Method and system for completing time series data of key equipment in substation based on cloud-edge collaboration

Abstract

The present invention provides a method and system for completing the time series data of key equipment in a substation based on cloud-edge collaboration, the method comprising: using a wireless sensor network detection device to collect the time series data of key equipment in the substation, and uploading the collected time series data to a cloud server; judging the time series data based on the time series data fault detection algorithm of the cloud server to obtain a data judgment result; using the data tensor reconstruction algorithm of the cloud server to perform tensor reconstruction on the time series data and its adjacent time series data that are faulty or missing in the data judgment result to form tensor time series data; constructing a tensor completion algorithm based on the core tensor nuclear norm of tensor decomposition, and using the tensor completion algorithm to complete the missing data of the tensor time series data. The present invention solves the problem that the hard-set core tensor cannot match the size change of the data model in time, thereby combining the constraints of the optimization problem to achieve the minimization calculation of the tensor nuclear norm.

Description

Technical Field

[0001] The present invention relates to the field of data processing technology, and in particular to a method and system for completing time series data of key equipment in a substation based on cloud-edge collaboration. Background Art

[0002] Due to the continuous development of digital power grids, unmanned intelligent substation functions are gradually being realized in substations, and the number and scale of smart substations are increasing day by day.

[0003] As an important hub connecting power systems at all levels, the operating status of the substation will directly affect the power quality of the upstream and downstream. For this reason, various equipment status monitoring systems are installed in the substation, such as the energy management system (EMS). Through SCADA monitoring, WSNs monitoring and other monitoring systems, various monitoring data streams in the substation can be obtained in a timely manner, and then the current operating status of the substation can be evaluated.

[0004] However, in reality, due to the large number of monitoring devices, their wide distribution, complex electromagnetic environment, high maintenance costs, and other reasons, as well as equipment failures, damage to transformers or measuring equipment, channel interference, and storage device failure, the data collected within the substation will be missing or erroneous.

[0005] In order for the substation to operate normally, the subsequent applications of the substation, such as equipment status perception, digital twin, system monitoring, digital transformation, load forecasting and unattended operation, all require complete, accurate and reliable data supply. Therefore, it is of great significance to conduct reliable status estimation of the monitoring data of substation equipment to ensure the reliability and integrity of the data, thereby providing reliable data support for subsequent data applications. Summary of the invention

[0006] Based on this, the purpose of the present invention is to provide a method and system for completing time series data of key equipment in a substation based on cloud-edge collaboration, so as to at least solve the shortcomings of the above-mentioned technology.

[0007] The present invention proposes a method for completing time series data of key equipment in a substation based on cloud-edge collaboration, comprising:

[0008] Use wireless sensor network detection equipment to collect time series data of key equipment in the substation, and upload the collected time series data to the cloud server;

[0009] Judging the time series data based on the time series data fault detection algorithm of the cloud server to obtain a corresponding data judgment result;

[0010] Using the data tensor reconstruction algorithm of the cloud server to reconstruct the tensor of the time series data with faults or missing and its adjacent time series data in the data judgment result to form tensor time series data;

[0011] A tensor completion algorithm based on the core tensor nuclear norm of tensor decomposition is constructed, and the tensor completion algorithm is used to complete missing data of the tensor time series data.

[0012] Furthermore, the steps of using the wireless sensor network detection equipment to collect the time series data of the key equipment in the substation and uploading the collected time series data to the cloud server include:

[0013] Monitoring the time series data of key equipment in the substation using preset monitoring rules and a state sensing device installed in the substation;

[0014] The collected time series data is uploaded via the wireless sensor network so that the cloud server can store and process the time series data.

[0015] Furthermore, the step of judging the time series data based on the time series data fault detection algorithm of the cloud server to obtain the corresponding data judgment result includes:

[0016] Performing fault detection on the time series data based on the time series data fault detection algorithm of the cloud server to determine whether the time series data is missing or faulty;

[0017] If the time series data is judged to be normal data, the normal data is directly applied to the processing of the time series detection data of the substation;

[0018] If the time series data is determined to be missing or faulty, the data tensor reconstruction algorithm of the cloud server is used to reconstruct the missing or faulty data.

[0019] Furthermore, the step of using the data tensor reconstruction algorithm of the cloud server to reconstruct the tensor of the faulty or missing time series data and its adjacent time series data in the data judgment result to form tensor time series data includes:

[0020] Reading the time series data with faults or missing in the judgment result and its adjacent historical time series data;

[0021] The time series data and its adjacent historical time series data are segmented into single-day segments according to date, and the data after the single-day segmentation is reconstructed to form time series data in the form of a third-order tensor.

