Cloud data-based intelligent integrated design method for box-type substation

Abstract

The present application relates to the technical field of data processing, and especially relates to a box-type transformer substation intelligent integrated design method based on cloud data, which comprises: integrated storage of historical system graph data, component data, cabinet type data and heat dissipation scheme data; through a machine learning algorithm, the correlation between each data after classification and arrangement is learned, and the machine learning algorithm is applied to a prediction model; the prediction model is trained, and according to new input system graph data, corresponding component data, cabinet type data and heat dissipation scheme are output; in the data transmission process, encryption processing is carried out, and the encryption processing adopts a transmission layer security protocol. The present application simplifies the design process, improves the accuracy and speed of the design, optimizes the matching of each component and the heat dissipation scheme, solves the parameter complex correlation problem through systematic data analysis, and can ensure the safety of data in the transmission and storage process, so as to comprehensively improve the design efficiency and quality of the box-type transformer substation.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of data processing, and in particular to a cloud data-based intelligent integrated design method for box-type substations. BACKGROUND

[0002] A box-type substation is an integrated and efficient power equipment widely used in urban power supply, industrial parks, commercial centers, etc. It integrates transformers, high-voltage distribution devices, low-voltage distribution devices, and auxiliary equipment in a sealed box, with the advantages of small footprint, quick installation, and easy mobility. However, the design and configuration of a box-type substation need to consider multiple factors, such as equipment heat dissipation, size matching, ventilation conditions, etc., which directly affect the operational efficiency and safety of the substation.

[0003] According to the current technical development, by introducing artificial intelligence technology, the design efficiency and reliability of the box-type substation can be significantly improved. For example, by using a large amount of historical system diagram design data, component heat dissipation data, component size data, box-type substation cabinet type data, and heat dissipation scheme data stored in the cloud, the correlation between these data can be learned and analyzed through machine learning algorithms, so that when generating new single data, the most suitable other data scheme can be automatically output. This not only speeds up the design process, but also improves the accuracy of the design, ultimately ensuring the good matching of each component in actual use, and ultimately improving the overall performance and safety of the substation.

[0004] However, at present, there is no mature technical solution for the above-mentioned content, and the design of many box-type substations still relies on traditional manual experience and static data, resulting in a tedious and error-prone design process. Designers need to manually search and match component information, making it difficult to quickly and accurately select the appropriate design scheme. In addition, due to the lack of systematic analysis of a large amount of historical data, the complex correlation between various parameters cannot be fully considered in the design process, which may lead to poor heat dissipation, space waste, or equipment mismatch, ultimately affecting the operational stability and maintenance cost of the substation.

[0005] How to provide a specific scheme by introducing cloud data and intelligent integrated design to effectively overcome the above limitations and improve design efficiency and quality has become a technical problem to be solved at present. SUMMARY

[0006] The present application provides a cloud data-based intelligent integrated design method for box-type substations, which can effectively solve the problems in the background art.

[0007] To achieve the above purpose, the technical solution adopted by the present application is:

[0008] The cloud data-based intelligent integrated design method of box-type substations comprises:

[0009] Integrating and storing historical grouped system diagram data, component data, cabinet type data and heat dissipation scheme data to form a cloud data warehouse;

[0010] Classifying and structurally arranging the data stored in the cloud data warehouse in sequence;

[0011] Learning the correlation between each data after classification and arrangement through a machine learning algorithm, and applying the machine learning algorithm to a prediction model;

[0012] Training the prediction model, and outputting corresponding component data, cabinet type data and heat dissipation scheme according to newly input system diagram data through the trained prediction model.

