photoelectric conversion, the voltage generated by one cell unit is about 0.45V, and the working current is about 20-25 mA/cm ² . And packaging the photovoltaic battery monomers in series and parallel to form the photovoltaic array.

Standard test conditions (STC, standard Test Conditions): the standard test condition for the photovoltaic array is AM (Air-Mass, atmospheric quality) =1.5; irradiance =1000W/m ² The method comprises the steps of carrying out a first treatment on the surface of the Temperature=25℃. Here, am=1.5 means that the actual distance of light passing through the atmosphere is 1.5 times the vertical thickness of the atmosphere.

Volt-ampere characteristic curve (IV curve): under the condition that the working temperature and the irradiation degree are unchanged, when the load in the photovoltaic array changes, the interrelation between the output current and the output voltage of the photovoltaic array is realized. Fig. 1 shows an IV curve of a photovoltaic array in a healthy operating state.

Short circuit current (Isc): the maximum output current at a given solar intensity and temperature is shown in fig. 1, and the intersection point of the IV curve and the I axis is Isc, where u=0.

Open circuit voltage (Voc): the maximum output voltage at a given solar intensity and temperature is shown in fig. 1, where the intersection point of the IV curve and the V axis is Voc, and i=0.

Maximum power point power (Pmax): at a given solar radiation and temperature, the maximum power that the photovoltaic array may output; maximum power point current (Imp): at a given solar intensity and temperature, the photovoltaic array corresponds to the current at the maximum power point; maximum power point voltage (Vmp): at a given solar intensity and temperature, the photovoltaic array corresponds to the voltage at the maximum power point.

In practical applications, the photovoltaic array may fail due to environmental conditions or anomalies in its own elements. For example, failure conditions may result when the photovoltaic array is obscured by defoliation, dirt (shading), or when the circuitry of the photovoltaic array is open, shorted, or aged. Fig. 2 shows the IV curve morphology of the photovoltaic array in the healthy state (normal) and four fault states (shaded, open, short, aged), and the corresponding maximum power point. The maximum power point may be obtained by maximum power point tracking (MPPT, maximum Power Point Tracking). The maximum power point tracking means tracking control of the change of output voltage and current generated by the change of the surface temperature and the change of solar irradiation degree of the photovoltaic array, so that the array is always kept in a maximum output working state, and the automatic adjustment action of the maximum power output is obtained. As shown in fig. 2, since the power generation capability of the photovoltaic array may be affected by a fault, the IV curve of the photovoltaic array may take on a unique form corresponding to the fault state. Therefore, the operating state of the photovoltaic array can be identified by analyzing the morphological characteristics of the IV curve of the photovoltaic array, so that the fault detection of the photovoltaic array is realized. The photovoltaic array can be a photovoltaic array of a household roof solar device, a ground photovoltaic power station, or a photovoltaic array in a scene such as a photovoltaic power generation system (BAPV, building Attached Photovoltaic) attached to a building, a photovoltaic building integrated (BIPV, building Integrated Photovoltaics) and the like, and the embodiment of the application does not limit specific application scenes. The fault detection scheme of the photovoltaic power generation equipment provided by the embodiment of the application can rapidly and accurately provide a fault detection result for operation and maintenance managers of the photovoltaic array, so that the operation and maintenance managers can timely maintain the photovoltaic array with faults, and the photovoltaic array with the faults can return to a health state capable of providing maximum output as soon as possible, and economic loss and potential safety hazard caused by faults are reduced.

Fig. 3 is a schematic diagram of an exemplary application scenario for implementing the method of the embodiments of the present application. As shown in fig. 3, the photovoltaic array is connected to an Inverter (Inverter). The inverter can convert direct current generated by the photovoltaic array into alternating current, and the direct current generated by the photovoltaic array can be output outwards only through the processing of the inverter. In practical applications, the inverter may be used to obtain values of electrically related power generation parameters, such as voltage and current, output from the photovoltaic array. In this embodiment of the present application, a communication module may be configured for an inverter, and data may be sent to a data system (e.g., a database disposed on a cloud computing platform) connected to the communication module through a data transmission capability of the communication module. In order to obtain the temperature and the irradiation degree of the photovoltaic array during operation, a temperature sensor and an illuminometer are also arranged in the working environment of the photovoltaic array. Similarly, a communication module may also be configured for the temperature sensor and the light meter. When the photovoltaic array works, the inverter, the temperature sensor and the illuminometer can report the power generation parameter values, the temperature and the illumination degree of the photovoltaic array to the data system through the preset data acquisition frequency (such as 1 time every 1 hour).

In order to perform fault detection on the photovoltaic array, data reported by the inverter, the temperature sensor and the illuminometer first need to be acquired from the data system. In the embodiment of the application, the data to be acquired includes an output voltage value and a current value of the photovoltaic array, and a current irradiation degree and a current temperature corresponding to the acquired voltage value and current value. After the above data is acquired, data correction may be performed on the acquired data. The data correction method is to correct the voltage value and the current value to the standard irradiance (1000W/m ² ) And the corresponding values at standard temperature (25 ℃).

And then, resampling the data of the corrected voltage value and the corrected current value to further improve the accuracy of fault detection. Specifically, the sampling interval [0, vmax ] of the resampling voltage value and the sampling point number N may be set first. Where Vmax is the maximum value that the voltage value can take in data resampling. In setting Vmax, the open circuit voltage value Voc of the photovoltaic array in the STC environment and in a healthy state may be first determined, and then Vmax may be set to a value slightly greater than Voc. After the value interval [0, vmax ] and the sampling point number N are determined, resampling the voltage value at the step length of [0, vmax ] and the like, and constructing a voltage vector V containing N points for the voltage value obtained by resampling. And constructing an initial parameter fitting curve, namely an initial volt-ampere characteristic curve, according to the corrected voltage value and current value. Based on the initial volt-ampere characteristic curve and the voltage vector V, current values corresponding to the voltage values in the voltage vector V are obtained to construct a current vector I containing N points. And finally, constructing a volt-ampere characteristic curve for fault detection based on the voltage vector V and the current vector I obtained by data resampling. Because the photovoltaic array corresponds to a corresponding unique volt-ampere characteristic curve under different running states, the fault detection result with higher accuracy can be obtained by carrying out fault detection on the photovoltaic array based on the volt-ampere characteristic curve.

After the voltammetric characteristic curve is obtained, a recursion graph algorithm is used for converting the voltammetric characteristic curve into a parameter change recursion graph so as to extract morphological characteristics of the voltammetric characteristic curve. The parameter change recursion map is added with information represented by image pixel points on the basis of keeping the full information of the volt-ampere characteristic curve, so that compared with the direct use of the volt-ampere characteristic curve, the fault detection is carried out by using the parameter change recursion map with richer information, and the accuracy of the fault detection can be improved. Then, the parameter variation recursion chart is input into a failure detection model trained in advance to obtain a failure detection result. The fault detection result comprises identification information and fault type of the detection object (namely the photovoltaic array), and the obtained fault detection result is sent to the terminal equipment of the operation and maintenance manager of the power station, so that the operation and maintenance manager can be timely prompted to timely overhaul the photovoltaic array with the fault.