[0022] Furthermore, the step of using the tensor completion algorithm to complete missing data of the tensor time series data includes:

[0023] A tensor completion model is constructed for the tensor time series data:

[0024] ;

[0025] In the formula, X is a complete third-order tensor after missing data completion, min rank represents the minimum rank, Represents the orthogonal projection operator, which can fill missing data with 0;

[0026] Transform the optimization problem of the tensor completion model into:

[0027] ;

[0028] In the formula, , and represents the weighting factor of the corresponding tensor expansion nuclear norm, , and is a constant greater than 0, and , X 1. X 2 and X 3 are tensors X The corresponding mode-1, mode-2 and mode-3 matrices, is a third-order tensor X The sum of the nuclear norms of each modular expansion matrix, is a third-order core tensor, R represents the real number field, I1, I2 and I3 represent the third-order tensors The dimensions of each order in, U1, U2 and U3 are the factor matrices of the third-order tensor T, r i and c i Respectively represent the matrix U i The number of rows and columns of (i=1, 2, 3), Represents a third-order core tensor Multiply each mode with the corresponding factor matrices U1, U2 and U3;

[0029] An auxiliary variable is introduced to convexly relax the optimization problem of the tensor completion model to obtain a simplified model;

[0030] ;

[0031] In the formula, , and Auxiliary variables , and The nuclear norm of is the Frobenius norm, that is, the square root of the sum of the squares of the absolute values ​​of all elements in the matrix, and λ is the penalty coefficient;

[0032] The simplified model is solved using the augmented Lagrangian algorithm to obtain a partial Lagrangian function:

[0033] ;

[0034] In the formula, represents the Lagrangian function, express and The inner product of (i=1, 2, 3), , and is an auxiliary variable, , and represents the Lagrange tensor multiplier, ω is the penalty coefficient;

[0035] The Lagrangian function is solved using the alternating direction multiplier method:

[0036] ;

[0037] In the formula, represents the variable value that makes the function achieve the minimum value, ω and μ are penalty coefficients;

[0038] The factor matrix of the Lagrangian function is updated using the Tucker decomposition algorithm in the pre-rank calculation-assisted HOOI algorithm:

[0039] ;

[0040] An optimal completion tensor is obtained, and the original missing data in the tensor time series data is completed using the optimal completion tensor.

[0041] The present invention also proposes a substation key equipment time series data completion system based on cloud-edge collaboration, including:

[0042] The data acquisition module is used to collect the time series data of key equipment in the substation using wireless sensor network detection equipment, and upload the collected time series data to the cloud server;

[0043] A data judgment module, used to judge the time series data based on the time series data fault detection algorithm of the cloud server to obtain a corresponding data judgment result;

[0044] A tensor reconstruction module, used to use the data tensor reconstruction algorithm of the cloud server to perform tensor reconstruction on the time series data with faults or missing and its adjacent time series data in the data judgment result to form tensor time series data;

[0045] The data completion module is used to construct a tensor completion algorithm based on the core tensor nuclear norm of tensor decomposition, and use the tensor completion algorithm to complete missing data of the tensor time series data.

[0046] Furthermore, the data acquisition module includes:

[0047] A data monitoring unit, used to monitor the time series data of key equipment in the substation using preset monitoring rules and a state sensing device installed in the substation;

[0048] The data uploading unit is used to upload the collected time series data through the wireless sensor network so that the cloud server can store and process the time series data.

[0049] Furthermore, the data judgment module includes:

[0050] A fault detection unit, configured to perform fault detection on the time series data based on the time series data fault detection algorithm of the cloud server to determine whether the time series data is missing or faulty;

[0051] A first processing unit, configured to directly apply the normal data to the processing of the time series detection data of the substation if the time series data is judged to be normal data;

[0052] The second processing unit is used to reconstruct the missing or faulty data using the data tensor reconstruction algorithm of the cloud server if the time series data is judged to be missing or faulty.

[0053] Furthermore, the tensor reconstruction module includes:

[0054] A data reading unit, used to read the time series data with faults or missing in the judgment result and its adjacent historical time series data;

[0055] The tensor reconstruction unit is used to segment the time series data and its adjacent historical time series data into single-day segments according to date, and reconstruct the data after the single-day segmentation to form time series data in the form of a third-order tensor.

[0056] The present invention also proposes a readable storage medium on which a computer program is stored. When the program is executed by a processor, it implements the above-mentioned method for completing the timing data of key equipment of a substation based on cloud-edge collaboration.

[0057] The present invention also proposes a computer, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, the method for completing the timing data of key equipment of a substation based on cloud-edge collaboration is implemented.

