A remote state intelligent sensing method, device and system for photovoltaic combiner box

Abstract

The present application relates to the field of photovoltaic power generation technology, and specifically to a remote state intelligent perception method, device and system of a photovoltaic combiner box, the method comprising: collecting the current of a photovoltaic cell array, the temperature and illumination of a photovoltaic combiner box, obtaining the current sequence of the photovoltaic cell array, the temperature sequence and the illumination sequence of the photovoltaic combiner box; determining the current proportion of the photovoltaic cell array at the current sampling moment; determining the current fluctuation characteristic value of the photovoltaic cell array; determining the current characteristic vector of the photovoltaic combiner box at the current collection moment and a preset number of historical moments, obtaining a trained random forest, and obtaining the state judgment value of the photovoltaic combiner box corresponding to the current collection moment, and realizing remote state intelligent perception of the photovoltaic combiner box according to the state judgment value of the photovoltaic combiner box corresponding to the current collection moment. The present application solves the problem of omission or false alarm in abnormal detection of photovoltaic combiner boxes caused by complex current changes in photovoltaic cell arrays due to changes in ambient temperature.

Description

Technical Field

[0001] The present application relates to the field of photovoltaic power generation technology, and specifically to a remote state intelligent sensing method, device and system for a photovoltaic combiner box. Background Art

[0002] The photovoltaic system contains a large number of photovoltaic cell arrays with complex connections. At the same time, the photovoltaic cell array in the photovoltaic system in a complex environment has a high frequency of internal short circuits and other faults. The faulty photovoltaic cell array is easy to cause damage to the photovoltaic system. The photovoltaic combiner box can connect the photovoltaic cell array and the inverter, reducing the connection of the photovoltaic cell array to the inverter, and facilitating the inspection and maintenance of the photovoltaic system when a fault occurs. In order to reduce the impact of faults on the photovoltaic system, the photovoltaic system can be monitored according to the measurement and control device set in the photovoltaic combiner box.

[0003] Due to the complex environment in which the photovoltaic system is located, the parameters of the photovoltaic cell array will change significantly due to changes in ambient temperature, and complex current changes will also occur in the photovoltaic combiner box, which may easily cause the photovoltaic combiner box to misjudge abnormal detection of the photovoltaic system, resulting in missed or false reports of faults, affecting the accuracy of photovoltaic system fault handling. Summary of the invention

[0004] In order to solve the above technical problems, the purpose of this application is to provide a remote state intelligent sensing method, device and system for a photovoltaic combiner box. The technical solutions adopted are as follows:

[0005] In a first aspect, an embodiment of the present application provides a method for intelligently sensing the remote state of a photovoltaic combiner box, the method comprising the following steps:

[0006] Collect the current of the photovoltaic cell array, the temperature of the photovoltaic combiner box, and the illumination of the photovoltaic combiner box at the current collection time and a preset number of collection times before the current collection time, and obtain the current sequence of the photovoltaic cell array, the temperature sequence of the photovoltaic combiner box, and the illumination sequence of the photovoltaic combiner box at the current sampling time;

[0007] According to the current sequence of all photovoltaic cell arrays connected in series to the photovoltaic combiner box at the current sampling moment, the current proportion of each photovoltaic cell array at the current sampling moment is determined respectively;

[0008] The current sequence of the photovoltaic cell array is divided into a first preset threshold number of subsequences, and a normalized subsequence is obtained. The current fluctuation ratio of the subsequence is determined according to the difference between the currents contained in the subsequences and the difference between the currents contained in the current sequence from which the subsequences are divided, and the current fluctuation characteristic value of the photovoltaic cell array is determined by combining the correlation between the normalized subsequences of all photovoltaic cell arrays connected in series with the photovoltaic combiner box corresponding to the same time period;

[0009] According to the current proportion, current fluctuation characteristic value, temperature sequence, illumination sequence and current of all photovoltaic cell arrays connected in series with the photovoltaic combiner box at the current collection moment, the current characteristic vector of the photovoltaic combiner box at the current collection moment is determined, the current characteristic vectors of the photovoltaic combiner box corresponding to a preset number of historical moments are determined, and the trained random forest is obtained. The current characteristic vector of the photovoltaic combiner box corresponding to the current collection moment is input into the trained random forest to obtain the decision value of the decision tree, and all photovoltaic combiner boxes in the same decision tree are obtained. Combined with the difference between the temperature sequences of two different photovoltaic combiner boxes, the difference and similarity between the illumination sequences, and the difference between the currents of all photovoltaic cell arrays connected in series with the photovoltaic combiner box at the current collection moment, the state judgment value of the photovoltaic combiner box corresponding to the current collection moment is obtained. According to the state judgment value of the photovoltaic combiner box corresponding to the current collection moment, remote state intelligent perception of the photovoltaic combiner box is realized.

[0010] Furthermore, the process of determining the current proportion of each photovoltaic cell array at the current sampling moment is as follows:

[0011] The average value of all currents contained in the current sequence of the photovoltaic cell array at the current sampling moment is recorded as the current average value of the photovoltaic cell array at the current sampling moment;

[0012] The cumulative sum of the current averages of all photovoltaic cell arrays connected in series to the photovoltaic combiner box at the current sampling moment is recorded as the total current average of all photovoltaic cell arrays connected in series to the photovoltaic combiner box;

[0013] The ratio of the current average of the photovoltaic cell array to the total current average is recorded as the current proportion of the photovoltaic cell array at the current sampling moment.

[0014] Furthermore, the process of determining the current fluctuation ratio of the subsequence is as follows:

[0015] The standard deviation of all currents contained in the current sequence is recorded as the total current fluctuation of all subsequences obtained by evenly dividing the current sequence;

[0016] The ratio of the standard deviation of all currents contained in the subsequence to the total current fluctuation is recorded as the current fluctuation ratio of the subsequence.

[0017] Furthermore, the specific process of obtaining the current fluctuation characteristic value of the photovoltaic cell array is as follows:

[0018]

[0019] in, Indicates Current fluctuation characteristic value of a photovoltaic cell array; Indicates The number of other photovoltaic cell arrays connected in parallel to the same photovoltaic combiner box; Indicates the current collection time. The current sequence of the photovoltaic cell array is divided into The current fluctuation ratio of each subsequence; represents a first preset threshold; Indicates the current collection time. The current sequence of the photovoltaic cell array is divided into The normalized subsequence is the same as the current acquisition time. The current sequence of the photovoltaic cell array is divided into The Nth value in the cross-correlation sequence of the normalized subsequences, where N represents the amount of data contained in the normalized subsequence.