[0013] Further, the transmission of grouped system diagram data, component data, cabinet type data and heat dissipation scheme data to the cloud data warehouse is realized through a data uploading tool, and the transmission process comprises:

[0014] Initiating a connection request to the cloud data warehouse through the data uploading tool, starting a transmission layer security protocol handshake process, and the connection request at least includes a transmission layer security protocol version, supported encryption algorithm and random number;

[0015] The cloud data warehouse responds to the connection request, selects the supported transmission layer security protocol version and encryption algorithm, and sends a message with a digital certificate;

[0016] The data uploading tool verifies the digital certificate, and generates a pre-master key after verification, which is encrypted by the public key of the cloud data warehouse and sent to the cloud data warehouse;

[0017] A session key is generated using the pre-master key and random number, which is used for encrypting and decrypting data;

[0018] The data uploading tool and the cloud data warehouse send a set message to complete the handshake process.

[0019] Further, before the transmission of grouped system diagram data, component data, cabinet type data and heat dissipation scheme data to the cloud data warehouse, the box-type substation corresponding to each group of data is confirmed, and a unique identifier is provided for the box-type substation; the unique identifier is added to the random number.

[0020] Further, the random number is determined, comprising:

[0021] Randomly extracting a part of the unique identifier as the first part of the random number;

[0022] Collecting dynamic characteristic data of the network during data transmission and converting it into a byte sequence as the second part of the random number;

[0023] Using the timestamp during data transmission as the third part of the random number.

[0024] Further, collecting dynamic characteristic data of the network during data transmission and converting it into a byte sequence, including:

[0025] Determining the type of network dynamic characteristic data that needs to be collected, and collecting real-time data for each type;

[0026] Converting the collected network dynamic characteristic data into a byte representation of a set length;

[0027] Combining the byte representations of all types of network dynamic characteristic data into a continuous byte sequence;

[0028] Hash processing the combined byte sequence, and the converted byte sequence as part of the random number.

[0029] Further, the machine learning algorithm is the XGBoost algorithm, which integrates several decision trees, and the prediction model applied by the algorithm is:

[0030] ;

[0031] Wherein, is the component data, cabinet type data and heat dissipation scheme to be predicted;

[0032] is the kth decision tree of the XGBoost algorithm;

[0033] K is the total number of decision trees;

[0034] is the feature vector corresponding to the newly input system graph data.

[0035] Further, the calculation formula of the total number of decision trees K is:

[0036] K=N / η⋅C

[0037] Wherein, N is the number of features in the feature vector corresponding to the newly input system graph data;

[0038] η is the learning rate set;

[0039] C is a regulation constant for adjusting the proportion of the calculation formula.

[0040] Further, the calculation formula of the regulation constant C is:

[0041] C=aN+b

[0042] Wherein, a and b are fitting coefficients, the relationship between C and N is obtained by linear regression method fitting.

[0043] Further, the stored data in the cloud data warehouse is classified, including: determining the classification standard of the stored data, and adding a classification label to each data item according to the classification standard.

[0044] Further, the stored data in the cloud data warehouse is classified, including:

[0045] According to the characteristics and classification results of the classified data, a data model is defined;

[0046] According to the data model, a database structure is designed, and corresponding tables and relationships are created;

[0047] Data cleaning and conversion are performed to realize the matching of data and tables and relationships in the database structure.

[0048] Through the technical scheme of the present application, the following technical effects can be achieved:

[0049] The box-type substation intelligent integrated design method based on cloud data in the present application effectively overcomes the limitations in the prior art, simplifies the design process, improves the accuracy and speed of the design, optimizes the matching and heat dissipation scheme of each component, and solves the parameter complex correlation problem through systematic data analysis, while ensuring the security of data in the transmission and storage process, thereby comprehensively improving the design efficiency and quality of the box-type substation. BRIEF DESCRIPTION OF DRAWINGS

[0050] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the following will briefly introduce the drawings needed to be used in the embodiment or prior art description. Obviously, the drawings in the following description are only some embodiments described in the present application, and for those skilled in the art, other drawings can also be obtained without creative labor on the basis of these drawings.

[0051] Figure 1 It is a flowchart of the box-type substation intelligent integrated design method based on cloud data;

[0052] Figure 2 It is a flowchart of the transmission process;

[0053] Figure 3 It is a flowchart of determining the random number. DETAILED DESCRIPTION

[0054] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments.