[0058] The method and system for completing the time series data of key equipment in a substation based on cloud-edge collaboration in the present invention collect the time series data of key equipment in the substation through wireless sensor network detection equipment, and send it to the cloud server, so as to determine whether the collected data is missing or erroneous by using the time series data fault detection algorithm deployed on the cloud server; for the data judged to be abnormal, the abnormal data and its adjacent time series data are tensor reconstructed by using the data tensor reconstruction algorithm to form tensor time series data; the tensor time series data is completed by using the tensor completion algorithm of the core tensor nuclear norm of the pre-rank Tucker decomposition to obtain the optimal completed data; according to tensor theory, the tensor completion algorithm of the core tensor nuclear norm of the pre-rank Tucker decomposition is used to complete the missing data in the time series monitoring data of key equipment in the substation, and the problem of the fixed initialization rank of the core tensor in the Tucker decomposition in CNNT is solved, thereby solving the problem that the hard-set core tensor cannot match the size change of the data model in time, thereby combining the constraints of the optimization problem to achieve the minimization calculation of the tensor nuclear norm. BRIEF DESCRIPTION OF THE DRAWINGS

[0059] Figure 1 It is a flow chart of a method for completing time series data of key equipment in a substation based on cloud-edge collaboration in the first embodiment of the present invention;

[0060] Figure 2 for Figure 1 Detailed flow chart of step S101;

[0061] Figure 3 It is a schematic diagram of a device for collecting key timing information of substation equipment in a first embodiment of the present invention;

[0062] Figure 4 for Figure 1 Detailed flow chart of step S102;

[0063] Figure 5 is a flow chart of a time series data fault detection algorithm in the first embodiment of the present invention;

[0064] Figure 6 for Figure 1 Detailed flow chart of step S103;

[0065] Figure 7 Flow chart of a tensor completion algorithm for the core tensor nuclear norm of the pre-rank Tucker decomposition in the first embodiment of the present invention;

[0066] Figure 8 It is a structural block diagram of a system for completing time series data of key equipment in a substation based on cloud-edge collaboration in a second embodiment of the present invention;

[0067] Fig. 9 FIG. 4 is a structural block diagram of a computer in a third embodiment of the present invention.

[0068] The following specific implementation manner will further illustrate the present invention in conjunction with the above-mentioned drawings. DETAILED DESCRIPTION

[0069] In order to facilitate the understanding of the present invention, the present invention will be described more fully below with reference to the relevant drawings. Several embodiments of the present invention are given in the drawings. However, the present invention can be implemented in many different forms and is not limited to the embodiments described herein. On the contrary, the purpose of providing these embodiments is to make the disclosure of the present invention more thorough and comprehensive.

[0070] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as those commonly understood by those skilled in the art to which the present invention belongs. The terms used herein in the specification of the present invention are only for the purpose of describing specific embodiments and are not intended to limit the present invention. The term "and / or" used herein includes any and all combinations of one or more of the related listed items.

[0071] Embodiment 1

[0072] See also Figure 1 , which shows a method for completing time series data of key equipment in a substation based on cloud-edge collaboration in a first embodiment of the present invention, and the method specifically includes steps S101 to S105:

[0073] S101, using wireless sensor network detection equipment to collect time series data of key equipment in the substation, and uploading the collected time series data to a cloud server;

[0074] For further information, see Figure 2 , the step S101 specifically includes steps S1011~S1012:

[0075] S1011, monitoring the time series data of key equipment in the substation using preset monitoring rules and a state sensing device installed in the substation;

[0076] S1012, uploading the collected time series data via the wireless sensor network so that the cloud server can store and process the time series data.

[0077] For specific implementation, please refer to Figure 3, the operating status of key equipment is monitored by a large number of status sensing units installed in the key equipment of the substation, and the data is transmitted to the cloud server and the substation integrated management system through the wireless sensor network. The substation integrated management system stores the operating status monitoring data of each device in CSV format files.

[0078] S102, judging the time series data based on the time series data fault detection algorithm of the cloud server to obtain a corresponding data judgment result;

[0079] For further information, see Figure 4 , the step S102 specifically includes steps S1021 to S1023:

[0080] S1021, performing fault detection on the time series data based on the time series data fault detection algorithm of the cloud server to determine whether the time series data is missing or faulty;

[0081] S1022, if the time series data is determined to be normal data, directly applying the normal data to the processing of the time series detection data of the substation;

[0082] S1023: If the time series data is judged to be missing or faulty, a data tensor reconstruction algorithm of the cloud server is used to reconstruct the missing or faulty data.