[0020] Furthermore, the current characteristic vector of the photovoltaic combiner box at the current collection moment is:

[0021]

[0022] in, They represent the first to the The current proportion of each photovoltaic cell array, Represents the number of all photovoltaic cell arrays connected in series with the photovoltaic combiner box; They represent the first to the Current fluctuation characteristic value of a photovoltaic cell array; It represents the cumulative sum of the currents of all photovoltaic cell arrays connected in series with the photovoltaic combiner box at the current collection moment; Indicates the average value of all temperatures contained in the temperature series of the PV combiner box at the current acquisition moment; Indicates the average value of all illuminances contained in the illuminance sequence of the PV combiner box at the current acquisition moment.

[0023] Furthermore, the state determination value of the photovoltaic combiner box corresponding to the current collection moment is obtained by combining the difference between the temperature sequences of two different photovoltaic combiner boxes, the difference and similarity between the illumination sequences, and the difference between the currents of all photovoltaic cell arrays connected in series with the photovoltaic combiner box at the current collection moment, including the specific method of:

[0024] The accumulated current of all photovoltaic cell arrays connected in series with the photovoltaic combiner box at the current collection time is recorded as the total current of the photovoltaic combiner box;

[0025] According to the difference between the temperature sequences of two different photovoltaic combiner boxes and the similarity between the illumination sequences, the environmental difference values ​​of the two different photovoltaic combiner boxes are determined, and the calculation formula is:

[0026]

[0027] in, Indicates The first PV combiner box and the Environmental difference value of each PV combiner box; Indicates The mean of all temperatures contained in the temperature series of the first PV combiner box is The absolute value of the difference between the mean values ​​of all temperatures contained in the temperature series of the PV combiner boxes; Indicates The average of all illuminances contained in the illuminance sequence of the first PV combiner box is The absolute value of the difference between the mean values ​​of all illuminances contained in the illuminance sequence of the PV combiner boxes; Indicates The temperature series of the first PV combiner box and the The cosine similarity of the temperature series of the PV combiner boxes; Indicates The illumination sequence of the first PV combiner box is the same as that of the The cosine similarity of the illumination sequence of the PV combiner boxes; represents a second preset threshold;

[0028] Determine the maximum value of the environmental difference of all different PV combiner boxes and minimum value , the interval Divide it into 50 sub-intervals, assign the environmental difference value within the sub-interval range as the value of the left endpoint of the sub-interval, and complete the quantification of the environmental difference value of the photovoltaic combiner box;

[0029] According to the difference between the total currents of different photovoltaic combiner boxes, the difference between the temperature sequences, the difference between the illumination sequences, and the current characteristic vector of the photovoltaic combiner box corresponding to the current collection moment, the state judgment value of the photovoltaic combiner box corresponding to the current collection moment is obtained.

[0030] Furthermore, according to the difference between the total currents of different photovoltaic combiner boxes, the difference between the temperature sequences, the difference between the illumination sequences, and the current characteristic vector of the photovoltaic combiner box corresponding to the current collection time, the state judgment value of the photovoltaic combiner box corresponding to the current collection time is obtained, and the specific process of obtaining is:

[0031] The absolute value of the difference between the total currents of two different photovoltaic combiner boxes is recorded as the current difference of the two different photovoltaic combiner boxes;

[0032] The cumulative sum of the absolute values ​​of the differences between the corresponding data in the temperature sequences of two different photovoltaic combiner boxes is recorded as the temperature difference of the two different photovoltaic combiner boxes, the cumulative sum of the absolute values ​​of the differences between the corresponding data in the illumination sequences of the two different photovoltaic combiner boxes is recorded as the illumination difference of the two different photovoltaic combiner boxes, and the sum of the temperature difference and the illumination difference of the two different photovoltaic combiner boxes is recorded as the environmental difference of the two different photovoltaic combiner boxes;

[0033] The mutual information between the current difference and the environmental difference of all different photovoltaic combiner boxes of the same decision tree is recorded as the decision coefficient of the same decision tree; the product of the decision coefficient of the decision tree and the decision value of the decision tree is recorded as the adjusted decision value of the decision tree; the cumulative sum of the adjusted decision values ​​of all decision trees is recorded as the state judgment value of the photovoltaic combiner box corresponding to the current collection moment.

[0034] Furthermore, the state judgment value of the photovoltaic combiner box corresponding to the current acquisition time is used to realize the remote state intelligent perception of the photovoltaic combiner box. The specific process of obtaining is:

[0035] When the state determination value of the photovoltaic combiner box corresponding to the current collection time is a negative number, the operation state of the photovoltaic combiner box at the current collection time is determined to be a fault state;

[0036] When the state determination value of the photovoltaic combiner box corresponding to the current collection time is a non-negative number, the operation state of the photovoltaic combiner box at the current collection time is determined to be a normal operation state.

[0037] In the second aspect, an embodiment of the present application provides a remote state intelligent sensing device for a photovoltaic combiner box, and the remote state intelligent sensing device includes: a data acquisition module, a current proportion acquisition module, a current fluctuation characteristic value acquisition module, and a remote state sensing module.

[0038] A data acquisition module is used to collect the current of the photovoltaic cell array, the temperature of the photovoltaic combiner box, and the illumination of the photovoltaic combiner box at the current acquisition moment and a preset number of acquisition moments before the current acquisition moment, and obtain the current sequence of the photovoltaic cell array, the temperature sequence of the photovoltaic combiner box, and the illumination sequence of the photovoltaic combiner box at the current sampling moment;

[0039] The current proportion acquisition module is used to determine the current proportion of each photovoltaic cell array at the current sampling moment according to the current sequence of all photovoltaic cell arrays connected in series with the photovoltaic combiner box at the current sampling moment;

[0040] A current fluctuation characteristic value acquisition module is used to divide the current sequence of the photovoltaic cell array into a first preset threshold number of subsequences, obtain normalized subsequences, determine the current fluctuation ratio of the subsequences according to the difference between the currents contained in the subsequences and the difference between the currents contained in the current sequence from which the subsequences are divided, and determine the current fluctuation characteristic value of the photovoltaic cell array in combination with the correlation between the normalized subsequences of all photovoltaic cell arrays connected in series with the photovoltaic combiner box corresponding to the same time period;

[0041] The remote state perception module is used to determine the current characteristic vector of the photovoltaic combiner box at the current collection moment according to the current proportion, current fluctuation characteristic value, temperature sequence, illumination sequence and current of all photovoltaic cell arrays connected in series with the photovoltaic combiner box at the current collection moment, determine the current characteristic vectors of the photovoltaic combiner box corresponding to a preset number of historical moments, obtain the trained random forest, input the current characteristic vector of the photovoltaic combiner box corresponding to the current collection moment into the trained random forest, obtain the decision value of the decision tree, and obtain all photovoltaic combiner boxes in the same decision tree, combine the difference between the temperature sequences of two different photovoltaic combiner boxes, the difference and similarity between the illumination sequences, and the difference between the currents of all photovoltaic cell arrays connected in series with the photovoltaic combiner box at the current collection moment, obtain the state judgment value of the photovoltaic combiner box corresponding to the current collection moment, and realize remote state intelligent perception of the photovoltaic combiner box according to the state judgment value of the photovoltaic combiner box corresponding to the current collection moment.