[0055] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains. The terminology used in this specification is for the purpose of describing particular embodiments only and is not intended to be limiting of the invention. The term "and / or" as used herein includes any and all combinations of one or more of the associated listed items.

[0056] like Figure 1 As shown, the intelligent integrated design method for prefabricated substations based on cloud data includes:

[0057] S1: Integrate and store historical grouped system diagram data, component data, cabinet type data, and heat dissipation solution data to form a cloud data warehouse;

[0058] Specifically, to centrally manage the collected data related to prefabricated substations, this step provides a unified data management platform by constructing a cloud-based data warehouse. This platform stores several sets of system diagram data, component data, cabinet type data, and heat dissipation scheme data for prefabricated substations, facilitating centralized data management, improving data reliability, and providing a foundation for subsequent data analysis and processing. The system diagram data, component data, cabinet type data, and heat dissipation scheme data for the same prefabricated substation constitute one set of data. System diagram data includes main wiring diagrams, protection and control circuit diagrams, equipment layout diagrams, cable wiring diagrams, and single-line diagrams; component data includes model and specifications, rated voltage and current, dimensions and weight, heat dissipation parameters, lifespan, and reliability data; cabinet type data includes cabinet type, dimensions and layout, material and structure, interfaces and connection points, heat dissipation characteristics, and installation methods; and heat dissipation scheme data includes heat dissipation technology types, heat dissipation components, heat dissipation effect evaluation, environmental adaptability, cost, and energy consumption.

[0059] S2: Classify and structure the data stored in the cloud data warehouse in sequence;

[0060] On the basis of the above embodiments, classification standards for various types of data can be established to organize and classify the data stored in the cloud data warehouse. The data stored in the cloud data warehouse has its own specific attributes and purposes. Through classification, these data can be better managed and utilized, improving data retrievability and operability. For the classification results, the data can be stored in different database tables or columns for structured arrangement to facilitate management and query. Therefore, this step integrates data classification and data structured arrangement to improve data quality and ensure the effectiveness and accuracy of data in subsequent analysis and machine learning processes.

[0061] S3: Through a machine learning algorithm, the correlation between the classified and arranged data is learned, and the machine learning algorithm is applied to a prediction model.

[0062] S4: The prediction model is trained, and the trained prediction model outputs corresponding component data, cabinet type data, and heat dissipation scheme according to new input system diagram data.

[0063] Specifically, by analyzing the arranged data, the internal relationship and pattern between different data can be found. Appropriate machine learning algorithms such as decision trees, random forests, neural networks, etc. can be used to construct a prediction model according to the selected machine learning algorithm. The data stored in the cloud data warehouse can be divided into training set and test set, and the prediction model can be repeatedly trained and verified. Techniques such as cross-validation can be used to evaluate model performance and continuously adjust parameters to improve model accuracy. Through sufficient training, the model can have high prediction ability and generalization ability, and can adapt to different input data to accurately predict the design scheme corresponding to new data. Through machine learning and prediction model, the system can learn and optimize autonomously, reducing human intervention, improving the intelligent level of design, and automatically predicting and outputting design scheme, improving design efficiency. At the same time, sufficient training and verification also ensure the prediction accuracy of the model, thereby improving the reliability and rationality of the box-type substation design scheme.

[0064] The box-type substation intelligent integrated design method based on cloud data in the present application effectively overcomes the limitations in the current technology, simplifies the design process, improves the accuracy and speed of design, optimizes the matching and heat dissipation scheme of each component, and solves the parameter complex correlation problem through systematic data analysis, thereby comprehensively improving the design efficiency and quality of the box-type substation.