After receiving the fault detection result, the operation and maintenance manager can quickly position the photovoltaic array with the fault according to the identification information in the fault detection result, and prepares an overhaul tool according to the fault type. For example, for shielding faults, a cleaning tool can be used for cleaning the shielding object for the photovoltaic panel in a manual cleaning mode, or a cleaning robot is called; for an aging failure, electronic components for replacing the aging components may be prepared for maintenance of the failed photovoltaic module. When the same type of faults frequently occur in a plurality of photovoltaic arrays in the power station, a unified upgrading and reforming scheme can be also formulated for the power station, or a retirement and recovery scheme can be formulated for the photovoltaic arrays with high maintenance cost.

In practical application, an interactive interface for displaying detection results and operation and maintenance management can be built based on the cloud computing platform according to the requirements of fault detection, so that operation and maintenance management staff can conveniently and uniformly detect the operation state of the photovoltaic array.

The execution body of the embodiment of the present application may be an application, a service, an instance, a functional module in a software form, a Virtual Machine (VM), a container, a cloud server, or the like, or a hardware device (such as a server or a terminal device) or a hardware chip (such as a CPU, GPU, FPGA, NPU, AI accelerator card or a DPU) with a data processing function, or the like. The device for implementing fault detection of the photovoltaic power generation equipment can be deployed on a computing device of an application party providing corresponding services or a cloud computing platform providing computing power, storage and network resources, and the mode of externally providing services by the cloud computing platform can be IaaS (Infrastructure as a Service ), paaS (Platform as a Service, platform as a service), saaS (Software as a Service ) or DaaS (Data as a Service, data as a service). Taking the example that the platform provides SaaS software as a service (Software as a Service), the cloud computing platform can provide training of a neural network model or functional execution of a fault detection module of the photovoltaic power generation equipment by utilizing own computing resources, and a specific application architecture can be built according to service requirements. For example, the platform may provide a build service based on the above model to an application or person using the platform resources, and further invoke the above model and implement functions of online or offline fault detection based on a fault detection request of the photovoltaic power generation device submitted by a device such as a relevant client or server.

An embodiment of the present application provides a method for detecting a fault of a photovoltaic power generation device, as shown in fig. 4, which is a flowchart of a method 400 for detecting a fault of a photovoltaic power generation device according to an embodiment of the present application, where the method 400 may include:

in step S401, a power generation parameter value generated by the photovoltaic power generation apparatus based on the power generation output is acquired.

Photovoltaic power plants refer to devices that can convert light energy into electrical energy, such as devices that use solar energy for power generation based on photovoltaic arrays. The photovoltaic power generation equipment correspondingly has corresponding power generation capacity under different running states, and the power generation capacity of the photovoltaic power generation equipment can be evaluated based on the power generation parameter value of power generation output. Therefore, the power generation parameter value generated by the photovoltaic power generation equipment can be analyzed to detect the operation state of the photovoltaic power generation equipment, so that the purpose of fault detection is achieved. In addition, no additional detecting instrument or manual operation is needed when the power generation parameter value is obtained, so that the cost of fault detection is reduced.

The power generation parameters involved may include voltage, current, power, etc. generated by the photovoltaic power generation device based on the power generation output. When obtaining the power generation parameter value, can obtain through the mode of gathering data from photovoltaic power generation equipment or the device that is connected with photovoltaic power generation equipment, this embodiment of the application does not do the restriction to specific collection system. For example, for a photovoltaic array, the voltage and current values of the light energy output by the photovoltaic array may be obtained by an inverter connected to the photovoltaic array. Further, when acquiring, some or all of the above parameters may be selected, and specific values may be acquired. For example, the voltage value and the current value output by the photovoltaic power generation device can be obtained, and the selection of specific power generation parameters is not limited in the embodiment of the application. In addition, other power generation parameter values may be further calculated based on the acquired power generation parameter values based on the electrical law. For example, after the voltage value and the power value are obtained, the corresponding current value may be calculated according to the formula P (power) =u (voltage) ×i (current). In the embodiment of the present application, how to describe a fault detection method for a photovoltaic power generation device mainly using a photovoltaic array as an example of the photovoltaic power generation device, in practical application, fault detection may also be performed on devices using other forms of energy for generating power, such as wind power generation devices, hydroelectric power generation devices, and the like, based on the same design concept, which is not limited in this application.

In one possible implementation, the power generation parameter value of the photovoltaic power generation device may be obtained by accessing an inverter of the photovoltaic power generation device and reading the power generation parameter value collected by the inverter. The inverter can convert direct current generated by the photovoltaic array into alternating current, and electric energy generated by the photovoltaic array can be output outwards through the processing of the inverter.

For the inverter configured with the communication module, the data acquisition frequency may be preset (for example, 1 time every 1 hour), and the photovoltaic array power generation parameter value acquired by the inverter may be reported to the data system connected with the communication module through the data transmission capability of the communication module. Therefore, the power generation parameter value can be obtained in real time by periodically reading the data system, and the on-line fault detection of the photovoltaic array is realized.

In step S402, a parameter fitting curve is constructed from the power generation parameter values.

The parameter fitting curve referred to in the embodiments of the present application refers to a curve formed by a fitting relationship between at least two obtained power generation parameters, for example, a volt-ampere characteristic curve formed by a fitting relationship between a voltage value and a current value. After the power generation parameter value is obtained, a linear fitting algorithm such as a least square method can be used for constructing the fitting relation, and the algorithm selection for constructing the fitting relation is not limited in the embodiment of the application. Because the fitting relation is determined by various power generation parameters, and the values of the power generation parameters can reflect the power generation capacity of the photovoltaic power generation equipment, a curve constructed by the parameter fitting relation of the power generation parameter values carries more information, so that the fault detection of the photovoltaic power generation equipment is carried out through the parameter fitting curve, and a more accurate result can be obtained.

For the photovoltaic array, a fitting curve between the current value and the voltage value, namely a volt-ampere characteristic curve of the photovoltaic array, can be constructed according to the acquired current value and voltage value. When a photovoltaic array fails, the voltammetric characteristic curve of the photovoltaic array can show a unique form corresponding to the failure state. Therefore, the operating state of the photovoltaic array can be identified by analyzing the morphological characteristics of the IV curve of the photovoltaic array, so that the fault detection of the photovoltaic array is realized.

