Photovoltaic protocol converter multi-system adaptive communication method, device and medium

Abstract

The invention discloses a multi-system adaptive communication method and device for a photovoltaic protocol converter, and a medium. The method comprises the following steps: S1, initializing a protocol library in which a protocol feature template is stored by the photovoltaic protocol converter; s2, the converter receives the communication message from the inverter and extracts specified features, message feature recognition is carried out through a built-in protocol library, and the protocol type to which the message belongs is judged; s3, dynamically loading an analysis engine corresponding to the identification result data for message unpacking, data extraction and recombination, executing protocol conversion, and mapping source protocol data to a target protocol format; and S4, outputting the converted message standardized data, uploading the converted message standardized data to an upper system, and monitoring the state. According to the invention, efficient and reliable communication interconnection between photovoltaic devices can be ensured, seamless compatibility is realized, the compatibility problem of heterogeneous photovoltaic device access is solved, and a converter can be allowed to load a specific engine module from a protocol library according to an identification result during operation instead of statically binding all protocols.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of photovoltaic power generation system communication, and particularly relates to a photovoltaic protocol converter multi-standard adaptive communication method, equipment and a storage medium. BACKGROUND

[0002] In a photovoltaic power generation system, an inverter as a core device is responsible for converting direct current into alternating current, and data interaction is carried out with an upper computer and a monitoring system through a communication interface. However, with the rapid development of photovoltaic technology, inverters of different manufacturers often adopt multiple communication protocols, such as the Modbus protocol used for industrial automation, the SunSpec protocol specially designed for photovoltaic device standardization, and the DL / T645 protocol commonly used in the field of power metering. These protocols have significant differences in data format, message structure, and verification mechanism, resulting in compatibility problems when heterogeneous devices are connected.

[0003] Traditional methods usually rely on manual configuration of protocol converters or development of special adapters, which not only increases the complexity of system integration, but also easily causes configuration errors, communication interruptions, and increased maintenance difficulty. In addition, in large-scale photovoltaic power stations, devices are frequently updated and replaced, and manual adaptation is difficult to keep up with the evolution of protocols, resulting in limited system scalability and low operational efficiency.

[0004] Meanwhile, based on traditional technical methods, the more severe challenge is that in actual applications, non-standard protocol "dialects" (i.e., minor modifications of standard protocols by manufacturers) and data frame damage or incompleteness problems caused by interference in complex electromagnetic environments often occur. These problems make traditional recognition methods based on fixed rules easily fail, and cannot guarantee the robustness of communication.

[0005] Therefore, the present application proposes a photovoltaic protocol converter multi-standard adaptive communication method. In view of the heterogeneous characteristics of devices such as inverters in photovoltaic systems using multiple communication protocols (such as Modbus, SunSpec, DL / T645, etc.), a converter design with an embedded protocol library is proposed. By automatically identifying message characteristics and dynamically loading corresponding analysis engines, adaptive access and conversion of different protocols are realized to solve the technical problems of poor compatibility of traditional photovoltaic device access, low efficiency caused by manual configuration of protocols, and insufficient system scalability. SUMMARY

[0006] The main purpose of the present application is to provide a photovoltaic protocol converter multi-standard adaptive communication method, which realizes seamless compatibility through automatic identification and dynamic loading mechanism, improves the overall reliability and intelligent level of photovoltaic systems, and solves the technical problems proposed in the background art.

[0007] The present application solves the above technical problems by adopting the following technical solutions: A multi-standard adaptive communication method for photovoltaic protocol converters, comprising the following steps performed via computer equipment: S1. The converter initializes the built-in protocol library, which pre-stores feature templates for various common protocols, including Modbus register address format, SunSpec model identifier, DL / T645 frame structure, etc. The Modbus template includes a function code sequence, and the SunSpec template contains model identifier bits. S2. The photovoltaic protocol converter receives communication messages from the inverter and extracts specified key features of the messages, such as the start byte, length field, checksum, and data load mode. Then, it automatically identifies the message features through the built-in protocol library to determine the protocol type. The similarity between the extracted features and the protocol library template is used for judgment: if the matching degree is higher than the threshold, the protocol type is confirmed; if there is no match, the learning mode is triggered to supplement the new protocol. S3. Dynamically load the parsing engine corresponding to the identification result data, which is used for message unpacking, data extraction and reassembly, and to perform protocol conversion, mapping the source protocol data to the target protocol format in order to achieve message decoding and conversion, such as converting Modbus register values ​​to SunSpec model parameters; S4. After output conversion, the standardized data of the message is uploaded to the host system and its status is monitored to achieve compatible communication between heterogeneous devices.

