Digitized supervision quality evaluation method based on big data model and storage medium

Abstract

The embodiment of the invention provides a digital supervision quality evaluation method based on a big data model and a storage medium, and belongs to the technical field of supervision quality evaluation. The evaluation method comprises the following steps: constructing an evaluation grade of a supervision model; obtaining an output time sequence characteristic and a real time sequence characteristic of the current supervision model; according to the method, the error time sequence characteristics of the current supervision model are obtained through the output time sequence characteristics and the real time sequence characteristics of the current supervision model, the evaluation matrix of the current supervision model can be obtained by combining the evaluation level of the constructed supervision model, and the evaluation value of the supervision model can be obtained according to the evaluation matrix. The current supervision model can be evaluated by using the evaluation value; by adopting the mode of quantifying the error time sequence characteristics of the supervision model into the specific evaluation values, effective evaluation, namely effective supervision, of the supervision model can be realized, and the reliability and the accuracy of power grid system supervision are further guaranteed.

Description

Technical Field

[0001] The present invention relates to the technical field of supervision quality evaluation, and in particular to a digital supervision quality evaluation method and storage medium based on a big data model. Background Art

[0002] The digital supervision model is based on digital technology, especially information technology and intelligent technology, and is applied to the field of power grid engineering supervision or power grid operation supervision. Traditional supervision mainly relies on manual or related supervision tools. The digital supervision model can improve the efficiency, accuracy and transparency of supervision through digital tools and systems, and has been applied by the power grid system. Specifically, the types of supervision models can include voltage supervision model, construction quality supervision model, etc.

[0003] At present, with the gradual application of supervision models, there is no standard to measure the supervision quality of the supervision models, which makes the optimization and iteration effect of the supervision models poor, thus affecting the reliability and accuracy of the supervision of the power grid system.

[0004] The inventors of the present application discovered during the process of implementing the present invention that the above-mentioned solutions in the prior art have the defects of poor reliability and low accuracy in monitoring the power grid system. Summary of the invention

[0005] The purpose of the embodiments of the present invention is to provide a digital supervision quality evaluation method and storage medium based on a big data model, which have the function of improving the reliability and accuracy of power grid system supervision.

[0006] In order to achieve the above object, an embodiment of the present invention provides a method for evaluating the quality of digital supervision based on a big data model, comprising:

[0007] Construct the evaluation level of the supervision model;

[0008] Obtain the output timing characteristics and real timing characteristics of the current supervision model;

[0009] Obtaining error timing characteristics of the current supervision model according to the output timing characteristics of the current supervision model and the real timing characteristics;

[0010] Acquire an evaluation matrix according to the error timing characteristics and the evaluation level;

[0011] Obtaining an evaluation value of the current supervision model according to the evaluation matrix;

[0012] The quality of the current supervision model is evaluated according to the evaluation value of the current supervision model.

[0013] Optionally, the evaluation levels for constructing the supervision model include:

[0014] Obtaining multiple historical output time series features of the supervision model and corresponding historical real time series features;

[0015] Acquire historical error timing characteristics according to the historical output timing characteristics and the historical real timing characteristics;

[0016] Obtaining the mean, maximum and minimum values ​​of the historical error time series characteristics;

[0017] The evaluation level of the supervision model is divided according to the mean value, maximum value and minimum value of the historical error time series characteristics.

[0018] Optionally, acquiring the historical error timing characteristics according to the historical output timing characteristics and the historical real timing characteristics includes:

[0019] According to formula (1), the historical error time series characteristics are obtained:

[0020]

[0021] in, is the value of the i-th historical error time series feature at time t, is the value of the i-th historical output time series feature at time t, is the value of the i-th historical real time series feature at time t, i is an integer number, and t is the time.