In one possible implementation, before the parameter fitting curve is constructed according to the power generation parameter values, the power generation parameter values may be corrected to values corresponding to the standard irradiation degree and the standard temperature according to the current irradiation degree and the current temperature corresponding to the time of collecting the power generation parameter values. The current illumination degree and the current temperature can be respectively acquired by an illuminometer and a temperature sensor which are arranged in the working environment of the photovoltaic power generation equipment. It will be appreciated that, similar to the inverter described above, the illumination meter and temperature sensor involved may be configured with a communication module, with the data transmission capabilities of the communication module, the current illumination level and current temperature collected by the illumination meter and temperature sensor may be transmitted to a data system connected to the communication module. The current illumination degree and the current temperature corresponding to the power generation parameter value can be obtained from the data system by setting the data acquisition frequency and the data acquisition time which are the same as those of the inverter for the illumination instrument and the temperature sensor.

The standard irradiance and standard temperature involved may be a set of irradiance and temperature that are preset. For example, the irradiation degree and temperature specified under standard test conditions, i.e., irradiation degree=1000W/m ² Temperature = 25 ℃; also can be the irradiance and temperature specified under other standards used in the industry, e.g. irradiance = 1000W/m ² Temperature = 20 ℃. In practical application, due to the influence of environmental reasons such as weather transformation and season rotation, different current illumination degrees and current temperatures are corresponding to the parameter values acquired for a plurality of times. The power generation capability of the photovoltaic array is greatly affected by the irradiation degree and the temperature, for example, the irradiation degree is low on overcast days, which results in poor power generation capability of the photovoltaic array. When the photovoltaic array is subjected to fault detection, the collected power generation parameter value is corrected to the value corresponding to the standard irradiation degree and the standard temperature, so that the condition that the capacity of the photovoltaic array is limited due to the two factors can be eliminated, and the fault detection accuracy is improved.

In one application example, the voltage value and the current value can be corrected to a standard irradiation degree (1000W/m) specified under standard test conditions by the following formula ² ) Values corresponding to standard temperature (25 ℃):

I ₂ ＝I ₁ (1+α _rel (T _m2 -T _m1 ))G ₂ /G ₁

V ₂ ＝V ₁ +V _OC1 [β _re l(T _m2 -T _m1 )+a·ln(G ₂ /G ₁ )]-R _s (I ₂ -I ₁ )-κ·I ₂ (T _m2 -T _m1 )

wherein I is ₁ 、V ₁ The current value and the voltage value before correction are respectively; i ₂ 、V ₂ The corrected current value and the corrected voltage value are respectively obtained; v (V) _OC1 Open circuit voltage for the photovoltaic array; t (T) _m1 And T _m2 The current temperature and the standard temperature are respectively; g ₁ And G ₂ Respectively the current irradiation degree and the standard irradiation degree; rs is the internal resistance of the photovoltaic array; alpha _rel And beta _rel The temperature coefficient of the current and the temperature coefficient of the voltage, respectively; a and k are constants, and the values of a and k can be set to be 1 during calculation.

Fig. 5A and 5B are schematic diagrams of IV curves before and after correction of two sets of data, wherein fig. 5A shows one set of IV curves of a photovoltaic array operating in a healthy state, and the other set of IV curves shown by fig. 5B is an IV curve of a photovoltaic array in a state with a partial occlusion failure. As can be seen from the IV curve before data correction (raw IV) and the IV curve after data correction (corrected IV) shown in fig. 5A and 5B, the power generation capability of the photovoltaic array in both the healthy state and the fault state is affected by illuminance and temperature, and the data before and after correction differ greatly.

In one possible implementation, after the parameter fitting curve is constructed according to the power generation parameter values, the power generation parameter values with uniform data distribution can be obtained through data resampling, and the parameter fitting curve is reconstructed based on the resampled power generation parameter values.

Specifically, when resampling data, the data of the partial distribution electric parameter value of the electric power generation parameter can be resampled according to the set step length and the parameter value range. The set step length refers to a resampling sampling step length, and a numerical value with uniform data distribution can be obtained by resampling data according to a certain step length. The parameter numerical range refers to the value range of the power generation parameter needing to be resampled, and the parameter numerical range can be determined according to the maximum value which can be taken by the corresponding power generation parameter under the health state of the photovoltaic power generation equipment. Alternatively, resampling of the power generation parameter values may be accomplished by presetting the number of sampling points (e.g., 50) and then performing equal step sampling within the parameter value range.

Since the parameter fitting curve is determined from at least two kinds of power generation parameter values, after resampling some of the power generation parameter values, the power generation parameter values of the remaining part of the power generation parameters can be determined from the fitting curve of the various parameters involved and the resampled power generation parameter values. For example, in the case where the parameter-fitted curve is determined from two kinds of power generation parameter values, one of the power generation parameter values is first resampled, and then the value of the other power generation parameter is redetermined based on the parameter-fitted curve and the resampled power generation parameter value.

Finally, reconstructing the parameter fitting curve based on the resampled power generation parameter values and the redetermined power generation parameter values to perform equipment fault detection according to the reconstructed parameter fitting curve. By performing fault detection on the photovoltaic power generation equipment based on the reconstructed parameter fitting curve, the robustness of fault detection can be improved.

In one example of an application, when resampling the voltage value and the current value acquired by the photovoltaic array, a voltage value range for resampling the voltage value may be determined first from an open circuit voltage value of the photovoltaic array in an STC environment and in a healthy state. The maximum value in the voltage value range is denoted as a maximum voltage value (Vmax), and in order to prevent inaccuracy of the fault detection result due to the voltage value exceeding the voltage limit, the maximum voltage value may be set to a value slightly larger than the open circuit voltage value. Then, 50 values are sampled in medium step in the voltage value range [0, vmax ] as resampled voltage values, and a vector V is constructed to record the 50 voltage values obtained by resampling, so that the volt-ampere characteristic curve constructed before resampling and the voltage vector V obtain current values corresponding to the voltage values in the voltage vector V, and a current vector I containing 50 current values is constructed. And finally reconstructing a volt-ampere characteristic curve for fault detection of the photovoltaic array based on the voltage vector V and the current vector I obtained by data resampling.

In step S403, device fault detection is performed on the parameter fitting curve based on a fault detection model, so as to obtain a corresponding fault detection result, where the fault detection model is used to extract the curve characteristic of the parameter fitting curve and perform device fault detection according to the curve characteristic.

The fault detection model provided by the embodiment of the application is a model trained in advance, and can perform feature extraction on input data to identify the running state of the photovoltaic power generation equipment, so that a fault detection result is output. The fault detection model may be, for example, a convolutional neural network (CNN, convolutional Neural Network) model, a residual network (res net, residual Neural Network) model, or a support vector machine (Support Vector Machine) model, or may be a machine learning model or a deep learning model integrated with multiple models. To improve the accuracy of the fault detection model, algorithms may also be adaptively added to the model, such as adding Regularization terms to the loss function of the model using Regularization (Regularization) algorithms, or incorporating attention mechanisms (Attention Mechanism) in constructing convolutional neural networks, etc. The embodiment of the application does not limit the selection of a specific machine learning model, a deep learning model or an algorithm.