[0008] Preferably, in step S1, the converter loads a built-in protocol library upon startup. The feature templates of the protocols in the protocol library are represented by vectors. ,in, For the first A standardized feature template vector for each protocol, used to represent the core patterns of the protocol. For the first Each characteristic component, such as byte value or field length. For the first The weights of each feature represent coefficients assigned based on protocol importance. The feature dimension is the total number of extracted features. Preferably, the specific operations for receiving communication messages from the inverter and extracting specified key features of the messages in step S2 include: The converter captures the message stream through an interface (such as RS485) and extracts the start byte, length, checksum, etc. For the communication messages captured by the photovoltaic protocol converter, modeling them as signal filtering and performing noise filtering to obtain the pure feature sequence, we have:

[0009] in, For time The extracted feature vector at point represents the filtered message features. The input message signal sequence contains the raw byte stream. This refers to a filter kernel function used for noise suppression, such as a Gaussian filter. For the sub-variable representing time offset; Then, principal component analysis (PCA) is used to reduce the dimensionality of the filtered feature vectors, quantifying the dimensionality compression of the feature vectors. The dimensionality reduction formula is as follows:

[0010] in, The feature vectors of the message after dimensionality reduction are used to reduce computational complexity. The principal component matrix is ​​composed of the eigenvectors of the covariance matrix. The original feature vector, This is the feature mean vector used for centering.

[0011] Preferably, the similarity comparison in step S2 uses cosine similarity values. If the similarity exceeds a specified threshold, the protocol type is confirmed, as shown in the formula:

[0012] in, The cosine similarity value ranges from -1 to 1. A higher cosine similarity value indicates a better match. For the first Protocol template vector, It is represented as a vector norm, usually the Euclidean norm.

[0013] Preferably, in step S2, a probability model is used to estimate the matching confidence level for the similarity judgment matching results, so as to improve the matching accuracy. The probability model formula is as follows:

[0014] in, For a given feature Under the condition of the first The conditional probability of the protocol. For the first Protocol similarity, This is a temperature parameter used to control the smoothness of the distribution. This represents the total number of protocols in the protocol library.

[0015] Preferably, during the process of dynamically loading the parsing engine for the corresponding recognition result data in step S3, if there is ambiguity in the recognition result (e.g., multiple protocols), In cases where both are relatively high, multiple candidate engines are loaded in parallel. The loading process can be modeled as resource allocation optimization, with the following allocation optimization formula:

[0016] in, To allocate the optimized selected engine modules, For the engine The computational cost, such as CPU usage, As a trade-off factor, used to balance cost and delay, For the engine Loading delay time, This is the set of available engines.

[0017] Furthermore, based on the optimized allocation of engine modules, if a multi-engine collaborative reasoning scenario exists, the confidence scores of the candidate engines are calculated and ranked, and the engine with the highest confidence score is selected. The formula for calculating the confidence score is as follows:

[0018] in, For the engine The final confidence score, The score is assigned based on syntax; 1 is awarded for a successful CRC check, and 0 is awarded for an unsuccessful CRC check. This is a semantic score, used to measure whether the parsed data is within a predefined reasonable range, such as whether the voltage value is within the range of [0, 1000V]. and These are the weights for the corresponding scores.