[0083] In specific implementation, Figure 5 As shown, the cloud server deploys a time series data fault detection algorithm to continuously detect faults in the time series data of key equipment and determine whether the monitored time series data of key equipment is missing or faulty;

[0084] Specifically, if the time series data is judged to be normal data, the normal data is directly applied to the subsequent processing and application of substation time series detection data; if the time series data is judged to be faulty or missing, the tensor reconstruction module is used to reconstruct the data tensor of the faulty or missing data.

[0085] S103, using the data tensor reconstruction algorithm of the cloud server to perform tensor reconstruction on the time series data with faults or missing and its adjacent time series data in the data judgment result to form tensor time series data;

[0086] For further information, see Figure 6 , the step S103 specifically includes steps S1031~S1032:

[0087] S1031, reading the time series data with faults or missing in the judgment result and its adjacent historical time series data;

[0088] S1032, segmenting the time series data and its adjacent historical time series data into single-day segments according to date, and reconstructing the data after the single-day segmentation to form time series data in the form of a third-order tensor.

[0089] In the specific implementation, the time series data of detected faults or missing and the historical time series data of the adjacent days stored in the substation integrated management system in CSV format are read, and the time series data is segmented by day to form a For example, 7 days of data is generated. ; In time series data, Represents the single monitoring type of the monitoring unit, for example: Indicates that the monitoring unit monitors 18 types of data at a time; represents the number of monitoring times in a single day, for example Representatives monitored 96 times in a single day;

[0090] Furthermore, the time series data is reconstructed by the above-mentioned day, monitoring type, and monitoring number, and the reconstructed data forms a third-order tensor time series data, which is expressed as , R is the field of real numbers.

[0091] S104, constructing a tensor completion algorithm based on the core tensor nuclear norm of tensor decomposition, and using the tensor completion algorithm to complete missing data of the tensor time series data.

[0092] In specific implementation, Figure 7 As shown, based on the above tensor timing data Construct a tensor completion model. The data content in the tensor time series data is missing. Assume that Ω represents the location index set of known data, that is, . Then the missing data position index set in the tensor model can be used Represents. Let the tensor For the complete third-order tensor after missing data completion, its low-rank property is completed, expressed as:

[0093] ;

[0094] In the formula, min rank represents the minimum rank, Represents the orthogonal projection operator, which can fill missing data with 0:

[0095] ;

[0096] In the formula, X dct Representing a tensor X The observation value of any point in the location index collection Ω of known data, Representing a tensorX The projection operator of the observed value of any point in the known data position index collection Ω.

[0097] in, Requesting Tensors T and tensors X The observed values ​​are equal, that is, , satisfying the above two formulas represents the tensorized time series data T Satisfying low-rank validity; then the nuclear norm can be minimized to approximate the solution To obtain the rank of the tensorized time series data T:

[0098] ;

[0099] In the formula, represents the nuclear norm of the tensor T, , and represents the weighting factor corresponding to the expanded nuclear norm, , and is a constant greater than 0, and , T1, T2 and T3 are tensors T The corresponding mode-1, mode-2 and mode-3 matrices, , and are the nuclear norms of matrices T1, T2, and T3, respectively, Represents a third-order tensor T The sum of the nuclear norms of the matrices expanded in each module.

[0100] Third-order tensor constructed for time series data ,So The core tensor norm of can be defined as , so the optimization problem can be transformed into:

[0101] ;

[0102] In the formula, , and represents the weighting factor of the corresponding tensor expansion nuclear norm, , and is a constant greater than 0, and , X 1. X 2 and X 3 are tensors X The corresponding mode-1, mode-2 and mode-3 matrices, is a third-order tensorX The sum of the nuclear norms of each modular expansion matrix, is a third-order core tensor, R represents the real number field, I1, I2 and I3 represent the third-order tensors The dimensions of each order in, U1, U2 and U3 are the factor matrices of the third-order tensor T, r i and c i Respectively represent the matrix U i The number of rows and columns of (i=1, 2, 3), Represents a third-order core tensor Multiply each mode with the corresponding factor matrices U1, U2 and U3;

[0103] Furthermore, auxiliary variables are introduced to perform convex relaxation on the optimization problem of the tensor completion model to obtain a simplified model;

[0104] ;

[0105] In the formula, , and Auxiliary variables , and The nuclear norm of is the Frobenius norm, that is, the square root of the sum of the squares of the absolute values ​​of all elements in the matrix, and λ is the penalty coefficient;

[0106] The simplified model is solved using the augmented Lagrangian algorithm to obtain a partial Lagrangian function:

[0107] ;

[0108] In the formula, represents the Lagrangian function, express and The inner product of (i=1, 2, 3), , and is an auxiliary variable, , and represents the Lagrange tensor multiplier, ω is the penalty coefficient;

[0109] Use the Alternating Direction Method of Multipliers (ADMM) to solve , , , , , , and T, in order to reduce the computational cost, this embodiment will Set to 1 / 3, that is, the fixed weight parameters are equal. For the penalty coefficient ω, set the initial value of the penalty coefficient ω to 10-5, and the update method is , the maximum value =105.