[0022] Optionally, dividing the evaluation level of the supervision model according to the mean value, maximum value and minimum value of the historical error time series characteristics includes:

[0023] According to formula (2), the level base value of the historical error time series feature is obtained:

[0024]

[0025] Among them, Δε is the level base value of the historical error time series feature, ε max is the maximum value of the historical error time series characteristic, ε min is the minimum value of the historical error time series characteristic, and σ is the interval base value;

[0026] According to formula (3), the level value of the historical error time series feature is obtained:

[0027]

[0028] Among them, ε l is the level value of the historical error time series feature, l is the level sequence number of the historical error time series feature, l={1,2}, is the mean value of the historical error time series characteristics;

[0029] Determine whether there is the error time series feature between the current level value of the historical error time series feature and its closest maximum or minimum value;

[0030] In the case where it is determined that there is an error timing feature between the level value of the current historical error timing feature and its closest maximum value or minimum value, obtaining the level value of the current historical error timing feature;

[0031] When it is determined that there is no error timing feature between the level value of the current historical error timing feature and its closest maximum or minimum value, the interval base value is updated according to formula (4):

[0032] σ=σ+δ, (4)

[0033] Among them, δ is the adjustment base value;

[0034] Return to the step of obtaining the level base value of the historical error time series feature according to formula (2).

[0035] Optionally, obtaining the error time series characteristics of the current supervision model according to the output time series characteristics and the real time series characteristics of the current supervision model includes:

[0036] According to formula (5), the error time series characteristics of the current supervision model are obtained:

[0037]

[0038] in, is the value of the error time series characteristic of the current supervision model at time t, Y t d is the value of the output time series characteristic of the current supervision model at time t, is the value of the real time series characteristics of the current supervision model at time t.

[0039] Optionally, acquiring an evaluation matrix according to the error timing characteristics and the evaluation level includes:

[0040] Classifying the error time series characteristics of the current supervision model according to the evaluation level;

[0041] According to formula (6), the evaluation probability of the error time series characteristics of the current supervision model at each level is obtained:

[0042]

[0043] Among them, p kis the evaluation probability of the error time series characteristics of the current supervision model at level k, k is the level interval, and k = {1, 2, 3}, T n2 k is the end time of the nth segment of the error time series characteristic of the current supervision model at level k, T n1 k is the start time of the nth segment of the error timing characteristics of the current supervision model at level k, n is an integer number, j is an integer number, N is the total number of segments of the error timing characteristics of the current supervision model at level k, and T is the total duration of the error timing characteristics of the current supervision model;

[0044] The evaluation matrix of the current supervision model is obtained according to the evaluation probability of the error time series characteristics of the current supervision model at each level.

[0045] Optionally, obtaining the evaluation value of the current supervision model according to the evaluation matrix includes:

[0046] Obtain the decision value of the current supervision model at each level and construct a decision matrix;

[0047] According to formula (7), the evaluation value of the current supervision model is obtained:

[0048] S=P 3×1 ×Q 1×3 , (7)

[0049] Among them, S is the evaluation value of the current supervision model, P 3×1 is the evaluation matrix of the current supervision model, Q 1×3 is the decision matrix.

[0050] Optionally, obtaining the decision value of the current supervision model at each level includes:

[0051] Randomize the decision value of the current supervision model at each level and construct an initial population;

[0052] According to formula (8), the fitness value of each individual in the population is obtained:

[0053]

[0054] Among them, F fitness is the fitness value, K is the number of levels of the supervision model, and K=3, q k is the decision value of the error time series characteristic of the supervision model at level k, p k·l is the evaluation probability of the historical error time series feature at level k;

[0055] Selecting the best individual in the population according to the fitness value;

[0056] Iterate the remaining individuals in the population according to the optimal individual;

[0057] Determine whether the number of iterations of the current population is greater than or equal to a threshold number of iterations;

[0058] When it is determined that the number of iterations of the current population is greater than or equal to the iteration number threshold, outputting the optimal solution of the decision value of the current supervision model;

[0059] When it is determined that the current number of iterations of the population is less than the iteration number threshold, return to the step of obtaining the fitness value of each individual in the population according to formula (8).