[0065] On the basis of the above embodiments, as shown in Figure 2 The transmission of grouped system diagram data, component data, cabinet type data, and heat dissipation scheme data to the cloud data warehouse is realized through a data uploading tool. The transmission process includes:

[0066] A1: Initiate a connection request to the cloud data warehouse through the data upload tool, start the transmission layer security protocol handshake process, and the connection request at least includes the transmission layer security protocol version, supported encryption algorithm and random number;

[0067] This step can select command line tools, APIs and SDKs, and data synchronization tools to initiate a connection request and start a transmission layer security protocol handshake process to establish a secure connection, ensuring that data is encrypted and complete during transmission. The transmission layer security protocol version, supported encryption algorithm and random number are indispensable elements in the handshake process. The transmission layer security protocol is used to ensure the security of data transmission. The protocol version and encryption algorithm information in the connection request are used to negotiate the security communication standard used by both parties. The random number is used for subsequent key generation to ensure the uniqueness and security of each connection communication.

[0068] A2: The cloud data warehouse responds to the connection request and selects the supported transmission layer security protocol version and encryption algorithm, and sends a message with a digital certificate;

[0069] After receiving the connection request initiated by the data upload tool, the cloud data warehouse can parse the request content, select the optimal version and encryption algorithm supported by both parties to ensure smooth communication, and send a message with a digital certificate. This can be used to verify the identity of the cloud data warehouse, avoid connecting to a fake server, and prevent man-in-the-middle attacks. The digital certificate contains a public key and other authentication information, which is used to encrypt sensitive data in subsequent communications.

[0070] A3: Verify the digital certificate through the data upload tool, and generate a pre-master key after verification is complete, which is encrypted by the public key of the cloud data warehouse and sent to the cloud data warehouse;

[0071] After receiving the digital certificate sent by the cloud data warehouse, the data upload tool can verify it to ensure its authenticity and validity. Verifying the digital certificate ensures that the identity of the cloud data warehouse is authentic and trustworthy, preventing forgery. After verification, a pre-master key is generated, which is a temporary key used to generate a session key later. The pre-master key can then be encrypted by the public key and sent to the cloud data warehouse, ensuring the security of the key exchange process.

[0072] A4: Generate a session key using the pre-master key and random number, which is used to encrypt and decrypt data;

[0073] The pre-master key and random number can be used to generate a symmetrically encrypted session key through a key generation algorithm, such as a pseudo-random function, ensuring that the key for each communication is unique. The session key can be used for subsequent data encryption and decryption, ensuring the confidentiality and integrity of data transmission.

[0074] A5: The data uploading tool and the cloud data warehouse send a setting message to complete the handshake process.

[0075] Specifically, the two parties confirm the successful completion of the handshake process by exchanging setting messages. The setting message can contain a handshake message digest to ensure that the data transmitted during the entire handshake process has not been tampered with. After the handshake is completed, the two parties can use a session key for secure data transmission. The entire transmission process is carried out through a transport layer security protocol to ensure the security of data transmission. Through digital certificate verification and encryption key exchange, data leakage and man-in-the-middle attacks are prevented. The verification and confirmation mechanisms included in each step ensure that the connection between the data uploading tool and the cloud data warehouse is reliable, and the integrity and confidentiality of the data during transmission are guaranteed. Although the transport layer security protocol handshake process is complex, it runs quickly in a modern computing environment, ensuring the efficiency of the data transmission process.

[0076] Further, before the grouped system diagram data, component data, cabinet type data, and heat dissipation scheme data are transmitted to the cloud data warehouse, the box-type substation corresponding to each group of data is confirmed, and a unique identifier is provided for the box-type substation. The unique identifier is added to the random number.

[0077] In this embodiment, the integrity and accuracy of each group of data can be checked and verified to ensure that the data can be correctly mapped to the corresponding box-type substation, and a universally unique identifier or other unique coding method can be used to generate a unique identifier for each box-type substation. In the transport layer security protocol handshake process, the unique identifier of the box-type substation can be embedded in the random number to ensure that the data source can be identified during the handshake process, and certain format processing can be performed to ensure the validity of the random number.

[0078] Further, as shown in Figure 3 determining the random number, comprising:

[0079] randomly extracting a part of the unique identifier as the first part of the random number;

[0080] As a preferred embodiment of the above, during the random number generation process, a specific position of the unique identifier or a few characters, numbers, or specific digits can be randomly selected as the first part of the random number. Since the identifier of each box-type substation is unique, extracting a part of the unique identifier as part of the random number can increase the uniqueness of the generated random number.