[0019] Preferably, in step S3, during the protocol conversion process, the data mapping process employs a linear transformation representation, which has the following: ,in, For the target protocol data vector, This is a mapping matrix used to correspond to specified defined fields. For source protocol data vectors, This is the bias vector, used to handle offsets; For complex data relationships, including those involving identity transformations, that cannot be expressed using linear transformation mappings, polynomial extensions are used, and the extended formula is as follows: ,in, For the first Target field value, For polynomial coefficients, For the input source value, It is the order of the polynomial, usually 2-3.

[0020] Preferably, after the data conversion, uploading, and integrity verification in step S4, the conversion accuracy and protocol diversity are evaluated, and anomalies in the time series model prediction are monitored, wherein: (1) The conversion accuracy is evaluated using an error function, and the calculation formula is as follows: ,in, The root mean square error is used to represent the accuracy of the conversion. For data The actual output value, For data The expected output value, For the number of data points; (2) Introducing entropy calculation to assess protocol diversity, the calculation formula is as follows: ,in, Information entropy is used to measure protocol uncertainty. For the first The probability of the protocol occurring For the number of protocols; (3) By constructing a time series model to predict and monitor anomalies, we have: ,in, To predict the state for the next moment, Let α be the smoothing factor, and 0 < α < 1. This represents the current state value, such as communication delay. This represents the predicted state at this moment; If anomalies are detected, a Z-score is calculated to quantify the abnormal deviation. The formula for calculating the Z-score is: ,in, For standardized scores, z > 3 indicates an anomaly. This is the current signal. To output the mean, To output the standard deviation.

[0021] Preferably, if an anomaly is detected after data conversion and uploading in step S4, and the upper-level system is a multi-device environment, then the communication throughput is optimized, with the following optimization formula:

[0022] in, For throughput, For the first Message bytes This represents the total time.

[0023] In another aspect, the present invention also discloses a computer-readable storage medium storing a computer program, which, when executed by a processor, causes the processor to perform the steps of the method described above.

[0024] In another aspect, the present invention also discloses a computer device, including a memory and a processor, wherein the memory stores a computer program, and when the computer program is executed by the processor, the processor performs the steps of the method described above.

[0025] As can be seen from the above technical solution, the present invention provides a multi-standard adaptive communication method for photovoltaic protocol converters. Compared with the prior art, the present invention has the following advantages: 1. The method of the present invention can improve system integration, reduce maintenance costs, and support future protocol expansion through automated identification and dynamic loading mechanisms, ensuring efficient and reliable communication and interconnection between photovoltaic devices, achieving seamless compatibility, and improving the overall reliability and intelligence level of the photovoltaic system.

[0026] 2. The method of this invention achieves accurate matching by constructing a protocol semantic fingerprint, avoiding the limitations of traditional methods based on fixed ports or preset rules. It also further incorporates semantic information such as the statistical characteristics of data load and the correlation between fields, making it more robust to partial damage to messages and protocol variants. This solves the compatibility problem of heterogeneous photovoltaic equipment access, especially in hybrid protocol environments, reducing human error and improving system response speed.

[0027] 3. By constructing a dynamic loading parsing engine and a multi-engine collaborative reasoning mechanism, this invention enables the converter to load specific engine modules from the protocol library based on the identification results during runtime, rather than statically binding all protocols. This not only optimizes resource utilization but also supports hot updates of new protocol engines, thus enhancing the flexibility of photovoltaic systems and the accuracy of decision-making in uncertain environments, making it suitable for dynamically changing equipment environments.

[0028] It should be understood that the descriptions in this section are not intended to identify key or essential features of embodiments of the invention, nor are they intended to limit the scope of the invention. Other features of the invention will become readily apparent from the following description. Of course, implementing any product of the invention does not necessarily require achieving all of the advantages described above simultaneously. Attached Figure Description

[0029] The accompanying drawings, which form part of this application, are used to provide a further understanding of the invention. The illustrative embodiments of the invention and their descriptions are used to explain the invention and do not constitute an undue limitation of the invention. In the drawings: Figure 1 This is a schematic diagram of the overall process operation steps of the present invention. Detailed Implementation

[0030] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present invention, and not all of them. Unless otherwise specified, the embodiments and features in the embodiments of this application can be combined with each other. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0031] For details in the embodiments, please refer to Figure 1 .