[0060] Optionally, evaluating the quality of the current supervision model according to the evaluation value of the current supervision module includes:

[0061] Determine whether the evaluation value of the current supervision model is less than or equal to the evaluation threshold;

[0062] In the case where it is determined that the evaluation value of the current supervision model is less than or equal to the evaluation threshold, determining that the supervision quality of the current supervision model is good;

[0063] When it is determined that the evaluation value of the current supervision model is greater than the evaluation threshold, it is determined that the supervision quality of the current supervision model is poor.

[0064] On the other hand, the present invention further provides a computer-readable storage medium, wherein the computer-readable storage medium stores instructions, wherein the instructions are used to be read by a machine so that the machine executes any of the evaluation methods described above.

[0065] Through the above technical scheme, the digital supervision quality evaluation method and storage medium based on the big data model provided by the present invention obtain the error timing characteristics of the current supervision model through the output timing characteristics and real timing characteristics of the current supervision model, and combine the evaluation level of the constructed supervision model to obtain the evaluation matrix of the current supervision model. According to the evaluation matrix, the evaluation value of the supervision model can be obtained, and then the current supervision model can be evaluated by using the evaluation value; by adopting this method of quantifying the error timing characteristics of the supervision model into specific evaluation values, it is possible to achieve effective evaluation of the supervision model, that is, effective supervision, thereby ensuring the reliability and accuracy of power grid system supervision.

[0066] Other features and advantages of the embodiments of the present invention will be described in detail in the subsequent detailed description. BRIEF DESCRIPTION OF THE DRAWINGS

[0067] The accompanying drawings are used to provide a further understanding of the embodiments of the present invention and constitute a part of the specification. Together with the following specific implementations, they are used to explain the embodiments of the present invention, but do not constitute a limitation on the embodiments of the present invention. In the accompanying drawings:

[0068] Figure 1 is a flow chart of a method for evaluating digital supervision quality based on a big data model according to an embodiment of the present invention;

[0069] Figure 2 It is a flow chart for obtaining evaluation grades in a method for evaluating digital supervision quality based on a big data model according to an embodiment of the present invention;

[0070] Figure 3 It is a flow chart of obtaining a grade value in a method for evaluating digital supervision quality based on a big data model according to an embodiment of the present invention;

[0071] Figure 4 It is a flow chart of obtaining an evaluation matrix in a method for evaluating digital supervision quality based on a big data model according to an embodiment of the present invention;

[0072] Figure 5 It is a flow chart of obtaining evaluation values ​​in a method for evaluating digital supervision quality based on a big data model according to an embodiment of the present invention;

[0073] Figure 6 It is a flowchart of obtaining a decision matrix in a method for evaluating digital supervision quality based on a big data model according to an embodiment of the present invention;

[0074] Figure 7 It is a flowchart of evaluating the supervision model in the evaluation method of digital supervision quality based on the big data model according to one embodiment of the present invention. DETAILED DESCRIPTION

[0075] The specific implementation of the embodiment of the present invention is described in detail below in conjunction with the accompanying drawings. It should be understood that the specific implementation described here is only used to illustrate and explain the embodiment of the present invention, and is not used to limit the embodiment of the present invention.

[0076] It should be noted that the acquisition, transmission, storage, use, and processing of data in the technical solution of this application are in compliance with the relevant provisions of national laws and regulations. In the embodiments of this application, some existing solutions in the industry such as certain software, components, and models may be mentioned, which should be considered as exemplary. Their purpose is only to illustrate the feasibility of implementing the technical solution of this application, but it does not mean that the applicant has or will necessarily use the solution.

[0077] Figure 1is a flowchart of a method for evaluating digital supervision quality based on a big data model according to an embodiment of the present invention. Figure 1 The evaluation method may include:

[0078] In step S10, the evaluation level of the supervision model is constructed. The evaluation level of the supervision model can be determined according to the output error of the supervision model in the historical data, so that the multiple evaluation levels divided into the supervision model can be more accurate and effective.