[0042] In the third aspect, an embodiment of the present application also provides a remote status intelligent perception system for a photovoltaic combiner box, the system comprising a memory, a processor, and a computer program stored in the memory and running on the processor, and the processor implements the steps of any one of the above methods when executing the computer program.

[0043] It can be seen from the above embodiments that the method, device and system for intelligently sensing the remote state of a photovoltaic combiner box provided in the embodiments of the present application have at least the following beneficial effects:

[0044] The present application evaluates the degree of current fluctuation of the photovoltaic cell array based on the characteristic that the current of the photovoltaic cell array decreases when a short circuit, open circuit or aging lamp problem occurs in the photovoltaic cell array in the photovoltaic system, obtains the current proportion of each photovoltaic cell array at the current sampling moment, divides the current sequence into subsequences and evaluates the correlation between the subsequences and the difference between the currents contained in the subsequences based on the characteristic that the current fluctuation lasts for a short period of time when the photovoltaic cell array in the photovoltaic system initially fails, and obtains the current fluctuation characteristic value of the photovoltaic cell array; determines the current characteristic vector of the photovoltaic combiner box at the current collection moment and a preset number of historical moments based on the current proportion of all photovoltaic cell arrays connected in series with the photovoltaic combiner box at the current collection moment, the current fluctuation characteristic value, the temperature sequence, the illumination sequence and the current; further, based on The random forest is trained according to the current feature vector of the photovoltaic combiner box at the historical moment to obtain the trained random forest. In order to avoid the problem that the trained random forest is affected by the sample imbalance corresponding to the operating state of the photovoltaic combiner box in the data set during the training process, the difference between the temperature series of two different photovoltaic combiner boxes, the difference and similarity between the illumination series, and the difference between the currents of all photovoltaic cell arrays connected in series with the photovoltaic combiner box at the current collection moment are combined to obtain the state judgment value of the photovoltaic combiner box corresponding to the current collection moment. Finally, according to the state judgment value of the photovoltaic combiner box corresponding to the current collection moment, the remote state intelligent perception of the photovoltaic combiner box is realized to solve the problem of missed or false alarms in abnormal detection of photovoltaic system caused by complex current changes in photovoltaic cell arrays due to changes in ambient temperature. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] In order to more clearly illustrate the technical solutions and advantages in the embodiments of the present application or the prior art, the drawings required for use in the embodiments or the prior art descriptions are briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0046] Figure 1 A flowchart of a method for intelligently sensing the remote state of a photovoltaic combiner box provided in one embodiment of the present application;

[0047] Figure 2 A schematic diagram of the structure of a remote state intelligent sensing device for a photovoltaic combiner box provided in one embodiment of the present application. DETAILED DESCRIPTION

[0048] In order to further explain the technical means and effects adopted by this application to achieve the predetermined invention purpose, the following is a detailed description of the remote state intelligent sensing method, device and system of a photovoltaic combiner box proposed in this application, its specific implementation method, structure, characteristics and effects in combination with the accompanying drawings and preferred embodiments. In the following description, different "one embodiment" or "another embodiment" does not necessarily refer to the same embodiment. In addition, specific features, structures or characteristics in one or more embodiments may be combined in any suitable form.

[0049] Unless otherwise specified and limited, terms such as "comprises", "comprising" or any other variants thereof are intended to cover non-exclusive inclusion, so that a circuit structure, article or device including a series of elements includes not only those elements, but also includes other elements not explicitly listed, or also includes elements inherent to such articles or devices. In the absence of further restrictions, an element defined by the sentence "comprising a ..." does not exclude the presence of other identical elements in the article or device including the element. In addition, the term "and\or" used herein includes any and all combinations of one or more related listed items. All technical and scientific terms used herein have the same meaning as those commonly understood by technicians in the technical field of this application.

[0050] The following is a detailed description of a method, device and system for intelligently sensing the remote state of a photovoltaic combiner box provided by the present application in conjunction with the accompanying drawings.

[0051] See also Figure 1 , which shows a flowchart of a method for remote state intelligent perception of a photovoltaic combiner box provided by an embodiment of the present application, the method comprising the following steps:

[0052] S001: Collect the current of the photovoltaic cell array, the temperature of the photovoltaic combiner box, and the illumination of the photovoltaic combiner box at the current collection moment and a preset number of collection moments before the current collection moment, and obtain the current sequence of the photovoltaic cell array, the temperature sequence of the photovoltaic combiner box, and the illumination sequence of the photovoltaic combiner box at the current sampling moment.

[0053] The photovoltaic system of this embodiment includes a total of 200 photovoltaic combiner boxes, each of which is connected in series with 16 photovoltaic cell arrays, and all photovoltaic cell arrays connected in series with the photovoltaic combiner boxes are connected in parallel. Each photovoltaic cell array includes 20 230W photovoltaic modules.

[0054] Among them, in the actual application process, as other implementation methods, the implementer can set the number of photovoltaic junction boxes included in the photovoltaic system, the number of photovoltaic cell arrays connected in series to each photovoltaic junction box, and the number of photovoltaic modules included in the photovoltaic cell array according to actual conditions.

[0055] For each PV combiner box, the suspended wire Hall open-loop current sensor installed in the PV combiner box is used to collect the current of each PV cell array connected in series with the PV combiner box. The temperature sensor installed in the PV combiner box is used to collect the temperature of the PV combiner box, and the illumination sensor installed in the PV combiner box is used to collect the illumination of the PV combiner box.

[0056] Arrange the currents of the 599 photovoltaic cell arrays adjacent to the current sampling moment and before in the order of the collection moments to obtain the current sequence of the photovoltaic cell array at the current sampling moment. Arrange the temperatures of the 59 photovoltaic combiner boxes adjacent to the current sampling moment and before in the order of the collection moments to obtain the temperature sequence of the photovoltaic combiner box at the current sampling moment; arrange the illuminations of the 59 photovoltaic combiner boxes adjacent to the current sampling moment and before in the order of the collection moments to obtain the illumination sequence of the photovoltaic combiner box at the current sampling moment.