In one possible implementation manner, when performing equipment fault detection on the parameter fitting curve based on a fault detection model to obtain a corresponding fault detection result, the parameter fitting curve may be first converted into a parameter variation recursion map. The parameter variation recursion map characterizes the variation of the power generation parameter values using the pixel values of the image pixels. Fig. 6 is a schematic diagram of IV curves (IV curves as marked in fig. 6) and respective corresponding RP graphs (RP plot) in three different states. The states involved include a health status (health status), a first occlusion fault status (PS 1), and a second occlusion fault status (PS 2). By comparing the IV curves in the three states with the RP graphs obtained by corresponding conversion of the IV curves, it can be found that the differences between the three RP graphs are more obvious than those of the three IV curves, and the RP graphs can carry more characteristic information, so that more accurate results can be obtained by using the RP graphs for fault detection.

The curve features of the parameter fitting curve, that is, the image features including the parameter variation recursion chart, are input into the fault detection model, and a corresponding fault detection result can be obtained. The fault detection model is used for extracting image features of the parameter variation recursion chart and predicting equipment faults of the photovoltaic power generation equipment based on the extracted image features.

Fig. 7 shows a schematic structural diagram of a fault detection model provided in an embodiment of the present application. As shown in fig. 7, the fault detection model is a machine learning model constructed based on a convolutional neural network. After the parameter variation recursion map is input into the model, a first convolution layer (CNN 1) in the model performs preliminary feature extraction on the input parameter variation recursion map. The extracted features are subjected to dimension reduction compression through the pooling layer so as to reduce the calculation amount of the subsequent network. Wherein, the Pooling layer can be selected from the maximum Pooling layer (Max Pooling). And then, further extracting the characteristics output by the pooling layer through a second convolution layer (CNN 2), obtaining the weight of the extracted characteristics aiming at each fault type through a full connection layer pair, judging the fault type through a cross entropy loss function (Cross Entropy Loss), and outputting a detection result.

Through the steps, the obtained fault detection result may include a fault type. After the fault detection result is obtained, the power generation module generating the fault in the photovoltaic power generation equipment can be positioned according to the fault type.

In addition, the communication application can be called to send a prompt message comprising the fault detection result so as to provide the prompt message for operation and maintenance management personnel of the photovoltaic power generation equipment. And the operation and maintenance manager of the photovoltaic power generation equipment can receive the prompt message through the terminal equipment. After receiving the prompt message, the photovoltaic power generation equipment generating the fault can be overhauled according to the fault equipment and the fault type indicated in the prompt message.

The embodiment of the present application further provides a fault detection model processing method, as shown in fig. 8, which is a flowchart of a fault detection model processing method 800 according to an embodiment of the present application, where the method 800 may include:

in step S801, power generation parameter sample values of the photovoltaic power generation apparatus in various fault states are acquired.

In order to train the fault detection model, first, a sample value of the power generation parameter for training needs to be obtained. The power generation parameter sample value can be a historical value generated by the photovoltaic power generation equipment in a historical time period based on power generation output, or can be obtained by an equivalent circuit model in a mode of constructing the equivalent circuit model for the power generation equipment. The obtained power generation parameter sample value corresponds to a label marked with a fault type, such as a shielding fault, an aging fault, a short circuit fault or an open circuit fault. It can be understood that, in order to avoid that the fault detection model erroneously reports the health status as a fault status, a numerical value generated when the photovoltaic power generation device or an equivalent circuit model corresponding to the photovoltaic power generation device does not have any fault may be added to the power generation parameter sample value, and the health status label is correspondingly marked.

In practical application, the data quantity of the generated parameter sample value is insufficient due to the fact that the historical value is true, so that training of the fault detection model cannot be completed. By constructing an equivalent circuit model for the photovoltaic power generation equipment, a power generation parameter sample value can be obtained from the output of the equivalent circuit model, so that the data volume of the power generation sample parameter value for training is supplemented to complete the training of the model. Under the condition that the historical values are sufficient, more power generation sample parameter values can be obtained by using the equivalent circuit model, and the accuracy of the training model is improved by increasing the data quantity of the power generation sample parameter values. Further, as components and allocation parameters can be flexibly added for the equivalent circuit model, faults which do not occur temporarily in various practical applications can be simulated through the equivalent circuit model, so that a more accurate fault detection result can be obtained when the photovoltaic power generation equipment fails truly.

In one application example, a neural network model may be selected as the fault detection model, and the fault detection model may be trained using the obtained power generation parameter sample values and the corresponding fault type labels. When the fault detection model is constructed, the depth of the neural network can be adjusted according to the data quantity of the parameter sample values and the quantity of fault types so as to improve the accuracy of the fault detection result. The embodiment of the application does not limit the specific structure inside the model.

In one possible implementation manner, when obtaining the power generation parameter sample values of the photovoltaic power generation equipment in various fault states, the equipment faults can be simulated through an equivalent circuit model of the photovoltaic power generation equipment, and the simulated power generation parameter sample values in the various fault states are collected.

When equipment faults are simulated through an equivalent circuit model of the photovoltaic power generation equipment, the equipment faults can be simulated through adding a target component corresponding to the fault type into the equivalent circuit model, the related target component can comprise a booster, a series resistor or a parallel resistor, and the related fault type can comprise a shielding fault, an aging fault, a short circuit fault or an open circuit fault. Further, for the same type of faults, the values of the target components can be adjusted to obtain the power generation parameter sample values provided by the equivalent circuit model under different fault degrees.

Fig. 9 shows an equivalent circuit model diagram of a photovoltaic power generation device according to an embodiment of the present application. The equivalent circuit model can be used to simulate the operating conditions of the photovoltaic array when in operation, including the health conditions when in normal operation, and the operating conditions when in failure. As shown in fig. 9, the photovoltaic array consists of two strings, each containing 3 photovoltaic modules (PV Cells). In parallel with each photovoltaic module is a Bypass Diode (Bypass Diode), which can be used to prevent thermal damage to the photovoltaic module in practical applications. The inputs to the equivalent circuit are two external environmental variables, namely, irradiance (Irradiance) and Temperature (Temperature). By adjusting the values of the two variables, the sample values of the power generation parameters corresponding to the operation states of the photovoltaic array under different irradiation degrees and temperature conditions can be simulated.