[0032] like Figure 1 As shown in the embodiments of the present invention, the multi-standard adaptive communication method for photovoltaic protocol converters is applicable to distributed photovoltaic power plants, photovoltaic monitoring systems, and other fields. It can improve system integration, reduce maintenance costs, support future protocol expansion, and ensure efficient and reliable communication interconnection between photovoltaic devices. It can play a key role in the standardization process of smart grid and renewable energy communication, specifically including the following steps: S1. The converter initializes the built-in protocol library, which pre-stores feature templates for various common protocols, including Modbus register address format, SunSpec model identifier, DL / T645 frame structure, etc. The Modbus template includes a function code sequence, and the SunSpec template contains model identifier bits.

[0033] This step is to ensure the integrity of the baseline data for subsequent identification. The template is represented by a vector, and in a specific embodiment, the standardized form of the template is calculated using a formula.

[0034] The converter loads a built-in protocol library upon startup. The feature templates of the protocols within the library are represented by vectors. ,in, For the first A standardized feature template vector for each protocol, used to represent the core patterns of the protocol. For the first Each characteristic component, such as byte value or field length. For the first The weights of each feature represent coefficients assigned based on protocol importance. The feature dimension is the sum of the number of extracted features. This formula uses weighted summation and normalization to ensure consistency of the template vector in the high-dimensional space, facilitating subsequent comparisons. In practice, the weights can be adjusted based on historical data to improve initialization accuracy.

[0035] S2. The photovoltaic protocol converter receives communication messages from the inverter and extracts specified key features of the messages, such as the start byte, length field, checksum, and data load mode. Then, it automatically identifies the message features through the built-in protocol library to determine the protocol type. The similarity between the extracted features and the protocol library template is compared: if the matching degree is higher than the threshold, the protocol type is confirmed; if there is no match, the learning mode is triggered to supplement the new protocol.

[0036] Since the message may be subject to electromagnetic interference at the photovoltaic site, this step uses digital signal processing to improve robustness.

[0037] The specific operations for receiving communication messages from the inverter and extracting specified key features of the messages include: The converter captures the message stream through an interface (such as RS485) and extracts the start byte, length, checksum, etc. For the communication messages captured by the photovoltaic protocol converter, modeling them as signal filtering and performing noise filtering to obtain the pure feature sequence, we have:

[0038] in, For time The extracted feature vector at point represents the filtered message features. The input message signal sequence contains the raw byte stream. This refers to a filter kernel function used for noise suppression, such as a Gaussian filter. To represent the time offset, this convolution formula is used to simulate the filtering process of feature extraction, ensuring that key fields are isolated from noisy environments; Then, principal component analysis (PCA) is used to reduce the dimensionality of the filtered feature vectors, quantifying the dimensionality compression of the feature vectors. The dimensionality reduction formula is as follows:

[0039] in, The feature vectors of the message after dimensionality reduction are used to reduce computational complexity. The principal component matrix is ​​composed of the eigenvectors of the covariance matrix. The original feature vector, The mean vector of features used for centering is derived from the PCA principle and is used to help preserve major variations in high-dimensional messages, thereby improving recognition efficiency.

[0040] Furthermore, it should be noted that the similarity comparison here uses cosine similarity values. If the similarity exceeds a specified threshold, the protocol type is confirmed, as shown in the formula:

[0041] in, The cosine similarity value ranges from -1 to 1. A higher cosine similarity value indicates a better match. For the first Protocol template vector, Represented as a vector norm, typically the Euclidean norm, this formula quantifies the matching degree. Through parallel computation across multiple protocols, the optimal match is selected in real-time. If all similarities are low at this point, the system can enter learning mode and update the library.

[0042] Therefore, further, for the matching results of similarity judgment, a probability model is set up to estimate the matching confidence level in order to improve the matching accuracy. The probability model formula is as follows:

[0043] in, For a given feature Under the condition of the first The conditional probability of the protocol. For the first Protocol similarity, This is a temperature parameter used to control the smoothness of the distribution. This represents the total number of protocols in the protocol library.