[0057] Preferably, in one embodiment of the present application, when collecting current data, temperature data and illumination data, the sampling frequency of current data in this embodiment is 10 times per second, the sampling frequency of temperature data and illumination data is 1 time per second, the current sequence of the photovoltaic cell array contains 600 currents, the temperature sequence of the photovoltaic combiner box contains 60 temperatures, and the illumination sequence of the photovoltaic combiner box contains 60 illuminations. In actual application, as other implementation methods, the implementer can decide the sampling frequency of current, the number of currents contained in the current sequence, the number of temperatures contained in the temperature sequence of the photovoltaic combiner box, and the number of illuminations contained in the illumination sequence of the photovoltaic combiner box according to actual conditions, and this application does not impose any special restrictions.

[0058] The collected data is transmitted to the data cloud platform using power line carrier communication technology to reduce the installation cost of the signal transmission equipment, wherein the collected data includes the current sequence of the photovoltaic cell array at the current sampling moment, the temperature sequence of the photovoltaic combiner box, and the illumination sequence of the photovoltaic combiner box. The use of power line carrier communication technology to transmit the collected data to the data cloud platform is a well-known technology and will not be described in detail. Specifically, the collected data is modulated using a photovoltaic combiner box and coupled to the power line, and the collected data is transmitted to the inverter. The concentrator set on the inverter obtains the carrier signal and the collected data is uploaded to the data cloud platform using the MODBUS protocol.

[0059] At this point, the current sequence of the photovoltaic cell array, the temperature sequence of the photovoltaic combiner box, and the illumination sequence of the photovoltaic combiner box at the current sampling moment are obtained.

[0060] S002: According to the current sequence of all photovoltaic cell arrays connected in series to the photovoltaic combiner box at the current sampling moment, the current proportion of each photovoltaic cell array at the current sampling moment is determined respectively.

[0061] When the photovoltaic system is operating normally, the current of the photovoltaic cell array connected in series with the photovoltaic combiner box is stable, so the ratio between the current of each photovoltaic cell array and the current passing through the photovoltaic combiner box is relatively stable. When the temperature and illumination in the environment change, the ratio between the current of each photovoltaic cell array and the current passing through the photovoltaic combiner box is still relatively stable. Therefore, when the photovoltaic system is operating normally, the current of the photovoltaic cell array connected in series with the photovoltaic combiner box is still relatively stable, and the ratio between the current of each photovoltaic cell array and the current passing through the photovoltaic combiner box is also relatively stable.

[0062] When a short circuit, open circuit or aging problem occurs in the photovoltaic cell array in the photovoltaic system, the current of the photovoltaic cell array will show a significant decrease. When the photovoltaic modules in the photovoltaic system are blocked by foreign objects, the power generation of the photovoltaic modules will be reduced, and the current of the photovoltaic cell array where the photovoltaic modules are located will also decrease. Foreign objects will also make the blocked photovoltaic modules become loads to generate heat, resulting in a hot spot effect. Therefore, when an abnormality occurs in the photovoltaic system, the current of the photovoltaic cell array connected in series with the photovoltaic combiner box will fluctuate significantly, and the ratio between the current of each photovoltaic cell array and the current passing through the photovoltaic combiner box will also fluctuate accordingly.

[0063] According to the current sequence of all photovoltaic cell arrays connected in series to the photovoltaic combiner box at the current sampling moment, the current proportion of each photovoltaic cell array at the current sampling moment is determined respectively.

[0064] The average value of all currents contained in the current sequence of the photovoltaic cell array at the current sampling moment is recorded as the current average value of the photovoltaic cell array at the current sampling moment; the cumulative sum of the current average values ​​of all photovoltaic cell arrays connected in series with the photovoltaic combiner boxes at the current sampling moment is recorded as the total current average value of all photovoltaic cell arrays connected in series with the photovoltaic combiner boxes; the ratio of the current average value of the photovoltaic cell array to the total current average value is recorded as the current proportion of the photovoltaic cell array at the current sampling moment.

[0065] The current proportion of the photovoltaic cell array reflects the proportion of the current value of the photovoltaic cell array in the current passing through the photovoltaic combiner box at the current sampling moment.

[0066] At this point, the current proportion of each photovoltaic cell array at the current sampling moment is obtained.

[0067] S003: Divide the current sequence of the photovoltaic cell array into a first preset threshold number of subsequences, obtain normalized subsequences, determine the current fluctuation ratio of the subsequences according to the difference between the currents contained in the subsequences and the difference between the currents contained in the current sequence from which the subsequences are divided, and determine the current fluctuation characteristic value of the photovoltaic cell array in combination with the correlation between the normalized subsequences of all photovoltaic cell arrays connected in series with the photovoltaic combiner box corresponding to the same time period.

[0068] When the photovoltaic cell array in the photovoltaic system begins to fail, the current of the photovoltaic cell array changes little. In order to more sensitively identify the failure of the photovoltaic combiner box, further analysis is performed on the features with a greater risk of failure to improve the sensitivity of identifying abnormal conditions of the photovoltaic combiner box.

[0069] When the photovoltaic cell array in the photovoltaic system initially fails, the current will fluctuate and the duration of the current fluctuation is short. For each current sequence of the photovoltaic cell array, the current sequence is divided into M subsequences, and the length of each subsequence divided by the current sequence is the same. The number of data contained in the subsequence is recorded as N.

[0070] Wherein, M represents the first preset threshold. In this embodiment, the value of the first preset threshold is 20, and the value of N is 30.

[0071] The standard deviation of all currents contained in the current sequence is recorded as the total current fluctuation of all subsequences obtained by evenly dividing the current sequence, and the ratio of the standard deviation of all currents contained in the subsequence to the total current fluctuation is recorded as the current fluctuation ratio of the subsequence.

[0072] In the process of calculating the ratio, in order to avoid the situation where the denominator is zero, a preset value needs to be added to the denominator, and an embodiment of the preset value is 2.

[0073] When the current fluctuation ratio of the subsequence is larger, the fluctuation degree of all currents contained in the subsequence is larger relative to the fluctuation degree of all currents contained in the current sequence, and at the acquisition moment corresponding to the subsequence, the possibility of failure of the photovoltaic cell array corresponding to the current contained in the subsequence is greater.

[0074] There are slight differences in the electrical parameters of photovoltaic modules in different photovoltaic cell arrays, and the currents generated under the same temperature and illumination will also be different. All currents contained in each subsequence are normalized respectively, and the normalized currents are arranged according to the order of the currents in the subsequence to obtain the normalized subsequence.