The string set consisting of PV Cells1, PV Cells2 and PV Cells3 is denoted as the first string set, and the string set consisting of PV Cells4, PV Cells5 and PV Cells6 is denoted as the second string set. A Gain is connected in front of the irradiation degree input interface of each photovoltaic module in the first string, and the input irradiation intensity of the photovoltaic module can be attenuated by adjusting the Gain so as to simulate shielding faults. The second series is connected in series with an open circuit resistor (Roc), and when the resistance value of Roc is small, the second series is normally connected, and when the resistance value of Roc is large, the second series is disconnected. By setting Roc to a larger resistance value, an open circuit failure of the second string group can be simulated.

The two strings are commonly connected with a series resistor (Rs_degra), when the resistance value of the Rs_degra is smaller, the photovoltaic array in the equivalent circuit is communicated normally, and when the resistance value of the Rs_degra is larger, the photovoltaic array fails due to consumption of the generated power by the Rs_degra, so that aging failure can be simulated by setting a larger resistance value for the Rs_degra. The two strings are also commonly connected with a parallel resistor (Rsh_degra) in parallel, when the resistance value of the Rsh_degra is large, the generated power is normally input to an external load, and when the resistance value of the Rsh_degra is small, the photovoltaic array is short-circuited. By setting a smaller resistance value for rsh_degra, a short circuit fault when a circuit short circuit occurs in the photovoltaic array can be simulated.

The two paths of photovoltaic strings are commonly connected to a voltmeter and an ammeter and are connected with a controlled voltage source for voltage regulation. In the equivalent circuit, the output values of the respective elements are uniformly output to the Out variable.

Table 1 shows a value range corresponding to a fault policy parameter corresponding to a fault state in the equivalent circuit provided in the embodiment of the present application.

TABLE 1 fault conditions and corresponding policy parameters

The corresponding fault codes, fault names, fault set descriptions, and fault simulation parameters for the 5 fault states are shown in table 1. The fault code is a tag name corresponding to different fault types used in the embodiment of the present application, and in order to distinguish faults of different degrees under the same fault type, a number is added to the fault code. For example, PS1 and PS2 respectively indicate a fault condition in which one PV module is partially blocked on one string and two PV modules are partially blocked on one string. The fault types referred to in table 1 include occlusion fault, open circuit fault (OC), aging fault (rs_degra), and short circuit fault (rsh_degra).

The contents shown in table 1 are specifically described below with reference to fig. 9. In order to obtain the sample values of the power generation parameters under various fault states, the values of the target components in the equivalent circuit model can be adjusted according to the values or the value ranges provided in the fault policy parameters shown in table 1, so that the sample values of the power generation parameters output by the equivalent circuit model are collected at the Out of the equivalent circuit model. For the shielding fault with the fault code PS1, the illuminance received by the photovoltaic module connected in series with any one of the gain 1, the gain 2 and the gain 3 can be controlled within the range of [10%,90% ] of the ambient illuminance by adjusting the value of the gain to simulate the fault state when one of the PV modules is partially shielded on the first string group as shown in fig. 9. Similarly, for an occlusion fault with a fault code of PS1, the values of any two of gain 1, gain 2, and gain 3 can be adjusted to simulate the fault condition when two PV modules are partially occluded on the first string as shown in fig. 9.

For an open circuit fault with a fault code OC, the resistance value of the open circuit resistor Roc in series with the second string set may be set to 10 ⁵ Omega to simulate a fault condition in which the second string fails open. For aging faults with a fault code of Rs_degra, the resistance value of a series resistor (Rs_degra) commonly connected with two strings can be adjusted to simulate the fault state of the photovoltaic array under different aging degrees, wherein the resistance value of the series resistor can be in the range of Rs epsilon [0.5 omega, 5 omega ]]. Finally, aiming at the short-circuit fault with the fault code Rsh_degra, the electronic value of a parallel resistor (Rsh_degra) connected in parallel with two strings can be adjusted to simulate the fault state of the photovoltaic array in the short-circuit fault, wherein the resistance value of the parallel resistor can be Rs epsilon [0.5 omega, 5 omega ] in the range of]。

Besides the fault states shown above, other target components can be added to the equivalent circuit model, or other parameter forms can be adjusted, so that other possible fault states can be simulated to obtain corresponding power generation parameter sample values, and the specific target component configuration or parameter value range is not limited.

In one possible implementation, the current and voltage values output by the equivalent circuit model may be collected while collecting the simulated power generation parameter sample values under various fault conditions. In connection with fig. 9, the voltage and current values output by the equivalent circuit model can be collected at Out. When the voltage value and the current value are obtained, a plurality of groups of voltage values and current values of a certain type of faults can be obtained by taking different parameter values in the parameter value range, and meanwhile, the values of the irradiation degree and the temperature input by the equivalent circuit model can be adjusted to simulate the voltage values and the current values output under the environments with different current irradiation degrees and current temperatures.

In step S802, a parameter fitting curve sample is constructed from the power generation parameter sample values.

In step S803, the curve features of the parameter fitting curve sample are extracted, and a fault detection model is trained according to the curve features of the parameter fitting curve sample and fault information corresponding to the fault state.

The obtained parameter sample values are correspondingly marked with fault codes (namely labels) corresponding to the fault states, and when the model is trained, the fault codes can be directly used as fault information, and the model is trained.

In one possible implementation manner, before the fault detection model is trained according to the curve characteristics of the parameter fitting curve sample and the fault information corresponding to the fault state, the electrical parameter sample values are distributed in the middle of the power generation parameter sample values according to a set step length and a parameter numerical range, then the residual power generation parameter sample values in the power generation parameter sample values are determined according to the parameter fitting curve and the resampled power generation parameter sample values, finally the parameter fitting curve sample is reconstructed based on the resampled power generation parameter sample values and the redetermined power generation parameter sample values, so that the fault detection model is trained by using the reconstructed parameter fitting curve sample subsequently. Reference may be made to the foregoing embodiments for specific implementation of data resampling, which are not described here again.

In one possible implementation manner, before the parameter fitting curve sample is constructed according to the power generation parameter sample value, the power generation parameter sample value may be corrected to a value corresponding to the standard irradiation degree and the standard temperature according to the current irradiation degree and the current temperature corresponding to the time of collecting the power generation parameter sample value. In conjunction with fig. 9, the current irradiance and the current temperature can be acquired while the voltage value and the current value output by the equivalent circuit model are acquired at Out. For the specific implementation of data correction, reference may be made to the foregoing embodiments, and details are not repeated here.