[0079] In step S11, the output timing characteristics and real timing characteristics of the current supervision model are obtained. Among them, for the output timing characteristics of the current supervision model, the output time period can be pre-set to obtain the output data of the supervision model within the time period. Specifically, the timing characteristics can be discrete or continuous. If it is a discrete characteristic, it is necessary to obtain the output values ​​at equal time intervals. The time interval can be set smaller to ensure that there are relatively more discrete data. Similarly, the real timing characteristics of the supervision model, that is, the real timing characteristics of the supervised model, need to correspond to the output timing characteristics when collected.

[0080] In step S12, the error time series characteristics of the current supervision model are obtained according to the output time series characteristics and the real time series characteristics of the current supervision model. After obtaining the output time series characteristics and the real time series characteristics of the supervision model, the error time series characteristics of the supervision model can be obtained. Specifically, the error time series characteristics of the supervision model can be obtained according to formula (5):

[0081] According to formula (5), the error time series characteristics of the current supervision model are obtained:

[0082]

[0083] in, is the value of the error time series characteristic of the current supervision model at time t, Y t d is the value of the output time series feature of the current supervision model at time t, is the value of the real time series characteristics of the current supervision model at time t.

[0084] In step S13, an evaluation matrix is ​​obtained according to the error time series characteristics and the evaluation level. After the error time series characteristics of the current supervision model are obtained, the error time series characteristics can be graded in combination with the evaluation level of the supervision model to obtain the evaluation matrix.

[0085] In step S14, the evaluation value of the current supervision model is obtained according to the evaluation matrix. The evaluation matrix of the current supervision model is transformed to obtain the evaluation value of the current supervision model, which can reflect the output characteristics of the current supervision model over a period of time.

[0086] In step S15, the quality of the current supervision model is evaluated according to the evaluation value of the current supervision model. The quality of the current supervision model can be evaluated according to the evaluation value.

[0087] In step S10 to step S15, the evaluation level of the supervision model is first constructed based on the historical data of the supervision model, and then the output time series characteristics and the real time series characteristics of the current supervision model are obtained. The error time series characteristics of the current supervision model can be obtained by processing the two. According to the error time series characteristics and the constructed evaluation level, the evaluation matrix of the current supervision model can be obtained. The evaluation matrix is ​​further converted to obtain the evaluation value of the current supervision model, and the quality of the current supervision model can be evaluated according to the evaluation value.

[0088] In the process of supervising the power grid system, the traditional supervision model can only realize digital supervision, but the supervision quality / effect of the supervision model cannot be measured, which affects the reliability and accuracy of the supervision of the power grid system. In this embodiment of the present invention, by quantifying the error time series characteristics of the supervision model into specific evaluation values, it is possible to effectively evaluate the supervision model, that is, effectively supervise it, thereby ensuring the reliability and accuracy of the supervision of the power grid system.

[0089] In this embodiment of the present invention, when the evaluation level of the supervision model is divided, it is necessary to convert it according to the historical data of the supervision model. The specific conversion steps can be as follows: Figure 2 Specifically, Figure 2 In the above, the evaluation method may further include:

[0090] In step S100, multiple historical output time series features and corresponding historical real time series features of the supervision model are obtained. Among them, for the historical output time series features and historical real time series features of the supervision model, multiple groups corresponding to each other can be collected. Specifically, multiple groups of historical output time series features are continuous output time series features separated at equal intervals, that is, an overall time series feature is divided into multiple small time series features. Specifically, a historical output time series feature and a corresponding historical real time series feature can also be used here.

[0091] In step S101, the historical error time series features are obtained according to the historical output time series features and the historical real time series features. The acquisition of the historical error time series features can be shown as formula (1):

[0092] According to formula (1), the historical error time series characteristics are obtained:

[0093]

[0094] in, is the value of the i-th historical error time series feature at time t, is the value of the i-th historical output time series feature at time t, is the value of the i-th historical real time series feature at time t, i is an integer number, and t is the time.