[0081] Collecting dynamic characteristic data of the network during data transmission and converting it into a byte sequence as the second part of the random number;

[0082] Specifically, the network dynamic characteristic data, including dynamic properties of the network during data transmission, can be obtained using system functions or third-party libraries. These characteristics change over time and with changes in network status, so converting the collected dynamic characteristic data into a byte sequence as the second part of the random number increases the randomness of the random number.

[0083] The timestamp during data transmission is used as the third part of the random number.

[0084] The timestamp marks the specific time of random number generation or collection, which is crucial for recording and identifying the time sequence of data. As time passes, random numbers generated at different time points will be different, so using the timestamp during data transmission as part of the random number can increase the timeliness of the random number. At the same time, timeliness can ensure that the generated random number is associated with a specific time point, ensuring that each generation of the random number is unique in time.

[0085] For example, assume that the unique identifier of a box-type substation is ABC123, and a part of the unique identifier, ABC1, is extracted as the first part of the random number. The current network dynamic characteristic data is network delay 100 ms, bandwidth 100 Mbps, and packet loss rate 0.1%, and the network dynamic characteristic data is converted into a byte sequence b'100ms_100Mbps_0.1%' as the second part of the random number. The time of generating the random number is June 25, 2021, 12:25:02, and the timestamp 1735728002 is converted as the third part of the random number. Integrating the above three parts together, the final random number can be generated as follows:

[0086] Random number = ABC1 + b'100ms_100Mbps_0.1%' + 1735728002;

[0087] Where b is a byte sequence object.

[0088] In practical applications, the unique identifier, network dynamic characteristic data, and timestamp are independent and unrelated information. Their combination order does not affect the nature of each part or the integrity of the information. For example, the timestamp can also be used as the starting point first, then the network dynamic characteristic data is added, and finally the part of the unique identifier is added, so the order of the three parts is flexible. Various order forms are within the scope of the present application.

[0089] Further, the network dynamic characteristic data during data transmission is collected and converted into a byte sequence, including:

[0090] B1: Determine the type of network dynamic characteristic data to be collected, and collect real-time data for each type.

[0091] B2: Convert the collected network dynamic feature data into a byte representation of a set length;

[0092] B3: Combine the byte representations of all types of network dynamic feature data into a continuous byte sequence;

[0093] B4: Hash the combined byte sequence and use the converted byte sequence as part of the random number.

[0094] Specifically, the type of network dynamic feature data to be collected can be determined, including but not limited to network delay, bandwidth utilization, packet loss rate, packet transmission delay, etc. Each type of network dynamic feature data collected needs to be converted into a fixed-length byte representation. This conversion usually involves encoding the data into a byte sequence according to certain rules to ensure consistency and reliability of the data during transmission and processing; all collected network dynamic feature data byte representations are combined into a continuous byte sequence, which forms a complete data block containing all the information needed for subsequent hash processing; the combined byte sequence is hashed, and the hash processing can be implemented through an encryption hash function to map the complex byte sequence to a fixed-length hash value. The result of the hash processing is a fixed-length data digest that cannot be reversed and can uniquely represent the input data block; the result of the hash processing is used as part of the random number, which can ensure that the generated random number is unique and secure while retaining important features of the original data, making it suitable for authentication, security verification, and data integrity checking applications.

[0095] Further, the machine learning algorithm is the XGBoost algorithm, which integrates several decision trees, and the prediction model applied by the algorithm is:

[0096] ;

[0097] wherein, is the component data, cabinet type data, and cooling scheme to be predicted;

[0098] is the kth decision tree of the XGBoost algorithm;

[0099] K is the total number of decision trees;

[0100] is the feature vector corresponding to the newly input system graph data.