[0110] Specifically, for , we only need to update the following smoothly differentiable subproblems:

[0111] ;

[0112] In the formula, arg min represents the variable value that makes the function take the minimum value, and μ is the penalty coefficient.

[0113] Specifically, for , the singular value shrinkage algorithm (Singular Value Thresholding, SVT) can be used to solve the following formula:

[0114] ;

[0115] In the formula, yes Auxiliary variables when obtaining the minimum value The value of (i=1, 2, 3), , and They are , and The weighting factor corresponding to the expansion norm, is the penalty coefficient.

[0116] Specifically, for , that is, solve:

[0117] ;

[0118] In the formula, yes The value of the tensor T when the minimum value is obtained, is the core tensor, and U1, U2, and U3 are factor matrices.

[0119] Specifically, for , using the Tucker decomposition algorithm based on pre-rank calculation to assist in calculating the core tensor The specific calculation steps of the rank R can be based on the calculation steps of the HOSVD algorithm, and the rank size under the mode is obtained by comparing the ratio of the sum of the first i singular values ​​of a tensor X generated in different modes to the sum of all singular values ​​and the size of the set threshold U, and finally the better rank R is obtained:

[0120] Pseudocode of Pre-R algorithm

[0121]

[0122] Finally, the rank R of each mode obtained by the Pre-R algorithm is substituted into the Tucker decomposition of the high-order orthogonal iteration (HOOI) to update the above factor matrices U1, U2 and U3:

[0123]

[0124] Furthermore, an optimal completion tensor is obtained, and the optimal completion tensor is used to complete the original missing data in the tensor time series data, and the completed data is transmitted back to the substation integrated management system.

[0125] In summary, the method for completing the time series data of key equipment in a substation based on cloud-edge collaboration in the above-mentioned embodiments of the present invention collects the time series data of key equipment in the substation through a wireless sensor network detection device, and sends it to a cloud server, so as to determine whether the collected data is missing or erroneous by using a time series data fault detection algorithm deployed on the cloud server; for data judged to be abnormal, a data tensor reconstruction algorithm is used to perform tensor reconstruction on the abnormal data and its adjacent time series data to form tensor time series data; the tensor time series data is completed by a tensor completion algorithm of the core tensor nuclear norm of the pre-rank Tucker decomposition to obtain the optimal completed data; according to tensor theory, a tensor completion algorithm of the core tensor nuclear norm of the pre-rank Tucker decomposition is used to complete the missing data in the time series monitoring data of key equipment in the substation, and the problem of fixed initialization rank of the core tensor in the Tucker decomposition in CNNT is solved, so as to solve the problem that the hard-set core tensor cannot match the size change of the data model in time, thereby combining the constraints of the optimization problem to achieve the minimization calculation of the tensor nuclear norm.

[0126] Embodiment 2

[0127] On the other hand, the present invention also proposes a substation key equipment time series data completion system based on cloud-edge collaboration, please refer to Figure 8 , shown is a substation key equipment time series data completion system based on cloud-edge collaboration in the second embodiment of the present invention, the system includes:

[0128] The data acquisition module 11 is used to collect the time series data of key equipment in the substation using the wireless sensor network detection equipment, and upload the collected time series data to the cloud server;

[0129] Furthermore, the data acquisition module 11 includes:

[0130] A data monitoring unit, used to monitor the time series data of key equipment in the substation using preset monitoring rules and a state sensing device installed in the substation;

[0131] The data uploading unit is used to upload the collected time series data through the wireless sensor network so that the cloud server can store and process the time series data.

[0132] A data judgment module 12, configured to judge the time series data based on the time series data fault detection algorithm of the cloud server to obtain a corresponding data judgment result;

[0133] Furthermore, the data judgment module 12 includes:

[0134] A fault detection unit, configured to perform fault detection on the time series data based on the time series data fault detection algorithm of the cloud server to determine whether the time series data is missing or faulty;

[0135] A first processing unit, configured to directly apply the normal data to the processing of the time series detection data of the substation if the time series data is judged to be normal data;

[0136] The second processing unit is used to reconstruct the missing or faulty data using the data tensor reconstruction algorithm of the cloud server if the time series data is judged to be missing or faulty.