[0095] In step S102, the mean, maximum and minimum values ​​of the historical error time series characteristics are obtained. Among them, for the mean of multiple historical error time series characteristics, the multiple historical error time series characteristics can be merged into a larger overall error time series characteristic to directly obtain the mean, or the mean of each historical error time series characteristic can be obtained, and finally the multiple means are averaged. Similarly, the maximum and minimum values ​​of multiple historical error time series can also be obtained. Specifically, for the acquisition of the maximum and minimum values, the maximum or minimum K values ​​can be obtained and averaged to avoid the influence of abnormal points.

[0096] In step S103, the evaluation level of the supervision model is divided according to the mean, maximum and minimum values ​​of the historical error time series characteristics. The number of evaluation levels can be determined according to the actual evaluation accuracy requirements. In this embodiment of the present invention, two evaluation levels are divided into three evaluation areas. Specifically, Figure 3 As shown, the steps of obtaining the evaluation level may include:

[0097] In step S1030, the level base value of the historical error time series feature is obtained according to formula (2):

[0098]

[0099] Among them, Δε is the level base value of the historical error time series characteristics, ε max is the maximum value of the historical error time series characteristic, ε min is the minimum value of the historical error time series feature, and σ is the interval base value. Specifically, the value of the interval base value σ is generally σ>2, but it cannot be too large, generally within 10, to ensure that the intermediate level area is not too small.

[0100] In step S1031, the level value of the historical error time series feature is obtained according to formula (3):

[0101]

[0102] Among them, ε lis the level value of the historical error time series feature, l is the level sequence number of the historical error time series feature, l={1,2}, is the mean value of the historical error time series feature. Specifically, the number of the level values ​​is two, that is, two values ​​with symmetrical mean values, and the historical error time series feature can be divided into three level areas.

[0103] In step S1032, it is determined whether there is an error time series feature between the level value of the current historical error time series feature and its closest maximum or minimum value, wherein the maximum or minimum value close to the level value includes the maximum level value and the maximum value, and the minimum level value and the minimum value.

[0104] In step S1033, when it is determined that there is an error time series feature between the level value of the current historical error time series feature and its adjacent maximum or minimum value, the level value of the current historical error time series feature is obtained. If there is an error time series feature between the level value and its adjacent maximum or minimum value, it means that each level value has an analyzable error value, which further indicates that the level division is accurate, and the level value corresponding to the current division is obtained.

[0105] In step S1034, when it is determined that there is no error timing feature between the level value of the current historical error timing feature and its closest maximum or minimum value, the interval base value is updated according to formula (4):

[0106] σ=σ+δ, (4)

[0107] Among them, δ is the adjustment base value, and the adjustment base value satisfies 0<δ≤1. Specifically, if there is no error time series feature between the level value and its closest maximum or minimum value, it means that the division is unreasonable. This kind of division is constant at 0 at a certain level and has no reference significance.

[0108] In step S1035, return to the step of obtaining the level base value of the historical error time series feature according to formula (2).

[0109] In step S1030 to step S1035, the level base value is first obtained according to the maximum and minimum values ​​of the historical error time series characteristics, and based on the level base value, it is converted into two level values. According to the presence or absence of historical error time series characteristics between the two level values ​​and the similar maximum or minimum values, it is confirmed whether the division of the level value is reasonable. If there are historical error time series characteristics between the two level values ​​and the similar maximum or minimum values, it means that each level area has analytical value, and the current divided level value can be obtained. Otherwise, it means that some level areas do not have analytical value, and it is necessary to update the interval base value to continuously adjust the level base value, and finally obtain the optimal level value. In this way, reliable division of the historical error feature sequence can be achieved.

[0110] In step S100 to step S103, the historical error time series characteristics can be obtained according to the historical output time series characteristics and the historical real time series characteristics of the supervision model, and the mean, maximum and minimum values ​​are determined according to the historical error time series characteristics. By analyzing the mean, maximum and minimum values ​​of the historical time series characteristics, multiple evaluation levels of the supervision model can be reasonably determined.