[0075] The normalized subsequence is a sequence that only extracts and obtains the trend characteristics of the current changes in the subsequence, avoiding the differences in electrical parameters of photovoltaic modules in different photovoltaic cell arrays, as well as the influence of temperature and illumination differences in the environment on the abnormal identification of photovoltaic combiner boxes. It should be noted that this embodiment uses the Z-Score standard normalization method to calculate the normalized value. In actual application, the implementer can use other existing methods such as maximum and minimum value normalization method, sigmoid function, etc. to calculate the normalized value, which is not limited here.

[0076] According to the correlation between the normalized subsequences corresponding to the same time period of all photovoltaic cell arrays connected in series with the photovoltaic combiner box and the current fluctuation ratio of the subsequences, the current fluctuation characteristic value of the photovoltaic cell array is determined.

[0077]

[0078] in, Indicates Current fluctuation characteristic value of a photovoltaic cell array; Indicates The number of other photovoltaic cell arrays connected in parallel to the same photovoltaic combiner box; Indicates the current collection time. The current sequence of the photovoltaic cell array is divided into The current fluctuation ratio of each subsequence; Indicates the first preset threshold, which is the value of the current acquisition time. The number of subsequences divided from the current sequence of the photovoltaic cell array, in this embodiment, the value of the first preset threshold is 20; Indicates the current collection time. The current sequence of the photovoltaic cell array is divided into The normalized subsequence is the same as the current acquisition time. The current sequence of the photovoltaic cell array is divided into The Nth value in the cross-correlation sequence of the normalized subsequences, where N represents the amount of data contained in the normalized subsequence.

[0079] Calculating the mutual correlation sequence of two different subsequences is a well-known technique and will not be described in detail. The Nth value in the mutual correlation sequence is used to measure the correlation between the two different subsequences.

[0080] Used to measure the current collection time The fluctuation of the current sequence of a photovoltaic cell array, when the The greater the fluctuation of the current sequence of the photovoltaic array, the The smaller the similarity between the current sequence of the first photovoltaic cell array and the current sequence of other photovoltaic cell arrays, the The greater the risk of a minor fault in the photovoltaic array at the current collection time, the greater the risk of a minor fault in the photovoltaic array at the current collection time. The larger the current fluctuation characteristic value of the photovoltaic cell array.

[0081] At this point, the current fluctuation characteristic values ​​of all photovoltaic cell arrays are obtained.

[0082] S004: Determine the current characteristic vector of the photovoltaic combiner box at the current collection moment according to the current proportion, current fluctuation characteristic value, temperature sequence, illumination sequence and current of all photovoltaic cell arrays connected in series with the photovoltaic combiner box at the current collection moment, determine the current characteristic vectors of the photovoltaic combiner box corresponding to a preset number of historical moments, obtain the trained random forest, input the current characteristic vector of the photovoltaic combiner box corresponding to the current collection moment into the trained random forest, obtain the decision value of the decision tree, and obtain all photovoltaic combiner boxes in the same decision tree, combine the difference between the temperature sequences of two different photovoltaic combiner boxes, the difference and similarity between the illumination sequences, and the difference between the currents of all photovoltaic cell arrays connected in series with the photovoltaic combiner box at the current collection moment, obtain the state judgment value of the photovoltaic combiner box corresponding to the current collection moment, and realize remote state intelligent perception of the photovoltaic combiner box according to the state judgment value of the photovoltaic combiner box corresponding to the current collection moment.

[0083] According to the current proportion, current fluctuation characteristic value, temperature sequence, illumination sequence and current of all photovoltaic cell arrays connected in series with the photovoltaic combiner box at the current acquisition time, the current characteristic vector of the photovoltaic combiner box is determined. The current characteristic vector of the photovoltaic combiner box corresponding to the current acquisition time is:

[0084]

[0085] in, They represent the first to the The current proportion of each photovoltaic cell array, represents the number of all photovoltaic cell arrays connected in series with the photovoltaic combiner box. The value of is 16; They represent the first to the Current fluctuation characteristic value of a photovoltaic cell array; It represents the cumulative sum of the currents of all photovoltaic cell arrays connected in series with the photovoltaic combiner box at the current collection moment; Indicates the average value of all temperatures contained in the temperature series of the PV combiner box at the current acquisition moment; Indicates the average value of all illuminances contained in the illuminance sequence of the PV combiner box at the current acquisition moment.

[0086] According to the same method, the current characteristic vectors of the photovoltaic combiner box corresponding to 20,000 historical moments in the historical moments are obtained, and the operating status of the photovoltaic combiner box at these 20,000 historical moments is obtained. The operating status of the photovoltaic combiner box includes two types: normal operating status and fault status. Among them, the number of historical moments 20,000 is recorded as the preset number.

[0087] The number of decision trees is set to 200, the maximum number of features is set to 4, the maximum depth of the decision tree is set to 10, and the minimum number of samples is set to 5. The random forest algorithm is used to classify the current feature vectors of the photovoltaic combiner boxes corresponding to all historical moments. The decision value output of each decision tree is "-1" or "1", where "-1" indicates that the operation status of the photovoltaic combiner box is normal operation, and "1" indicates that the operation status of the photovoltaic combiner box is faulty. The training of the random forest is completed. The random forest algorithm uses the Bootstrap sampling method to generate 200 training sets to improve the accuracy of the algorithm's remote state perception of the photovoltaic combiner box. Each training set contains the current feature vectors of the photovoltaic combiner box corresponding to 1000 historical moments and the operation status of the photovoltaic combiner box at 1000 historical moments.

[0088] At this point, the trained random forest is obtained.

[0089] In the process of training the random forest, considering that there are few PV combiner boxes in fault state in the historical moment, the correlation between temperature, illumination and the total current of the PV combiner box is used to reduce the impact of the sample imbalance corresponding to the operating state of the PV combiner box in the data set on the obtained trained random forest. Among them, the total current of the PV combiner box is the cumulative sum of the currents of all photovoltaic cell arrays connected in series with the PV combiner box at the current acquisition moment.

[0090] According to the difference between the temperature sequences of two different photovoltaic combiner boxes and the similarity between the illumination sequences, the environment difference values ​​of the two different photovoltaic combiner boxes are determined.

[0091]

[0092] in, Indicates The first PV combiner box and the Environmental difference value of each PV combiner box; Indicates The mean of all temperatures contained in the temperature series of the first PV combiner box is The absolute value of the difference between the mean values ​​of all temperatures contained in the temperature series of the PV combiner boxes; Indicates The average of all illuminances contained in the illuminance sequence of the first PV combiner box is The absolute value of the difference between the mean values ​​of all illuminances contained in the illuminance sequence of the PV combiner boxes; Indicates The temperature series of the first PV combiner box and the The cosine similarity of the temperature series of the PV combiner boxes; Indicates The illumination sequence of the first PV combiner box is the same as that of the The cosine similarity of the illumination sequence of the PV combiner boxes; represents the second preset threshold. The function of the second preset threshold is to ensure that the denominator is a positive number. In this embodiment, the value of the second preset threshold is 2.