[0137] A tensor reconstruction module 13 is used to use the data tensor reconstruction algorithm of the cloud server to perform tensor reconstruction on the time series data with faults or missing and its adjacent time series data in the data judgment result to form tensor time series data;

[0138] Furthermore, the tensor reconstruction module 13 includes:

[0139] A data reading unit, used to read the time series data with faults or missing in the judgment result and its adjacent historical time series data;

[0140] The tensor reconstruction unit is used to segment the time series data and its adjacent historical time series data into single-day segments according to date, and reconstruct the data after the single-day segmentation to form time series data in the form of a third-order tensor.

[0141] The data completion module 14 is used to construct a tensor completion algorithm based on the core tensor nuclear norm of tensor decomposition, and use the tensor completion algorithm to complete missing data of the tensor time series data.

[0142] The functions or operation steps implemented when the above modules and units are executed are generally the same as those in the above method embodiments, and will not be repeated here.

[0143] The implementation principle and technical effects of the time series data completion system for key equipment of substations based on cloud-edge collaboration provided in the embodiment of the present invention are the same as those of the aforementioned method embodiment. For the sake of brief description, for matters not mentioned in the system embodiment, reference may be made to the corresponding contents in the aforementioned method embodiment.

[0144] Embodiment 3

[0145] The present invention also provides a computer, see Fig. 9 , shown is a computer in the third embodiment of the present invention, including a memory 10, a processor 20, and a computer program 30 stored in the memory 10 and executable on the processor 20. When the processor 20 executes the computer program 30, the above-mentioned method for completing the time series data of key equipment of a substation based on cloud-edge collaboration is implemented.

[0146] The memory 10 includes at least one type of readable storage medium, which includes a flash memory, a hard disk, a multimedia card, a card-type memory (e.g., an SD or DX memory, etc.), a magnetic memory, a magnetic disk, an optical disk, etc. In some embodiments, the memory 10 may be an internal storage unit of a computer, such as a hard disk of the computer. In other embodiments, the memory 10 may also be an external storage device, such as a plug-in hard disk, a smart memory card (Smart Media Card, SMC), a secure digital (Secure Digital, SD) card, a flash card, etc. Further, the memory 10 may also include both an internal storage unit of the computer and an external storage device. The memory 10 may be used not only to store application software and various types of data installed in the computer, but also to temporarily store data that has been output or is to be output.

[0147] Among them, in some embodiments, the processor 20 can be an electronic control unit (Electronic Control Unit, abbreviated as ECU, also known as a vehicle computer), a central processing unit (Central Processing Unit, CPU), a controller, a microcontroller, a microprocessor or other data processing chip, used to run the program code stored in the memory 10 or process data, such as executing access restriction programs, etc.

[0148] It should be pointed out that Fig. 9 The structure shown does not constitute a limitation on the computer. In other embodiments, the computer may include fewer or more components than shown in the figure, or combine certain components, or arrange the components differently.

[0149] An embodiment of the present invention also proposes a readable storage medium on which a computer program is stored. When the program is executed by a processor, it implements the method for completing the timing data of key equipment of a substation based on cloud-edge collaboration as described above.

[0150] Those skilled in the art will appreciate that the logic and / or steps represented in the flowchart or otherwise described herein, for example, may be considered as an ordered list of executable instructions for implementing logical functions, and may be specifically implemented in any computer-readable medium for use by an instruction execution system, device or apparatus (such as a computer-based system, a system including a processor, or other system that can fetch instructions from an instruction execution system, device or apparatus and execute instructions), or in conjunction with such instruction execution systems, devices or apparatuses. For purposes of this specification, "computer-readable medium" may be any device that can contain, store, communicate, propagate or transmit a program for use by an instruction execution system, device or apparatus, or in conjunction with such instruction execution systems, devices or apparatuses.

[0151] More specific examples of computer-readable media (a non-exhaustive list) include the following: an electrical connection with one or more wires (electronic device), a portable computer disk case (magnetic device), a random access memory (RAM), a read-only memory (ROM), an erasable and programmable read-only memory (EPROM or flash memory), an optical fiber device, and a portable compact disk read-only memory (CDROM). In addition, the computer-readable medium may even be a paper or other suitable medium on which the program is printed, since the program may be obtained electronically, for example, by optically scanning the paper or other medium, followed by editing, deciphering or, if necessary, processing in another suitable manner, and then stored in a computer memory.