[0101] As a preferred embodiment of the present embodiment, the XGBoost algorithm is selected to learn and predict the correlation between various data. The XGBoost algorithm models the complex nonlinear relationship between data by integrating multiple decision trees. Due to the complex interdependence and nonlinear relationship between various data types in the intelligent integrated design of box-type substations, the XGBoost algorithm can effectively capture and model these complex relationships, thereby improving the accuracy and reliability of the prediction. The XGBoost algorithm in this scheme can effectively handle the complex electrical connection relationships and layout requirements between different parts of the system graph data, thereby optimizing the design and layout of the electrical system to ensure safety and performance requirements; it can also analyze the correlation between component data and system operating parameters to predict the most suitable component selection and configuration; it can consider multiple aspects of cabinet type data to provide the best cabinet type selection suggestion to ensure safe and effective operation of the equipment; it can also analyze the efficiency and adaptability of different heat dissipation schemes and their performance under various environmental conditions to provide the best heat dissipation solution.

[0102] The selection of the XGBoost algorithm can help identify and understand the most influential features in the data based on the importance evaluation of each feature during the learning process. For box-type substation design, this means being able to determine which system graph data, component data, cabinet type data, and heat dissipation scheme data have a critical impact on the final prediction model, which helps to optimize the design and decision-making process; the XGBoost algorithm also has the ability to handle missing data and outliers, which can improve the robustness and stability of the model through reasonable splitting strategies and handling of missing values.

[0103] Further, the calculation formula of the total number of decision trees K is:

[0104] K=N / η⋅C

[0105] where N is the number of features in the feature vector corresponding to the newly input system graph data;

[0106] η is the learning rate set;

[0107] C is the adjustment constant used to adjust the proportion of the calculation formula.

[0108] N represents the number of features obtained after feature engineering of the newly input system diagram data, which can be numerical or categorical data describing system structure, performance, or other related factors; the learning rate η is a parameter set when training each decision tree, which controls the influence of each tree on the final model prediction, and a smaller learning rate usually leads to more decision trees being trained to achieve ideal model performance; the adjustment constant C is a scaling factor used to adjust the calculation formula, and the selection of the adjustment constant directly affects the size of the total number of decision trees K. The adjustment constant C is usually a pre-set parameter in machine learning and optimization algorithms, which can be obtained by empirical value setting, parameter search (such as grid search, random search, etc.), automatic adjustment (such as evolutionary algorithm, particle swarm optimization, etc.), interactive optimization, etc.

[0109] Taking the calculation of two different total numbers K as an example, the specific process is as follows:

[0110] Example 1, assuming the number of features N=50, the learning rate η=0.1, and the adjustment constant C=2, then K=N / η·C=1000.

[0111] Example 2, assuming the number of features N=80, the learning rate η=0.05, and the adjustment constant C=3, then K=N / η·C=4800.

[0112] From the above two total number K calculation processes, it can be seen that the increase of the number of features will lead to a significant increase in the total number of decision trees K. Larger adjustment constant and smaller learning rate usually lead to smaller step size of parameter update in each iteration, and more decision trees can be trained to capture complex data relationships to improve the accuracy and generalization ability of the model. This calculation formula combines the factors of feature quantity, learning rate and adjustment constant, which can provide an effective way to adjust the number of decision trees for box-type substation intelligent integrated design.

[0113] Further, the calculation formula of the adjustment constant C is:

[0114] C=aN+b

[0115] Where a and b are fitting coefficients, obtained by linear regression method to fit the relationship between C and N.

[0116] The step can realize intelligent optimization of the number K of decision trees by dynamically adjusting the adjustment constant C. At each time when new input data is received, the size of C can be adjusted according to different data quantities and feature complexities through dynamic calculation of a and b. Dynamic adjustment of C can intelligently adjust the number K of decision trees according to the characteristics and quantity of new data. In this way, it can be ensured that the model can always effectively capture the characteristics and relevance of the latest data, improving the accuracy and adaptability of prediction. Through dynamic optimization of the adjustment constant C and dynamic adjustment of the number K of decision trees, the complexity of the model can be effectively controlled, the utilization of computing resources can be optimized, and the problems of overfitting or underfitting can be avoided. At the same time, the process of adjusting the number K of decision trees by dynamically adjusting C using linear regression or other optimization methods can be automated. The system can automatically calculate and adjust the most suitable C value and K value according to the changes in real-time data, without human intervention, improving the intelligence and efficiency of the system, and making it suitable for different sizes and complexities of data sets and application scenarios.