[0044] Using the softmax formula for soft classification ensures that the probability distribution is output even in ambiguous situations, facilitating threshold decision-making.

[0045] In summary, automatic message feature identification achieves accurate matching by constructing a protocol semantic fingerprint, avoiding the limitations of traditional methods based on fixed ports or preset rules. This fingerprint not only includes syntactic features such as byte sequence patterns and field distribution, but also incorporates semantic information such as statistical characteristics of data load and relationships between fields, making it more robust to partial message corruption and protocol variations.

[0046] Specifically, the converter employs a machine learning-based feature extraction algorithm to scan packets in real time and calculate the matching score for each protocol template. For example, for Modbus packets, it focuses on checking function codes and CRC checks; for SunSpec, it focuses on model IDs and data block boundaries. This innovation solves the compatibility issues of heterogeneous photovoltaic device access, especially in hybrid protocol environments, reducing human error and improving system response speed.

[0047] S3. Dynamically load the parsing engine corresponding to the identification result data, which is used for message unpacking, data extraction and reassembly, and to perform protocol conversion, mapping the source protocol data to the target protocol format to achieve message decoding and conversion, such as converting Modbus register values ​​to SunSpec model parameters.

[0048] It can be further explained that during the process of loading the parsing engine for the corresponding recognition result data, if there is ambiguity in the recognition result (e.g., multiple protocols), In cases where both are relatively high, multiple candidate engines are loaded in parallel. The loading process can be modeled as resource allocation optimization, with the following allocation optimization formula:

[0049] in, To allocate the optimized selected engine modules, For the engine The computational cost, such as CPU usage, As a trade-off factor, used to balance cost and delay, For the engine Loading delay time, This is the set of available engines.

[0050] Furthermore, based on the optimized allocation of engine modules, if a multi-engine collaborative reasoning scenario exists, the confidence scores of the candidate engines are calculated and ranked, and the engine with the highest confidence score is selected. The formula for calculating the confidence score is as follows:

[0051] in, For the engine The final confidence score, The score is assigned based on syntax; 1 is awarded for a successful CRC check, and 0 is awarded for an unsuccessful CRC check. This is a semantic score, used to measure whether the parsed data is within a predefined reasonable range, such as whether the voltage value is within the range of [0, 1000V]. and These are the weights for the corresponding scores.

[0052] Furthermore, during the protocol conversion process, the data mapping process uses a linear transformation representation, which has the following: ,in, For the target protocol data vector, This is a mapping matrix used to correspond to specified defined fields. For source protocol data vectors, This is the bias vector, used to handle offsets; For complex data relationships, including those involving identity transformations, that cannot be expressed using linear transformation mappings, polynomial extensions are used, and the extended formula is as follows: ,in, For the first Target field value, For polynomial coefficients, For the input source value, Given a polynomial order, typically 2-3, this affine transformation formula is used to ensure lossless data conversion. In a specific embodiment, in photovoltaics, it facilitates the migration of voltage data from one protocol to another.

[0053] By constructing a dynamically loaded parsing engine and a multi-engine collaborative inference mechanism, the converter can load specific engine modules from the protocol library at runtime based on the recognition results, rather than statically binding all protocols. This not only optimizes resource utilization (such as memory and computational overhead) but also supports hot updates of new protocol engines.

[0054] Therefore, when message features are ambiguous or highly similar to multiple protocol templates, the system initiates multi-engine collaborative reasoning: it loads multiple candidate parsing engines in parallel, with each engine attempting to decode the message. The system scores the output of each engine based on dimensions such as decoding success rate (e.g., whether CRC check passes), the validity of the parsed data (e.g., whether voltage and current values ​​are within reasonable ranges), and continuity with historical data, ultimately selecting the engine with the highest score as the final parsing solution. Through modular design, the engines can be updated independently, ensuring the system's scalability for future protocols. For example, after identifying the DL / T645 protocol, the engine dynamically parses its multi-frame structure and processes encrypted fields, achieving seamless conversion. This innovation significantly improves the flexibility of photovoltaic systems and the accuracy of decision-making in uncertain environments, making it suitable for dynamically changing equipment environments.