[0152] It should be understood that the various parts of the present invention can be implemented by hardware, software, firmware or a combination thereof. In the above-mentioned embodiments, multiple steps or methods can be implemented by software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if implemented by hardware, as in another embodiment, it can be implemented by any one of the following technologies known in the art or a combination thereof: a discrete logic circuit having a logic gate circuit for implementing a logic function for a data signal, a dedicated integrated circuit having a suitable combination of logic gate circuits, a programmable gate array (PGA), a field programmable gate array (FPGA), etc.

[0153] The technical features of the above-described embodiments may be arbitrarily combined. To make the description concise, not all possible combinations of the technical features in the above-described embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0154] The above-mentioned embodiments only express several implementation methods of the present application, and the descriptions thereof are relatively specific and detailed, but they cannot be understood as limiting the scope of the invention patent. It should be pointed out that, for a person of ordinary skill in the art, several variations and improvements can be made without departing from the concept of the present application, and these all belong to the protection scope of the present application. Therefore, the protection scope of the patent of the present application shall be subject to the attached claims.

Claims

1. A method for completing time series data of key equipment in a substation based on cloud-edge collaboration, characterized in that: include: Use wireless sensor network detection equipment to collect time series data of key equipment in the substation, and upload the collected time series data to the cloud server; Judging the time series data based on the time series data fault detection algorithm of the cloud server to obtain a corresponding data judgment result; Using the data tensor reconstruction algorithm of the cloud server to reconstruct the tensor of the time series data with faults or missing and its adjacent time series data in the data judgment result to form tensor time series data; A tensor completion algorithm based on the core tensor nuclear norm of tensor decomposition is constructed, and the tensor completion algorithm is used to complete missing data of the tensor time series data, wherein the step of using the tensor completion algorithm to complete missing data of the tensor time series data includes: A tensor completion model is constructed for the tensor time series data: ； In the formula, X is a complete third-order tensor after missing data completion, min rank represents the minimum rank, Represents the orthogonal projection operator, which can fill missing data with 0; Transform the optimization problem of the tensor completion model into: ； In the formula, , and represents the weighting factor of the corresponding tensor expansion nuclear norm, , and is a constant greater than 0, and , X 1. X 2 and X 3 are tensors X The corresponding mode-1, mode-2 and mode-3 matrices, is a third-order tensor X The sum of the nuclear norms of each modular expansion matrix, is a third-order core tensor, R represents the real number field, I1, I2 and I3 represent the third-order tensors The dimensions of each order in, U1, U2 and U3 are the factor matrices of the third-order tensor T, r i and c i Respectively represent the matrix U i The number of rows and columns of the matrix U i i=1, 2, 3, Represents a third-order core tensor Multiply each mode with the corresponding factor matrices U1, U2 and U3; An auxiliary variable is introduced to convexly relax the optimization problem of the tensor completion model to obtain a simplified model; ； In the formula, , and Auxiliary variables , and The nuclear norm of is the Frobenius norm, that is, the square root of the sum of the squares of the absolute values ​​of all elements in the matrix, and λ is the penalty coefficient; The simplified model is solved using the augmented Lagrangian algorithm to obtain a partial Lagrangian function: ； In the formula, represents the Lagrangian function, express and , where i=1, 2, 3, , and is an auxiliary variable, , and represents the Lagrange tensor multiplier, ω is the penalty coefficient; The Lagrangian function is solved using the alternating direction multiplier method: ； In the formula, represents the variable value that makes the function achieve the minimum value, ω and μ are penalty coefficients; The factor matrix of the Lagrangian function is updated using the Tucker decomposition algorithm in the pre-rank calculation-assisted HOOI algorithm: ； An optimal completion tensor is obtained, and the original missing data in the tensor time series data is completed using the optimal completion tensor.

2. The method for completing time series data of key substation equipment based on cloud-edge collaboration according to claim 1 is characterized in that: The steps of using wireless sensor network detection equipment to collect time series data of key equipment in the substation and uploading the collected time series data to the cloud server include: Monitoring the time series data of key equipment in the substation using preset monitoring rules and a state sensing device installed in the substation; The collected time series data is uploaded via the wireless sensor network so that the cloud server can store and process the time series data.

3. The method for completing time series data of key substation equipment based on cloud-edge collaboration according to claim 1 is characterized in that: The step of judging the time series data based on the time series data fault detection algorithm of the cloud server to obtain the corresponding data judgment result includes: Performing fault detection on the time series data based on the time series data fault detection algorithm of the cloud server to determine whether the time series data is missing or faulty; If the time series data is judged to be normal data, the normal data is directly applied to the processing of the time series detection data of the substation; If the time series data is determined to be missing or faulty, the data tensor reconstruction algorithm of the cloud server is used to reconstruct the missing or faulty data.