In one possible implementation manner, the curve features of the parameter fitting curve sample are extracted, and when a fault detection model is trained according to the curve features of the parameter fitting curve sample and fault information corresponding to the fault state, the curve features of the parameter fitting curve sample are firstly converted into a parameter variation recursive sample graph, and the parameter variation recursive sample graph uses pixel values of image pixel points to represent variation of power generation parameter sample values. And then training a fault detection model for detecting fault information based on the parameter variation recursion sample graph and the fault information, wherein the curve characteristics of the parameter fitting curve sample, namely, the image characteristics comprising the parameter variation recursion sample graph, are used for extracting the image characteristics of the parameter variation recursion sample graph, and predicting the equipment fault of the photovoltaic power generation equipment based on the extracted image characteristics.

It will be appreciated that the acquired parameter variation recursive sample map may also be divided into two parts, one part being used for training the model and the remaining part being used for sample testing. In an application example, the Accuracy (Accuracy) of the fault detection model provided in the embodiment of the present application is 0.97, the Macro Average (Macro Average) is 0.97, and the weighted Average is 0.98. Therefore, the fault detection model provided by the embodiment of the application can achieve higher accuracy. The following table shows specific test results of the fault detection model provided by the embodiment of the application.

Fault code	Precision	Recall	F1-score	Sample data volume
					Healthy	1.00	0.93	0.96	367
PS1	0.94	1.00	0.97	767
					PS2	1.00	1.00	1.00	741
OC	1.00	1.00	1.00	374
					Rs_degra	1.00	0.87	0.93	383
Rsh_degra	0.92	1.00	0.96	363

TABLE 2 specific test results of failure detection models

Precision, recall, and F1-score is the harmonic mean of Precision and Recall. As shown in Table 2, the fault detection model provided by the embodiment of the application can provide more accurate detection results for various fault categories. In addition, the fault detection model accurately distinguishes faults of the same type and different degrees, and as shown by test results of corresponding fault codes PS1 and PS2, the fault detection model can accurately identify shielding faults of different degrees generated by different numbers of fault photovoltaic modules.

The embodiment of the present application further provides a method for detecting a fault of a photovoltaic power generation device, as shown in fig. 10, which is a flowchart of a method 1000 for detecting a fault of a photovoltaic power generation device according to an embodiment of the present application, where the method 1000 may include:

in step S1001, power generation parameter values generated by the photovoltaic power generation array based on the power generation output are acquired, the power generation parameter values being used to determine a voltage value and a current value.

In step S1002, a volt-ampere characteristic is constructed from the voltage value and the current value.

In step S1003, the volt-ampere characteristic curve is converted into a parameter variation recursion map that characterizes a variation trend of the power generation parameter values.

In step S1004, the parameter variation recursion chart is input into a convolutional neural network for fault detection to perform equipment fault detection, so as to obtain a corresponding fault detection result, where the fault detection result includes whether a fault exists or not and a fault type.

Specific embodiments may be referred to the foregoing examples, and will not be described herein.

Corresponding to the application scene and the method of the method provided by the embodiment of the application, the embodiment of the application also provides a fault detection device of the photovoltaic power generation equipment. As shown in fig. 11, which is a block diagram illustrating a structure of a fault detection apparatus 1100 of a photovoltaic power generation device according to an embodiment of the present application, the apparatus 1100 may include:

A parameter value obtaining module 1101, configured to obtain a power generation parameter value generated by the photovoltaic power generation device based on the power generation output;

a fitting curve construction module 1102, configured to construct a parameter fitting curve according to the power generation parameter value;

the result obtaining module 1103 is configured to perform equipment fault detection on the parameter fitting curve based on a fault detection model, so as to obtain a corresponding fault detection result, where the fault detection model is configured to extract a curve characteristic of the parameter fitting curve and perform equipment fault detection according to the curve characteristic.

In one possible implementation, the parameter obtaining module 1101 may be specifically configured to access an inverter of the photovoltaic power generation device, and read a power generation parameter value acquired by the inverter.

In one possible implementation, the apparatus 1100 may further include:

and the parameter value correction module is used for correcting the power generation parameter value to the value corresponding to the standard irradiation degree and the standard temperature according to the current irradiation degree and the current temperature corresponding to the power generation parameter value when the power generation parameter value is acquired before the parameter fitting curve is constructed according to the power generation parameter value.

In one possible implementation, the apparatus 1100 may further include:

The resampling module is used for resampling data of part of the power generation parameter values in the power generation parameter values according to a set step length and a parameter value range after the parameter fitting curve is constructed according to the power generation parameter values; determining the rest power generation parameter values in the power generation parameter values according to the parameter fitting curve and the resampled power generation parameter values; reconstructing the parameter fitting curve based on the resampled power generation parameter values and the redetermined power generation parameter values to perform equipment fault detection according to the reconstructed parameter fitting curve.

In one possible implementation, the result obtaining module 1103 may further include:

a recursive graph conversion sub-module for converting the parameter fitting curve into a parameter variation recursive graph, the parameter variation recursive graph characterizing a variation of a power generation parameter value using pixel values of image pixels;

the result obtaining sub-module is used for inputting the parameter variation recursion chart into a fault detection model to obtain a corresponding fault detection result, and the fault detection model is used for extracting image features of the parameter variation recursion chart and predicting equipment faults of the photovoltaic power generation equipment based on the extracted image features.

In one possible implementation, the fault detection result includes a fault type, and the apparatus 1100 may further include:

and the fault positioning module is used for positioning the power generation module generating faults in the photovoltaic power generation equipment according to the fault type.

In one possible implementation, the apparatus 1100 may further include:

and the message sending module is used for sending a prompt message comprising the fault detection result by calling the communication application.

Corresponding to the application scenario and the method of the method provided by the embodiment of the application, the embodiment of the application also provides a fault detection model processing device. As shown in fig. 12, which is a block diagram of a fault detection model processing apparatus 1200 according to an embodiment of the present application, the apparatus 1200 may include:

the parameter sample value obtaining module 1201 is configured to obtain a power generation parameter sample value of the photovoltaic power generation device in a plurality of fault states;

a fitting curve sample construction module 1202 for constructing a parameter fitting curve sample according to the power generation parameter sample values;

the model training module 1203 is configured to extract the curve characteristic of the parameter fitting curve sample, and train a fault detection model according to the curve characteristic of the parameter fitting curve sample and fault information corresponding to the fault state.

In one possible implementation, the parameter sample value obtaining module 1201 may further include:

and the parameter sample value acquisition submodule is used for simulating equipment faults through an equivalent circuit model of the photovoltaic power generation equipment and acquiring power generation parameter sample values under various simulated fault states.

In one possible implementation manner, the parameter sample value acquisition sub-module may be specifically configured to simulate a device fault by adding, to the equivalent circuit model, a target component corresponding to a fault type, where the target component includes a booster, a series resistor, or a parallel resistor, and the fault type includes an occlusion fault, an aging fault, a short circuit fault, or an open circuit fault.

In one possible implementation, the parameter sample value acquisition sub-module may be specifically configured to acquire a current value and a voltage value output by the equivalent circuit model.