[0111] In this embodiment of the present invention, after obtaining the error time series characteristics of the current supervision model, it is also necessary to obtain its evaluation matrix according to the evaluation level. The specific acquisition steps can be as follows: Figure 4 Specifically, Figure 4 In the above, the evaluation method may further include:

[0112] In step S130, the error time series characteristics of the current supervision model are divided according to the evaluation levels. The two evaluation levels divide the error time series characteristics of the current supervision model into three level areas.

[0113] In step S131, the evaluation probability of the error time series characteristics of the current supervision model at each level is obtained according to formula (6):

[0114]

[0115] Among them, p k is the evaluation probability of the error time series characteristics of the current supervision model at level k, k is the level interval, and k = {1, 2, 3}. n2 k is the end time of the nth segment of the error time series characteristic of the current supervision model at level k, T n1 k is the start time of the nth segment of the error timing characteristics of the current supervision model at level k, n is an integer number, j is an integer number, N is the total number of segments of the error timing characteristics of the current supervision model at level k, and T is the total duration of the error timing characteristics of the current supervision model.

[0116] In step S132, the evaluation matrix of the current supervision model is obtained according to the evaluation probability of the error time series characteristics of the current supervision model at each level. 3×1 =[p1p2p3].

[0117] In step S130 to step S132, the error time series characteristics of the current supervision model are first divided according to the evaluation level to obtain multiple error time series characteristics. There may be multiple discontinuous error time series characteristics in each evaluation level. Therefore, the multiple error time series characteristics in each evaluation level can be integrated, and the ratio of the total duration of the error time series characteristics to the error time series characteristics can be used as the distribution probability of the error time series characteristics at this level, that is, the evaluation probability. Finally, the evaluation probabilities at the three levels are summarized to obtain the evaluation matrix of the error time series characteristics.

[0118] In this embodiment of the present invention, after obtaining the evaluation matrix of the current supervision model, it needs to be converted into an evaluation value to evaluate the current established model. The specific conversion steps can be as follows: Figure 5 shown.

[0119] Specifically, in Figure 5 In the above, the evaluation method may further include:

[0120] In step S140, the decision value of the current supervision model at each level is obtained, and a decision matrix is ​​constructed. The decision value of the supervision model at each level can be determined according to the historical output time series characteristics and historical real time series characteristics of the supervision model. The specific steps can be as follows: Figure 6 Specifically, Figure 6 In the above, the evaluation method may further include:

[0121] In step S1400, the decision value of the current supervision model at each level is randomized, and an initial population is constructed. A particle swarm algorithm can be used to randomize the decision value of the supervision model at each level.

[0122] In step S1401, the fitness value of each individual in the population is obtained according to formula (8):

[0123]

[0124] Among them, F fitness is the fitness value, K is the number of levels of the supervision model, and K = 3, q k is the decision value of the error time series characteristics of the current supervision model at level k, p k·l is the evaluation probability of the historical error time series feature at level k. Specifically, for the evaluation probability p of the historical error time series feature at level k k·l The time series characteristics with smaller differences between the output value of the supervision model and the true value can be used, that is, the time series characteristics with better / optimal output effect of the supervision model can be used as the optimization calculation of the objective function.

[0125] In step S1402, the best individual in the population is selected according to the fitness value, wherein the best individual is the individual corresponding to the minimum fitness value.

[0126] In step S1403, the remaining individuals in the population are iterated according to the optimal individual.

[0127] In step S1404, it is determined whether the number of iterations of the current population is greater than or equal to an iteration number threshold, where the iteration number threshold is a preset value.

[0128] In step S1405, when it is determined that the number of iterations of the current population is greater than or equal to the iteration number threshold, the optimal solution of the decision value of the current supervision model is output. If the number of iterations of the current population is greater than the iteration number threshold, it means that the iteration is completed, and the global optimal solution is obtained as the optimal solution of the current supervision model.