4. The method for completing time series data of key substation equipment based on cloud-edge collaboration according to claim 1 is characterized in that: The step of using the data tensor reconstruction algorithm of the cloud server to reconstruct tensors of the faulty or missing time series data and its adjacent time series data in the data judgment result to form tensor time series data includes: Reading the time series data with faults or missing in the judgment result and its adjacent historical time series data; The time series data and its adjacent historical time series data are segmented into single-day segments according to date, and the data after the single-day segmentation is reconstructed to form time series data in the form of a third-order tensor.

5. A system for completing the time series data of key equipment in a substation based on cloud-edge collaboration, characterized in that: include: The data acquisition module is used to collect the time series data of key equipment in the substation using wireless sensor network detection equipment, and upload the collected time series data to the cloud server; A data judgment module, used to judge the time series data based on the time series data fault detection algorithm of the cloud server to obtain a corresponding data judgment result; A tensor reconstruction module, used to use the data tensor reconstruction algorithm of the cloud server to perform tensor reconstruction on the time series data with faults or missing and its adjacent time series data in the data judgment result to form tensor time series data; A data completion module is used to construct a tensor completion algorithm based on the core tensor nuclear norm of tensor decomposition, and use the tensor completion algorithm to complete missing data of the tensor time series data, wherein the data completion module is used to: A tensor completion model is constructed for the tensor time series data: ； In the formula, X is a complete third-order tensor after missing data completion, min rank represents the minimum rank, Represents the orthogonal projection operator, which can fill missing data with 0; Transform the optimization problem of the tensor completion model into: ； In the formula, , and represents the weighting factor of the corresponding tensor expansion nuclear norm, , and is a constant greater than 0, and , X 1. X 2 and X 3 are tensors X The corresponding mode-1, mode-2 and mode-3 matrices, is a third-order tensor X The sum of the nuclear norms of each modular expansion matrix, is a third-order core tensor, R represents the real number field, I1, I2 and I3 represent the third-order tensors The dimensions of each order in, U1, U2 and U3 are the factor matrices of the third-order tensor T, r i and c i Respectively represent the matrix U i The number of rows and columns of the matrix U i i=1, 2, 3, Represents a third-order core tensor Multiply each mode with the corresponding factor matrices U1, U2 and U3; An auxiliary variable is introduced to convexly relax the optimization problem of the tensor completion model to obtain a simplified model; ； In the formula, , and Auxiliary variables , and The nuclear norm of is the Frobenius norm, that is, the square root of the sum of the squares of the absolute values ​​of all elements in the matrix, and λ is the penalty coefficient; The simplified model is solved using the augmented Lagrangian algorithm to obtain a partial Lagrangian function: ； In the formula, represents the Lagrangian function, express and The inner product of, where i=1, 2, 3, , and is an auxiliary variable, , and represents the Lagrange tensor multiplier, ω is the penalty coefficient; The Lagrangian function is solved using the alternating direction multiplier method: ； In the formula, represents the variable value that makes the function achieve the minimum value, ω and μ are penalty coefficients; The factor matrix of the Lagrangian function is updated using the Tucker decomposition algorithm in the pre-rank calculation-assisted HOOI algorithm: ； An optimal completion tensor is obtained, and the original missing data in the tensor time series data is completed using the optimal completion tensor.

6. The substation key equipment time series data completion system based on cloud-edge collaboration according to claim 5 is characterized in that: The data acquisition module comprises: A data monitoring unit, used to monitor the time series data of key equipment in the substation using preset monitoring rules and a state sensing device installed in the substation; The data uploading unit is used to upload the collected time series data through the wireless sensor network so that the cloud server can store and process the time series data.

7. The substation key equipment time series data completion system based on cloud-edge collaboration according to claim 5 is characterized in that: The data judgment module comprises: A fault detection unit, configured to perform fault detection on the time series data based on the time series data fault detection algorithm of the cloud server to determine whether the time series data is missing or faulty; A first processing unit, configured to directly apply the normal data to the processing of the time series detection data of the substation if the time series data is judged to be normal data; The second processing unit is used to reconstruct the missing or faulty data using the data tensor reconstruction algorithm of the cloud server if the time series data is judged to be missing or faulty.

8. A readable storage medium having a computer program stored thereon, characterized in that: When the program is executed by a processor, it implements the method for completing the timing data of key equipment in a substation based on cloud-edge collaboration as described in any one of claims 1 to 4.

9. A computer comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the computer program, it implements the method for completing the timing data of key equipment of a substation based on cloud-edge collaboration as described in any one of claims 1 to 4.

Images