[0093] Calculating the cosine similarity between two sequences is a well-known technique and will not be described in detail.

[0094] Quantify the environmental difference values ​​of all different photovoltaic combiner boxes. The specific process of quantification is: determine the maximum value of the environmental difference values ​​of all different photovoltaic combiner boxes and minimum value , the maximum value and minimum value The interval formed The environmental difference value is evenly divided into U sub-intervals, and the environmental difference value within the sub-interval is assigned as the value of the left endpoint of the sub-interval to complete the quantification of the environmental difference value of the photovoltaic combiner box.

[0095] In this embodiment, the value of U is 50.

[0096] For example, when the maximum and minimum values ​​of the environmental difference values ​​of all different PV combiner boxes are 51 and 1 respectively, the maximum and minimum value The interval formed is ,Will The 50 sub-intervals divided equally are , , ,……, , , when the environmental difference value is 1.5, the environmental difference value is In the sub-interval range, the environmental difference value is assigned to The left endpoint of is 1.

[0097] When the PV combiner box is in normal operation, the total current of the PV combiner box is greatly affected by the temperature and illumination in the environment where the PV combiner box is located. Therefore, the difference between the total currents of two different PV combiner boxes has a strong correlation with the difference between the environments where the two different PV combiner boxes are located. However, the total current of the PV combiner box in a fault state will be affected by the fault, and the difference between the total current of the PV combiner box in a fault state and the total current of the PV combiner box in a normal operation state no longer shows an obvious correlation with the difference between the environments where the PV combiner boxes are located.

[0098] The absolute value of the difference between the total currents of two different photovoltaic combiner boxes is recorded as the current difference of the two different photovoltaic combiner boxes; the cumulative sum of the absolute values ​​of the differences between the corresponding data in the temperature sequences of the two different photovoltaic combiner boxes is recorded as the temperature difference of the two different photovoltaic combiner boxes; the cumulative sum of the absolute values ​​of the differences between the corresponding data in the illumination sequences of the two different photovoltaic combiner boxes is recorded as the illumination difference of the two different photovoltaic combiner boxes; and the sum of the temperature difference and the illumination difference of the two different photovoltaic combiner boxes is recorded as the environmental difference of the two different photovoltaic combiner boxes.

[0099] The current feature vector of the photovoltaic combiner box corresponding to the current collection moment is input into the trained random forest to obtain the decision value of the decision tree, and all photovoltaic combiner boxes of the same decision tree are obtained; the mutual information of the current difference and the environmental difference of all different photovoltaic combiner boxes of the same decision tree is recorded as the decision coefficient of the same decision tree; the product of the decision coefficient of the decision tree and the decision value of the decision tree is recorded as the adjusted decision value of the decision tree, and the cumulative sum of the adjusted decision values ​​of all decision trees is recorded as the state judgment value of the photovoltaic combiner box corresponding to the current collection moment.

[0100] Among them, mutual information is used to quantify the degree of mutual dependence between two random variables. Calculating the mutual information of current differences and environmental differences of all different photovoltaic combiner boxes is a well-known technology and will not be repeated here. Using the decision coefficient of the decision tree as the coefficient of the decision value of the decision tree can solve the problem of the impact of sample imbalance corresponding to the operating status of the photovoltaic combiner box in the data set on the random forest training process. The current feature vector of the photovoltaic combiner box corresponding to the current collection moment is input into the random forest to obtain the decision value of the decision tree. Using the random forest to obtain the decision value of the decision tree is a well-known technology and will not be repeated here.

[0101] It can be understood that, at this point, the state judgment value of the photovoltaic combiner box corresponding to the current acquisition moment is obtained, and the state judgment value can solve the problem of the influence of the sample imbalance corresponding to the operating state of the photovoltaic combiner box in the data set on the random forest training process.

[0102] When the state judgment value of the photovoltaic combiner box corresponding to the current collection time is a negative number, the operating state of the photovoltaic combiner box at the current collection time is judged to be a fault state; when the state judgment value of the photovoltaic combiner box corresponding to the current collection time is a non-negative number, the operating state of the photovoltaic combiner box at the current collection time is judged to be a normal operating state.

[0103] At this point, the operating status of the PV combiner box is obtained, and the remote status intelligent perception of the PV combiner box is realized.

[0104] See also Figure 2 , Figure 2 1 is a schematic diagram of the structure of a remote state intelligent sensing device for a photovoltaic combiner box provided by an embodiment of the present application. In this embodiment, the device includes various units for executing various steps in an embodiment corresponding to a remote state intelligent sensing method for a photovoltaic combiner box. Figure 2 The remote state intelligent perception device includes: a data acquisition module, a current proportion acquisition module, a current fluctuation characteristic value acquisition module, and a remote state perception module.

[0105] A data acquisition module is used to collect the current of the photovoltaic cell array, the temperature of the photovoltaic combiner box, and the illumination of the photovoltaic combiner box at the current acquisition moment and a preset number of acquisition moments before the current acquisition moment, and obtain the current sequence of the photovoltaic cell array, the temperature sequence of the photovoltaic combiner box, and the illumination sequence of the photovoltaic combiner box at the current sampling moment;

[0106] The current proportion acquisition module is used to determine the current proportion of each photovoltaic cell array at the current sampling moment according to the current sequence of all photovoltaic cell arrays connected in series with the photovoltaic combiner box at the current sampling moment;

[0107] A current fluctuation characteristic value acquisition module is used to divide the current sequence of the photovoltaic cell array into a first preset threshold number of subsequences, obtain normalized subsequences, determine the current fluctuation ratio of the subsequences according to the difference between the currents contained in the subsequences and the difference between the currents contained in the current sequence from which the subsequences are divided, and determine the current fluctuation characteristic value of the photovoltaic cell array in combination with the correlation between the normalized subsequences of all photovoltaic cell arrays connected in series with the photovoltaic combiner box corresponding to the same time period;

[0108] The remote state perception module is used to determine the current characteristic vector of the photovoltaic combiner box at the current collection moment according to the current proportion, current fluctuation characteristic value, temperature sequence, illumination sequence and current of all photovoltaic cell arrays connected in series with the photovoltaic combiner box at the current collection moment, determine the current characteristic vectors of the photovoltaic combiner box corresponding to a preset number of historical moments, obtain the trained random forest, input the current characteristic vector of the photovoltaic combiner box corresponding to the current collection moment into the trained random forest, obtain the decision value of the decision tree, and obtain all photovoltaic combiner boxes in the same decision tree, combine the difference between the temperature sequences of two different photovoltaic combiner boxes, the difference and similarity between the illumination sequences, and the difference between the currents of all photovoltaic cell arrays connected in series with the photovoltaic combiner box at the current collection moment, obtain the state judgment value of the photovoltaic combiner box corresponding to the current collection moment, and realize remote state intelligent perception of the photovoltaic combiner box according to the state judgment value of the photovoltaic combiner box corresponding to the current collection moment.