In one possible implementation, the model training module 1203 may be specifically configured to convert the curve features of the parameter fitting curve sample into a parameter variation recursive sample map, where the parameter variation recursive sample map characterizes a variation of the power generation parameter sample values using pixel values of image pixels; training a fault detection model for detecting fault information based on the parameter variation recursive sample graph and the fault information, wherein the curve characteristics of the parameter fitting curve sample comprise the image characteristics of the parameter variation recursive sample graph, and the fault detection model is used for extracting the image characteristics of the parameter variation recursive sample graph and predicting equipment faults of the photovoltaic power generation equipment based on the extracted image characteristics.

In one possible implementation, the apparatus 1200 may further include:

the resampling module is used for resampling data of a part of the power generation parameter sample values according to a set step length and a parameter numerical range before the fault detection model is trained according to the curve characteristics of the parameter fitting curve sample and the fault information corresponding to the fault state; determining the residual power generation parameter sample values in the power generation parameter sample values according to the parameter fitting curve samples and the resampled power generation parameter sample values; reconstructing the parameter fitting curve sample based on the resampled power generation parameter sample values and the redetermined power generation parameter sample values.

In one possible implementation, the apparatus 1200 may further include:

and the parameter sample value correction module is used for correcting the power generation parameter sample value to a value corresponding to the standard irradiation degree and the standard temperature according to the current irradiation degree and the current temperature corresponding to the power generation parameter sample value when the power generation parameter sample value is acquired before the parameter fitting curve sample is constructed according to the power generation parameter sample value.

Corresponding to the application scene and the method of the method provided by the embodiment of the application, the embodiment of the application also provides another fault detection device of the photovoltaic power generation equipment. As shown in fig. 13, a block diagram of a fault detection apparatus 1300 of a photovoltaic power generation device according to an embodiment of the present application, the apparatus 1300 may include:

A parameter value obtaining module 1301, configured to obtain a power generation parameter value generated by the photovoltaic power generation device based on a power generation output, where the power generation parameter value is used to determine a voltage value and a current value;

a volt-ampere characteristic curve construction module 1302 for constructing a volt-ampere characteristic curve based on the voltage value and the current value;

a recursive graph conversion module 1303, configured to convert the volt-ampere characteristic curve into a parameter variation recursive graph, where the parameter variation recursive graph characterizes a variation trend of a power generation parameter value;

the result obtaining module 1304 is configured to input the parameter variation recursion chart into a convolutional neural network for fault detection to perform equipment fault detection, and obtain a corresponding fault detection result, where the fault detection result includes whether a fault exists and a fault type.

The functions of each module in each device of the embodiments of the present application may be referred to the corresponding descriptions in the above methods, and have corresponding beneficial effects, which are not described herein.

Fig. 14 is a block diagram of an electronic device used to implement an embodiment of the present application. As shown in fig. 14, the electronic device includes: a memory 1401 and a processor 1402, the memory 1401 storing a computer program executable on the processor 1402. The processor 1402, when executing the computer program, implements the methods of the embodiments described above. The number of memories 1401 and processors 1402 may be one or more.

The electronic device further includes:

the communication interface 1403 is used for communicating with external devices and performing data interaction transmission.

If the memory 1401, the processor 1402, and the communication interface 1403 are implemented independently, the memory 1401, the processor 1402, and the communication interface 1403 can be connected to each other through buses and perform communication with each other. The bus may be an industry standard architecture (Industry Standard Architecture, ISA) bus, an external device interconnect (Peripheral Component Interconnect, PCI) bus, or an extended industry standard architecture (Extended Industry Standard Architecture, EISA) bus, among others. The bus may be classified as an address bus, a data bus, a control bus, etc. For ease of illustration, only one thick line is shown in fig. 14, but not only one bus or one type of bus.

Alternatively, in a specific implementation, if the memory 1401, the processor 1402, and the communication interface 1403 are integrated on a chip, the memory 1401, the processor 1402, and the communication interface 1403 may perform communication with each other through internal interfaces.

The present embodiments provide a computer-readable storage medium storing a computer program that, when executed by a processor, implements the methods provided in the embodiments of the present application.

The embodiment of the application also provides a chip, which comprises a processor and is used for calling the instructions stored in the memory from the memory and running the instructions stored in the memory, so that the communication device provided with the chip executes the method provided by the embodiment of the application.

The embodiment of the application also provides a chip, which comprises: the input interface, the output interface, the processor and the memory are connected through an internal connection path, the processor is used for executing codes in the memory, and when the codes are executed, the processor is used for executing the method provided by the application embodiment.

It should be appreciated that the processor may be a central processing unit (Central Processing Unit, CPU), but may also be other general purpose processors, digital signal processors (Digital Signal Processor, DSP), application specific integrated circuits (Application Specific Integrated Circuit, ASIC), field programmable gate arrays (Field Programmable Gate Array, FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, or the like. A general purpose processor may be a microprocessor or any conventional processor or the like. It is noted that the processor may be a processor supporting an advanced reduced instruction set machine (Advanced RISC Machines, ARM) architecture.

Further alternatively, the memory may include a read-only memory and a random access memory. The memory may be volatile memory or nonvolatile memory, or may include both volatile and nonvolatile memory. The nonvolatile Memory may include Read-Only Memory (ROM), programmable ROM (PROM), erasable Programmable ROM (EPROM), electrically Erasable EPROM (EEPROM), or flash Memory, among others. Volatile memory can include random access memory (Random Access Memory, RAM), which acts as external cache memory. By way of example, and not limitation, many forms of RAM are available. For example, static RAM (SRAM), dynamic RAM (Dynamic Random Access Memory, DRAM), synchronous DRAM (SDRAM), double Data Rate Synchronous DRAM (DDR SDRAM), enhanced SDRAM (ESDRAM), synchronous Link DRAM (SLDRAM), and Direct RAM (DR RAM).

In the above embodiments, it may be implemented in whole or in part by software, hardware, firmware, or any combination thereof. When implemented in software, may be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, the processes or functions in accordance with the present application are produced in whole or in part. The computer may be a general purpose computer, a special purpose computer, a computer network, or other programmable apparatus. Computer instructions may be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another.

In the description of the present specification, a description referring to terms "one embodiment," "some embodiments," "examples," "specific examples," or "some examples," etc., means that a particular feature, structure, material, or characteristic described in connection with the embodiment or example is included in at least one embodiment or example of the present application. Furthermore, the particular features, structures, materials, or characteristics described may be combined in any suitable manner in any one or more embodiments or examples. Furthermore, the different embodiments or examples described in this specification and the features of the different embodiments or examples may be combined and combined by those skilled in the art without contradiction.