[0117] Further, the data stored in the cloud data warehouse is classified, including: determining the classification standard of the stored data, and adding a classification label to each data item according to the classification standard.

[0118] Specifically, in order to make the data more orderly and efficient when storing and retrieving, the data can be classified and labeled according to the type, purpose or characteristics of the data, reducing the time of data search and processing; system diagram data can be classified according to its functional area, equipment type or electrical connection relationship, and systemically organized and managed system diagram data, for example, system diagram data can be classified according to its functional area into power generation area system diagram, power transformation area system diagram, power distribution area system diagram, control area system diagram and auxiliary equipment area system diagram, and add classification labels to each category, for example, the wiring diagram of the generator set and its connection to the transformer can add the classification label of power generation area system diagram; component data can be classified according to its type, function or manufacturer, for example, component data can be classified according to its type into switch, cable, capacitor, transformer and circuit breaker, and add classification labels to each category, for example, load switch can add the classification label of switch type component; cabinet type data can be classified according to its purpose, material or installation location, for example, cabinet type data can be classified according to its purpose into low-voltage cabinet, medium-voltage cabinet, high-voltage cabinet and control cabinet, and add classification labels to each category, for example, low-voltage power distribution cabinet can add the classification label of low-voltage cabinet; heat dissipation scheme data can be classified according to the type of heat sink, heat dissipation efficiency or applicable environment, for example, heat dissipation scheme data can be classified according to the type of heat sink into air-cooled, water-cooled, phase-change heat dissipation and hybrid, and add classification labels to each category, for example, air-cooled can add the classification label of air-cooled heat dissipation; for different types of data, by determining the classification standard and classification rule, the corresponding classification label is added to each data item, which helps subsequent data management, analysis and application, and improves the operability of the data and the intelligent degree of the overall system design.

[0119] Further, the structured arrangement of the classified data stored in the cloud data warehouse includes:

[0120] C1: defining a data model according to the characteristics and classification results of the classified data;

[0121] Defining a data model is the process of converting real-world entities and their relationships into a data structure. According to the determined classification data characteristics, the corresponding data model is defined, and the clear data model defines the data structure, ensuring the consistency and integrity of the data, which can effectively manage the data in the cloud data warehouse.

[0122] C2: design the database structure according to the data model, and create the corresponding tables and relationships;

[0123] Designing the database structure and converting the data model into tables and relationships in the database can ensure that the data in the cloud data warehouse is stored and managed in an efficient, orderly and flexible manner. Each data type defines a corresponding table according to its characteristics, and creates the necessary fields for each table. When designing the database structure, not only independent tables need to be created, but also the relationships between tables need to be defined to ensure the relevance and integrity of the data. These relationships can be one-to-many, many-to-many or one-to-one. Clear table and relationship design makes database queries more efficient, allowing quick retrieval and processing of relevant data. Through reasonable table design and relationship definition, data storage space can be optimized, redundant data can be reduced, and storage efficiency can be improved. The design of data structure can also add new tables and fields according to demand without affecting the integrity of existing data.

[0124] C3: Perform data cleaning and conversion to match the data with the tables and relationships in the database structure.

[0125] Data cleaning includes operations such as data deduplication, error correction and missing value filling. Data conversion includes operations such as format conversion, field matching and data mapping. Data cleaning and conversion can transform raw data into data that conforms to the database structure, eliminate errors and redundant data, ensure data quality, and improve data reliability and consistency.

[0126] The basic principles, main features and advantages of the present application are shown and described above. Those skilled in the art should understand that the present application is not limited by the above examples. The above examples and descriptions in the specification are only to illustrate the principles of the present application. Without departing from the spirit and scope of the present application, various changes and improvements can be made to the present application, and these changes and improvements fall within the scope of the claimed present application. The scope of protection of the present application is defined by the appended claims and their equivalents.