[0055] S4. After outputting the standardized data of the converted message, upload it to the host system and monitor its status. If there is an abnormality, back off and retry, ultimately achieving compatible communication between heterogeneous devices.

[0056] After data conversion, uploading, and integrity verification, the conversion accuracy and protocol diversity are evaluated, and anomalies in time series model predictions are monitored. (1) The conversion accuracy is evaluated using an error function, and the calculation formula is as follows: ,in, The root mean square error is used to represent the accuracy of the conversion. For data The actual output value, For data The expected output value, For the number of data points, the RMSE formula here is used for quality control, and if e is high, retry is performed; (2) Introducing entropy calculation to assess protocol diversity, the calculation formula is as follows: ,in, Information entropy is used to measure protocol uncertainty. For the first The probability of the protocol occurring The number of protocols; if there is a high H, it indicates that stronger adaptation is required. Furthermore, when the system continuously receives unrecognizable messages, this method will trigger an unsupervised learning mode, using clustering algorithms such as DBSCAN to cluster the semantic fingerprints of unknown messages. When a stable new cluster is formed, the system will mark it as a new protocol to be learned, thereby realizing the discovery of unknown protocols and ensuring system stability. (3) By constructing a time series model to predict and monitor anomalies, we have: ,in, To predict the state for the next moment, Let α be the smoothing factor, and 0 < α < 1. This represents the current state value, such as communication delay. This represents the predicted state at this moment; If anomalies are detected, a Z-score is calculated to quantify the abnormal deviation. The formula for calculating the Z-score is: ,in, For standardized scores, z > 3 indicates an anomaly. This is the current signal. To output the mean, To output the standard deviation.

[0057] Furthermore, it should be noted that if any anomalies are detected after data conversion and upload, and the host system is a multi-device environment, then the communication throughput should be optimized using the following formula:

[0058] in, For throughput, For the first Message bytes This represents the total time.

[0059] In summary, this method, through automated identification and dynamic loading mechanisms, can improve system integration, reduce maintenance costs, support future protocol expansion, ensure efficient and reliable communication and interconnection between photovoltaic devices, achieve seamless compatibility, and enhance the overall reliability and intelligence level of the photovoltaic system.

[0060] The core of the above method lies in the adaptive mechanism, which avoids traditional manual intervention and ensures efficient access. In addition, in practical applications, this method can be integrated into embedded converter hardware, supporting multiple physical interfaces such as RS485 and Ethernet, further enhancing the robustness of the photovoltaic monitoring system.

[0061] Based on the above embodiments, this application further proposes a comparative test example of a multi-standard adaptive communication method for photovoltaic protocol converters to demonstrate the technical effects of this application: Construct the following test scenario: Simulate a hybrid photovoltaic power station including 10 inverters, using Modbus (4 inverters), SunSpec (4 inverters), and DL / T645 (2 inverters) protocols respectively. One Modbus inverter is configured to use non-standard function codes to simulate a protocol "dialect". Specific testing process: The converter connects to these devices, receives real-time messages (such as voltage and current data), performs adaptive identification and conversion, and outputs the data to a unified monitoring system. The test lasts for 24 hours, injecting noise (such as 5% message loss and random byte errors) and new protocol variants, comparing the performance of this invention with traditional manual configuration methods, including identification accuracy, conversion latency, and compatibility success rate.

[0062] Based on the above tests, the technical comparison results are shown in the table below:

[0063] It can be concluded that the present invention outperforms traditional methods in terms of accuracy and latency, mainly due to the robustness of protocol semantic fingerprints and the multi-engine collaborative reasoning mechanism, which reduces human intervention and improves overall efficiency, especially in environments with protocol variants and communication noise.

[0064] In another aspect, the present invention also discloses a computer-readable storage medium storing a computer program, which, when executed by a processor, causes the processor to perform the steps of the method described above.

[0065] In another aspect, the present invention also discloses a computer device, including a memory and a processor, wherein the memory stores a computer program, and when the computer program is executed by the processor, the processor performs the steps of the method described above.