Furthermore, the terms "first," "second," and the like, are used for descriptive purposes only and are not to be construed as indicating or implying a relative importance or implicitly indicating the number of technical features indicated. Thus, a feature defining "a first" or "a second" may explicitly or implicitly include at least one such feature. In the description of the present application, the meaning of "a plurality" is two or more, unless explicitly defined otherwise.

Any process or method described in flow charts or otherwise herein may be understood as representing modules, segments, or portions of code which include one or more executable instructions for implementing specific logical functions or steps of the process. And the scope of the preferred embodiments of the present application includes additional implementations in which functions may be performed in a substantially simultaneous manner or in an opposite order from that shown or discussed, including in accordance with the functions that are involved.

Logic and/or steps described in the flowcharts or otherwise described herein, e.g., may be considered a ordered listing of executable instructions for implementing logical functions, can be embodied in any computer-readable medium for use by or in connection with an instruction execution system, apparatus, or device, such as a computer-based system, processor-containing system, or other system that can fetch the instructions from the instruction execution system, apparatus, or device and execute the instructions.

It is to be understood that portions of the present application may be implemented in hardware, software, firmware, or a combination thereof. In the above-described embodiments, the various steps or methods may be implemented in software or firmware stored in a memory and executed by a suitable instruction execution system. All or part of the steps of the methods of the embodiments described above may be performed by a program that, when executed, comprises one or a combination of the steps of the method embodiments, instructs the associated hardware to perform the method.

In addition, each functional unit in each embodiment of the present application may be integrated in one processing module, or each unit may exist alone physically, or two or more units may be integrated in one module. The integrated modules may be implemented in hardware or in software functional modules. The integrated modules described above, if implemented in the form of software functional modules and sold or used as a stand-alone product, may also be stored in a computer-readable storage medium. The storage medium may be a read-only memory, a magnetic or optical disk, or the like.

The foregoing is merely exemplary embodiments of the present application, but the scope of the present application is not limited thereto, and any person skilled in the art can easily think of various changes or substitutions within the technical scope of the present application, which should be covered in the scope of the present application. Therefore, the protection scope of the present application shall be subject to the protection scope of the claims.

Claims

1. A fault detection method for a photovoltaic power plant, comprising:

2. The method of claim 1, wherein the obtaining a power generation parameter value generated by the photovoltaic power generation device based on a power generation output comprises:

and accessing an inverter of the photovoltaic power generation equipment, and reading the power generation parameter values acquired by the inverter.

3. The method of claim 1, wherein prior to said constructing a parameter fit curve from said power generation parameter values, the method further comprises:

And correcting the power generation parameter value to a value corresponding to the standard irradiation degree and the standard temperature according to the current irradiation degree and the current temperature corresponding to the power generation parameter value when the power generation parameter value is acquired.

4. The method of claim 1, wherein after said constructing a parameter fit curve from said power generation parameter values, the method further comprises:

resampling data of partial power generation parameter values in the power generation parameter values according to a set step length and a parameter value range;

determining the rest power generation parameter values in the power generation parameter values according to the parameter fitting curve and the resampled power generation parameter values;

reconstructing the parameter fitting curve based on the resampled power generation parameter values and the redetermined power generation parameter values to perform equipment fault detection according to the reconstructed parameter fitting curve.

5. The method of claim 1, wherein the performing equipment fault detection on the parameter fitting curve based on a fault detection model, obtaining a corresponding fault detection result comprises:

converting the parameter fitting curve into a parameter variation recursion diagram, wherein the parameter variation recursion diagram uses pixel values of image pixel points to represent the variation of power generation parameter values;

Inputting the parameter variation recursion graph into a fault detection model to obtain a corresponding fault detection result, wherein the curve characteristics of the parameter fitting curve comprise the image characteristics of the parameter variation recursion graph, and the fault detection model is used for extracting the image characteristics of the parameter variation recursion graph and predicting the equipment fault of the photovoltaic power generation equipment based on the extracted image characteristics.

6. The method of claim 1, wherein the fault detection result comprises a fault type, the method further comprising:

and positioning a power generation module generating faults in the photovoltaic power generation equipment according to the fault types.

7. A fault detection model processing method, comprising:

8. The method of claim 7, wherein the obtaining power generation parameter sample values for the photovoltaic power plant under the plurality of fault conditions comprises:

And simulating equipment faults through an equivalent circuit model of the photovoltaic power generation equipment, and collecting simulated power generation parameter sample values under various fault states.

9. The method of claim 8, wherein simulating an equipment failure by an equivalent circuit model of a photovoltaic power plant comprises:

and simulating equipment faults through adding a target component corresponding to the fault type in the equivalent circuit model, wherein the target component comprises a gain, a series resistor or a parallel resistor, and the fault type comprises shielding faults, aging faults, short circuit faults or open circuit faults.

10. The method of claim 8, wherein the collecting simulated power generation parameter sample values for a plurality of fault conditions comprises:

and collecting the current value and the voltage value output by the equivalent circuit model.

11. The method of claim 7, wherein the extracting the curve features of the parameter-fitted curve sample and training a fault detection model based on the curve features of the parameter-fitted curve sample and fault information corresponding to the fault condition comprises:

converting the curve characteristics of the parameter fitting curve sample into a parameter variation recursive sample graph, wherein the parameter variation recursive sample graph uses pixel values of image pixel points to represent the variation of the power generation parameter sample values;

Training a fault detection model for detecting fault information based on the parameter variation recursive sample graph and the fault information, wherein the curve characteristics of the parameter fitting curve sample comprise the image characteristics of the parameter variation recursive sample graph, and the fault detection model is used for extracting the image characteristics of the parameter variation recursive sample graph and predicting equipment faults of the photovoltaic power generation equipment based on the extracted image characteristics.

12. The method of claim 11, wherein prior to training the fault detection model based on the curve characteristics of the parameter-fitted curve samples and fault information corresponding to the fault condition, the method further comprises:

resampling data of a part of the power generation parameter sample values according to a set step length and a parameter value range;

determining the residual power generation parameter sample values in the power generation parameter sample values according to the parameter fitting curve samples and the resampled power generation parameter sample values;

reconstructing the parameter fitting curve sample based on the resampled power generation parameter sample values and the redetermined power generation parameter sample values.

13. The method of claim 7, wherein prior to said constructing a parametric fit curve sample from said power generation parameter sample values, the method further comprises:

And correcting the power generation parameter sample value to a value corresponding to the standard irradiation degree and the standard temperature according to the current irradiation degree and the current temperature corresponding to the power generation parameter sample value.

14. A fault detection method for a photovoltaic power plant, comprising:

15. An electronic device comprising a memory, a processor and a computer program stored on the memory, the processor implementing the method of any one of claims 1-14 when the computer program is executed.

16. A computer readable storage medium having stored therein a computer program which, when executed by a processor, implements the method of any of claims 1-14.