[0109] Based on the same inventive concept as the above method, an embodiment of the present application also provides a remote state intelligent perception system for a photovoltaic combiner box, comprising a memory, a processor, and a computer program stored in the memory and running on the processor, wherein the processor implements the steps of any one of the above-mentioned methods for remote state intelligent perception of a photovoltaic combiner box when executing the computer program.

[0110] It is to be understood that the sequence of the embodiments of the present application described above is for description only and does not represent the advantages and disadvantages of the embodiments. The above describes specific embodiments of the present specification. In addition, the processes depicted in the accompanying drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0111] The various embodiments in this specification are described in a progressive manner, and the same or similar parts between the various embodiments can be referenced to each other, and each embodiment focuses on the differences from other embodiments.

[0112] The above content is only an implementation method of the present application and is not intended to limit the scope of the present application. Any equivalent structure or equivalent process transformation made using the contents of the present application specification and drawings, or directly or indirectly used in other related technical fields, are also included in the protection scope of the present application.

Claims

1. A remote state intelligent sensing method for a photovoltaic combiner box, characterized in that: The method comprises the following steps: Collect the current of the photovoltaic cell array, the temperature of the photovoltaic combiner box, and the illumination of the photovoltaic combiner box at the current collection time and a preset number of collection times before the current collection time, and obtain the current sequence of the photovoltaic cell array, the temperature sequence of the photovoltaic combiner box, and the illumination sequence of the photovoltaic combiner box at the current sampling time; According to the current sequence of all photovoltaic cell arrays connected in series to the photovoltaic combiner box at the current sampling moment, the current proportion of each photovoltaic cell array at the current sampling moment is determined respectively; The current sequence of the photovoltaic cell array is divided into a first preset threshold number of subsequences, and a normalized subsequence is obtained. The current fluctuation ratio of the subsequence is determined according to the difference between the currents contained in the subsequences and the difference between the currents contained in the current sequence from which the subsequences are divided, and the current fluctuation characteristic value of the photovoltaic cell array is determined by combining the correlation between the normalized subsequences of all photovoltaic cell arrays connected in series with the photovoltaic combiner box corresponding to the same time period; According to the current proportion, current fluctuation characteristic value, temperature sequence, illumination sequence and current of all photovoltaic cell arrays connected in series with the photovoltaic combiner box at the current collection moment, the current characteristic vector of the photovoltaic combiner box at the current collection moment is determined, the current characteristic vectors of the photovoltaic combiner box corresponding to a preset number of historical moments are determined, and the trained random forest is obtained. The current characteristic vector of the photovoltaic combiner box corresponding to the current collection moment is input into the trained random forest to obtain the decision value of the decision tree, and all photovoltaic combiner boxes in the same decision tree are obtained. Combined with the difference between the temperature sequences of two different photovoltaic combiner boxes, the difference and similarity between the illumination sequences, and the difference between the currents of all photovoltaic cell arrays connected in series with the photovoltaic combiner box at the current collection moment, the state judgment value of the photovoltaic combiner box corresponding to the current collection moment is obtained. According to the state judgment value of the photovoltaic combiner box corresponding to the current collection moment, remote state intelligent perception of the photovoltaic combiner box is realized.

2. A method for remote intelligent sensing of a photovoltaic combiner box according to claim 1, characterized in that: The process of determining the current proportion of each photovoltaic cell array at the current sampling moment is as follows: The average value of all currents contained in the current sequence of the photovoltaic cell array at the current sampling moment is recorded as the current average value of the photovoltaic cell array at the current sampling moment; The cumulative sum of the current averages of all photovoltaic cell arrays connected in series to the photovoltaic combiner box at the current sampling moment is recorded as the total current average of all photovoltaic cell arrays connected in series to the photovoltaic combiner box; The ratio of the current average of the photovoltaic cell array to the total current average is recorded as the current proportion of the photovoltaic cell array at the current sampling moment.

3. A method for remote intelligent sensing of the status of a photovoltaic combiner box according to claim 1, characterized in that: The process of determining the current fluctuation ratio of the subsequence is as follows: The standard deviation of all currents contained in the current sequence is recorded as the total current fluctuation of all subsequences obtained by evenly dividing the current sequence; The ratio of the standard deviation of all currents contained in the subsequence to the total current fluctuation is recorded as the current fluctuation ratio of the subsequence.

4. A method for remote intelligent sensing of a photovoltaic combiner box according to claim 1, characterized in that: The specific process of obtaining the current fluctuation characteristic value of the photovoltaic cell array is as follows: in, Indicates Current fluctuation characteristic value of a photovoltaic cell array; Indicates The number of other photovoltaic cell arrays connected in parallel to the same photovoltaic combiner box; Indicates the current collection time. The current sequence of the photovoltaic cell array is divided into The current fluctuation ratio of each subsequence; represents a first preset threshold; Indicates the current collection time. The current sequence of the photovoltaic cell array is divided into The normalized subsequence is the same as the current acquisition time. The current sequence of the photovoltaic cell array is divided into The Nth value in the cross-correlation sequence of the normalized subsequences, where N represents the amount of data contained in the normalized subsequence.

5. A method for remote intelligent sensing of a photovoltaic combiner box according to claim 1, characterized in that: The current characteristic vector of the photovoltaic combiner box at the current collection moment is: in, They represent the first to the The current proportion of each photovoltaic cell array, Represents the number of all photovoltaic cell arrays connected in series with the photovoltaic combiner box; They represent the first to the Current fluctuation characteristic value of a photovoltaic cell array; It represents the cumulative sum of the currents of all photovoltaic cell arrays connected in series with the photovoltaic combiner box at the current collection moment; Indicates the average value of all temperatures contained in the temperature series of the PV combiner box at the current acquisition moment; Indicates the average value of all illuminances contained in the illuminance sequence of the PV combiner box at the current acquisition moment.