[0066] In another embodiment provided in this application, a computer program product containing instructions is also provided, which, when run on a computer, causes the computer to execute any of the photovoltaic protocol converter multi-standard adaptive communication methods described in the above embodiments.

[0067] It is understood that the system provided in the embodiments of the present invention corresponds to the method provided in the embodiments of the present invention, and the explanation, examples and beneficial effects of the relevant content can be referred to the corresponding parts of the above methods.

[0068] This application also provides an electronic device, including a processor, a communication interface, a memory, and a communication bus, wherein the processor, communication interface, and memory communicate with each other via the communication bus. Memory, used to store computer programs; When the processor executes the program stored in the memory, it implements the above-mentioned multi-standard adaptive communication method for photovoltaic protocol converters.

[0069] The communication bus mentioned in the above-mentioned electronic devices can be a standard bus for interconnecting peripheral components or an extended industrial standard structure bus, etc. This communication bus can be divided into address bus, data bus, control bus, etc.

[0070] The communication interface is used for communication between the aforementioned electronic devices and other devices.

[0071] The memory may include random access memory or non-volatile memory, such as at least one disk storage device. Optionally, the memory may also be at least one storage device located remotely from the aforementioned processor.

[0072] The processors mentioned above can be general-purpose processors, including central processing units, network processors, etc.; they can also be digital signal processors, application-specific integrated circuits, field-programmable gate arrays or other programmable logic devices, discrete gate or transistor logic devices, or discrete hardware components.

[0073] In the above embodiments, implementation can be achieved entirely or partially through software, hardware, firmware, or any combination thereof. When implemented using software, it can be implemented entirely or partially 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, all or part of the processes or functions described in the embodiments of this application are generated. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, the computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via wired or wireless means. The computer-readable storage medium can be any available medium accessible to a computer or a data storage device such as a server or data center that integrates one or more available media. The available medium can be a magnetic medium, an optical medium, or a semiconductor medium, etc.

[0074] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

[0075] Furthermore, it should be noted that if any directional indication (such as up, down, left, right, front, back, etc.) is involved in the embodiments of the present invention, the directional indication is only used to explain the relative positional relationship and movement of each component in a specific posture. If the specific posture changes, the directional indication will also change accordingly.

[0076] Furthermore, if the embodiments of this invention involve descriptions such as "first" or "second," these descriptions are for descriptive purposes only and should not be construed as indicating or implying their relative importance or implicitly specifying the number of technical features indicated. Therefore, a feature defined with "first" or "second" may explicitly or implicitly include at least one of those features. Additionally, the meaning of "and / or" throughout the text includes three parallel solutions; for example, "A and / or B" includes solution A, solution B, or a solution where both A and B are satisfied simultaneously. Furthermore, in the embodiments of this invention, "multiple" refers to two or more. Moreover, the technical solutions of the various embodiments can be combined with each other, but this must be based on the ability of those skilled in the art to implement them. When the combination of technical solutions is contradictory or cannot be implemented, it should be considered that such a combination of technical solutions does not exist and is not within the scope of protection claimed by this invention.

Claims

1. A multi-standard adaptive communication method for a photovoltaic protocol converter, characterized in that, include: S1. The converter initializes the built-in protocol library, which pre-stores feature templates for specified protocols; S2. The photovoltaic protocol converter receives communication messages from the inverter and extracts the specified features of the messages. Then, it automatically identifies the message features through the built-in protocol library to determine the protocol type. The similarity between the extracted features and the protocol library template is compared: if the matching degree is higher than the threshold, the protocol type is confirmed. If no match is found, a learning mode is triggered to supplement a new protocol; S3. Dynamically load the parsing engine corresponding to the identification result data, which is used for message unpacking, data extraction and reassembly, and performs protocol conversion to map the source protocol data to the target protocol format; S4. After outputting the standardized data of the converted message, upload it to the host system and monitor its status.