Claims

1. A cloud data-based intelligent integrated design method for a box-type substation, characterized in that, The application relates to a system and method for integrating and storing historical system diagram data, component data, cabinet type data and heat dissipation scheme data to form a cloud data warehouse. The data stored in the cloud data warehouse is classified and structured in sequence. A machine learning algorithm is used to learn the correlation between the classified and structured data, and the machine learning algorithm is applied to a prediction model. The machine learning algorithm is an XGBoost algorithm, and the prediction model is a decision tree. The total number K of the decision trees is calculated according to the following formula: K=N / eta*C. ； wherein, the component data, cabinet type data and heat dissipation scheme to be predicted; is the kth decision tree for the XGBoost algorithm; Wherein, N is the number of features in the feature vector corresponding to the newly input system diagram data. a feature vector corresponding to the newly input system graph data; Eta is the learning rate. C is an adjustment constant used to adjust the proportion of the calculation formula. The calculation formula of the adjustment constant C is: C=aN+b. Wherein, a and b are fitting coefficients obtained by fitting the relationship between C and N through linear regression. The prediction model is trained, and the trained prediction model is used to output corresponding component data, cabinet type data and heat dissipation scheme according to newly input system diagram data. The system diagram data, component data, cabinet type data and heat dissipation scheme data are transmitted to the cloud data warehouse through a data uploading tool. The transmission process includes: A connection request is initiated to the cloud data warehouse through the data uploading tool, a transmission layer security protocol handshake process is started, and the connection request at least includes a transmission layer security protocol version, supported encryption algorithm and random number. The cloud data warehouse responds to the connection request, selects the supported transmission layer security protocol version and encryption algorithm, and sends a message with a digital certificate.

2. The cloud data based integrated design method of a box-type substation intelligence according to claim 1, wherein, The digital certificate is verified through the data uploading tool, and a pre-master key is generated after verification is completed, which is encrypted through the public key of the cloud data warehouse and sent to the cloud data warehouse. A session key is generated using the pre-master key and random number, which is used for data encryption and decryption. The data uploading tool and the cloud data warehouse send a set message to complete the handshake process. Before the system diagram data, component data, cabinet type data and heat dissipation scheme data are transmitted to the cloud data warehouse, the corresponding box-type substation of each group of data is confirmed, and a unique identifier is provided for the box-type substation. The unique identifier is added to the random number. The random number is determined by:

3. The cloud data based integrated design method of a box-type substation intelligence according to claim 2, characterized in that, Randomly extracting a part of the unique identifier as the first part of the random number.

4. The cloud data based kiosk substation intelligent integrated design method of claim 3, wherein, Collecting network dynamic characteristic data during data transmission and converting it into a byte sequence as the second part of the random number. Using the timestamp during data transmission as the third part of the random number. The network dynamic characteristic data during data transmission is collected and converted into a byte sequence, including: Determining the type of network dynamic characteristic data to be collected and collecting real-time data for each type.

5. The cloud data based kiosk substation intelligent integrated design method of claim 4, wherein, The collected network dynamic characteristic data is converted into a byte representation of a set length. The byte representations of all types of network dynamic characteristic data are combined into a continuous byte sequence. ​ ​ Hashing the combined byte sequence and using the converted byte sequence as part of the random number.

6. The cloud data based integrated design method of a box-type substation intelligence according to any one of claims 1-5, characterized in that, Classifying the data stored in the cloud data warehouse, including: determining the classification standard of the stored data, and adding a classification label to each data item according to the classification standard.

7. The cloud data based integrated design method of a box-type substation according to claim 6, wherein, Structuring the classified data stored in the cloud data warehouse, including: Defining a data model according to the characteristics and classification results of the classified data; Designing a database structure according to the data model, creating corresponding tables and relationships; Performing data cleaning and conversion to match the data with the tables and relationships in the database structure.

Images