6. A method for remote intelligent sensing of a photovoltaic combiner box according to claim 1, characterized in that: The method of combining the difference between the temperature sequences of two different photovoltaic combiner boxes, the difference and similarity between the illumination sequences, and the difference between the currents of all photovoltaic cell arrays connected in series with the photovoltaic combiner box at the current collection time to obtain the state determination value of the photovoltaic combiner box corresponding to the current collection time includes the following specific methods: The accumulated current of all photovoltaic cell arrays connected in series with the photovoltaic combiner box at the current collection time is recorded as the total current of the photovoltaic combiner box; According to the difference between the temperature sequences of two different photovoltaic combiner boxes and the similarity between the illumination sequences, the environmental difference values ​​of the two different photovoltaic combiner boxes are determined, and the calculation formula is: in, Indicates The first PV combiner box and the Environmental difference value of each PV combiner box; Indicates The mean of all temperatures contained in the temperature series of the first PV combiner box is The absolute value of the difference between the mean values ​​of all temperatures contained in the temperature series of the PV combiner boxes; Indicates The average of all illuminances contained in the illuminance sequence of the first PV combiner box is The absolute value of the difference between the mean values ​​of all illuminances contained in the illuminance sequence of the PV combiner boxes; Indicates The temperature series of the first PV combiner box and the The cosine similarity of the temperature series of the PV combiner boxes; Indicates The illumination sequence of the first PV combiner box is the same as that of the The cosine similarity of the illumination sequence of the PV combiner boxes; represents a second preset threshold; Determine the maximum value of the environmental difference of all different PV combiner boxes and minimum value , the interval Divide it into 50 sub-intervals, assign the environmental difference value within the sub-interval range as the value of the left endpoint of the sub-interval, and complete the quantification of the environmental difference value of the photovoltaic combiner box; According to the difference between the total currents of different photovoltaic combiner boxes, the difference between the temperature sequences, the difference between the illumination sequences, and the current characteristic vector of the photovoltaic combiner box corresponding to the current collection moment, the state judgment value of the photovoltaic combiner box corresponding to the current collection moment is obtained.

7. A method for remote intelligent sensing of a photovoltaic combiner box according to claim 6, characterized in that: According to the difference between the total currents of different photovoltaic combiner boxes, the difference between the temperature sequences, the difference between the illumination sequences, and the current characteristic vector of the photovoltaic combiner box corresponding to the current collection moment, the state judgment value of the photovoltaic combiner box corresponding to the current collection moment is obtained, and the specific process of obtaining is: The absolute value of the difference between the total currents of two different photovoltaic combiner boxes is recorded as the current difference of the two different photovoltaic combiner boxes; The cumulative sum of the absolute values ​​of the differences between the corresponding data in the temperature sequences of two different photovoltaic combiner boxes is recorded as the temperature difference of the two different photovoltaic combiner boxes, the cumulative sum of the absolute values ​​of the differences between the corresponding data in the illumination sequences of the two different photovoltaic combiner boxes is recorded as the illumination difference of the two different photovoltaic combiner boxes, and the sum of the temperature difference and the illumination difference of the two different photovoltaic combiner boxes is recorded as the environmental difference of the two different photovoltaic combiner boxes; The mutual information between the current difference and the environmental difference of all different photovoltaic combiner boxes of the same decision tree is recorded as the decision coefficient of the same decision tree; the product of the decision coefficient of the decision tree and the decision value of the decision tree is recorded as the adjusted decision value of the decision tree; the cumulative sum of the adjusted decision values ​​of all decision trees is recorded as the state judgment value of the photovoltaic combiner box corresponding to the current collection moment.

8. A method for remote intelligent sensing of a photovoltaic combiner box according to claim 1, characterized in that: According to the state judgment value of the photovoltaic combiner box corresponding to the current acquisition time, the remote state intelligent perception of the photovoltaic combiner box is realized, and the specific process of obtaining is: When the state determination value of the photovoltaic combiner box corresponding to the current collection time is a negative number, the operation state of the photovoltaic combiner box at the current collection time is determined to be a fault state; When the state determination value of the photovoltaic combiner box corresponding to the current collection time is a non-negative number, the operation state of the photovoltaic combiner box at the current collection time is determined to be a normal operation state.

9. A remote state intelligent sensing device for a photovoltaic combiner box, implementing the method as claimed in claim 1, characterized in that: The remote state intelligent sensing device comprises: A data acquisition module is used to collect the current of the photovoltaic cell array, the temperature of the photovoltaic combiner box, and the illumination of the photovoltaic combiner box at the current acquisition moment and a preset number of acquisition moments before the current acquisition moment, and obtain the current sequence of the photovoltaic cell array, the temperature sequence of the photovoltaic combiner box, and the illumination sequence of the photovoltaic combiner box at the current sampling moment; The current proportion acquisition module is used to determine the current proportion of each photovoltaic cell array at the current sampling moment according to the current sequence of all photovoltaic cell arrays connected in series with the photovoltaic combiner box at the current sampling moment; A current fluctuation characteristic value acquisition module is used to divide the current sequence of the photovoltaic cell array into a first preset threshold number of subsequences, obtain normalized subsequences, determine the current fluctuation ratio of the subsequences according to the difference between the currents contained in the subsequences and the difference between the currents contained in the current sequence from which the subsequences are divided, and determine the current fluctuation characteristic value of the photovoltaic cell array in combination with the correlation between the normalized subsequences of all photovoltaic cell arrays connected in series with the photovoltaic combiner box corresponding to the same time period; The remote state perception module is used to determine the current characteristic vector of the photovoltaic combiner box at the current collection moment according to the current proportion, current fluctuation characteristic value, temperature sequence, illumination sequence and current of all photovoltaic cell arrays connected in series with the photovoltaic combiner box at the current collection moment, determine the current characteristic vectors of the photovoltaic combiner box corresponding to a preset number of historical moments, obtain the trained random forest, input the current characteristic vector of the photovoltaic combiner box corresponding to the current collection moment into the trained random forest, obtain the decision value of the decision tree, and obtain all photovoltaic combiner boxes in the same decision tree, combine the difference between the temperature sequences of two different photovoltaic combiner boxes, the difference and similarity between the illumination sequences, and the difference between the currents of all photovoltaic cell arrays connected in series with the photovoltaic combiner box at the current collection moment, obtain the state judgment value of the photovoltaic combiner box corresponding to the current collection moment, and realize remote state intelligent perception of the photovoltaic combiner box according to the state judgment value of the photovoltaic combiner box corresponding to the current collection moment.

10. A remote state intelligent sensing system for a photovoltaic combiner box, comprising a memory, a processor, and a computer program stored in the memory and running on the processor, characterized in that: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 8 are implemented.

Images