2. The multi-standard adaptive communication method for photovoltaic protocol converters as described in claim 1, characterized in that, In step S1, the converter loads a built-in protocol library upon startup. The feature templates of the protocols within this library are represented by vectors. ,in, For the first Standardized feature template vectors of each protocol, For the first Each feature component For the first The weights of each feature Let be the feature dimension.

3. The multi-standard adaptive communication method for photovoltaic protocol converters as described in claim 2, characterized in that, The specific operations for extracting the specified key features of the message in step S2 include: For the communication messages captured by the photovoltaic protocol converter, modeling them as signal filtering and performing noise filtering to obtain the pure feature sequence, we have: in, For time Extracting feature vectors at the location, For the input message signal sequence, Here is the filter kernel function. For the sub-variable representing time offset; Then, principal component analysis is used to reduce the dimensionality of the filtered feature vectors, quantifying the dimensionality compression of the feature vectors. The dimensionality reduction formula is as follows: in, The feature vector after dimensionality reduction of the message. The principal component matrix is ​​composed of the eigenvectors of the covariance matrix. The original feature vector, This is the feature mean vector used for centering.

4. The multi-standard adaptive communication method for photovoltaic protocol converters as described in claim 3, characterized in that, The similarity comparison in step S2 uses a cosine similarity value. If the similarity exceeds a specified threshold, the protocol type is confirmed, as shown in the formula: in, The cosine similarity value ranges from -1 to 1. A higher cosine similarity value indicates a better match. For the first Protocol template vector, It is represented as a vector norm.

5. The multi-standard adaptive communication method for photovoltaic protocol converters as described in claim 4, characterized in that, In step S2, for the similarity judgment matching results, a probability model is set up to estimate the matching confidence, and the probability model formula exists: in, For a given feature Under the condition of the first The conditional probability of the protocol. For the first Protocol similarity, For temperature parameters, This represents the total number of protocols in the protocol library.

6. The multi-standard adaptive communication method for photovoltaic protocol converters as described in claim 1, characterized in that, In step S3, during the dynamic loading of the parsing engine for the corresponding recognition result data, if there are cases of blurred recognition results, multiple candidate engines are loaded in parallel. The loading process can be modeled as resource allocation optimization, with the following allocation optimization formula: in, To allocate the optimized selected engine modules, For the engine The computational cost As a weighting factor, For the engine Loading delay time, This is the set of available engines.

7. The multi-standard adaptive communication method for photovoltaic protocol converters as described in claim 6, characterized in that, Based on the optimized allocation of engine modules, if a multi-engine collaborative reasoning scenario exists, the confidence scores of the candidate engines are calculated and ranked, and the engine with the highest confidence score is selected. The formula for calculating the confidence score is as follows: in, For the engine The final confidence score, For grammar score, For semantic score, and These are the weights for the corresponding scores.

8. The multi-standard adaptive communication method for photovoltaic protocol converters as described in claim 1, characterized in that, After the data conversion and upload in step S4, the conversion accuracy and protocol diversity are evaluated, and anomalies in the time series model prediction are monitored. (1) The conversion accuracy is evaluated using an error function, and the calculation formula is as follows: ,in, The root mean square error is used to represent the accuracy of the conversion. For data The actual output value, For data The expected output value, For the number of data points; (2) Introducing entropy calculation to assess protocol diversity, the calculation formula is as follows: ,in, For information entropy, For the first The probability of the protocol occurring For the number of protocols; (3) By constructing a time series model to predict and monitor anomalies, we have: ,in, To predict the state for the next moment, As a smoothing factor, This is the current state value. This represents the predicted state at this moment; If anomalies are detected, a Z-score is calculated to quantify the abnormal deviation. The formula for calculating the Z-score is: ,in, For standardized scores, z > 3 indicates an anomaly. This is the current signal. To output the mean, To output the standard deviation.

9. A computer-readable storage medium, characterized in that, The device stores a computer program that, when executed by a processor, causes the processor to perform the steps of the method as described in any one of claims 1 to 8.

10. A computer device, characterized in that, It includes a memory and a processor, the memory storing a computer program that, when executed by the processor, causes the processor to perform the steps of the method as described in any one of claims 1 to 8.

Images