[0129] In step S1406, when it is determined that the number of iterations of the current population is less than the iteration number threshold, the step of obtaining the fitness value of each individual in the population according to formula (8) is returned. If the number of iterations of the current population is less than the iteration number threshold, it means that the iteration is not completed, and it is necessary to update the current number of iterations and return to recalculate the fitness value of the next population.

[0130] In step S1400 to step S1406, the decision value is first randomized to obtain the initial population, and then the fitness value of each individual in the population is calculated to select the optimal individual of the current population. The remaining individuals in the current population are iterated according to the optimal individual, and the number of iterations of the current population is compared with the iteration number threshold. If the number of iterations of the current population is greater than or equal to the iteration number threshold, it means that the iteration is completed and the global optimal solution can be output. Otherwise, the current number of iterations needs to be updated and the fitness value of the population needs to be recalculated, and this cycle is repeated. In this way, the optimal decision value of the error time series feature in each level area can be effectively obtained. Specifically, the decision matrix for the supervision model can be

[0131]

[0132] In step S141, the evaluation value of the current supervision model is obtained according to formula (7):

[0133] S=P 3×1 ×Q 1×3 , (7)

[0134] Among them, S is the evaluation value of the current supervision model, P 3×1 is the evaluation matrix of the current supervision model, Q 1×3 is the decision matrix.

[0135] In step S140 to step S141, the decision matrix of the supervision model is first obtained according to the historical error time series characteristics, and then the decision matrix is ​​multiplied by the evaluation matrix of the current supervision model to obtain the evaluation value of the current supervision model, and then the current supervision model can be evaluated.

[0136] In this embodiment of the present invention, after obtaining the evaluation value of the current supervision model, it can be compared with the evaluation threshold to determine the quality of the current supervision model. The specific evaluation steps can be as follows: Figure 7 Specifically, Figure 7 In the above, the evaluation method may further include:

[0137] In step S150, it is determined whether the evaluation value of the current supervision model is less than or equal to the evaluation threshold.

[0138] In step S151, when it is determined that the evaluation value of the current supervision model is less than or equal to the evaluation threshold, the supervision quality of the current supervision model is determined to be good. If the evaluation value of the current supervision model is less than or equal to the evaluation threshold, it means that the output error of the current supervision model is small, that is, the supervision quality is good.

[0139] In step S152, when it is determined that the evaluation value of the current supervision model is greater than the evaluation threshold, the supervision quality of the current supervision model is determined to be poor. If the evaluation value of the current supervision model is greater than the evaluation threshold, it means that the output error of the current supervision model is large, that is, the supervision quality is poor.

[0140] In step S150 to step S152, the evaluation value of the current supervision model is compared with the evaluation threshold. If the evaluation value is less than or equal to the evaluation threshold, it means that the output error of the current supervision model is small and the supervision quality / effect of the supervision model is good. Otherwise, the supervision quality / effect is poor.

[0141] On the other hand, the present invention further provides a computer-readable storage medium, wherein the computer-readable storage medium stores instructions, and the instructions are used to be read by a machine so that the machine executes any of the above evaluation methods.

[0142] Through the above technical scheme, the digital supervision quality evaluation method and storage medium based on the big data model provided by the present invention obtain the error timing characteristics of the current supervision model through the output timing characteristics and real timing characteristics of the current supervision model, and combine the evaluation level of the constructed supervision model to obtain the evaluation matrix of the current supervision model. According to the evaluation matrix, the evaluation value of the supervision model can be obtained, and then the current supervision model can be evaluated by using the evaluation value; by adopting this method of quantifying the error timing characteristics of the supervision model into specific evaluation values, it is possible to achieve effective evaluation of the supervision model, that is, effective supervision, thereby ensuring the reliability and accuracy of power grid system supervision.

[0143] Those skilled in the art will appreciate that the embodiments of the present application may be provided as methods, systems, or computer program products. Therefore, the present application may adopt the form of a complete hardware embodiment, a complete software embodiment, or an embodiment in combination with software and hardware. Moreover, the present application may adopt the form of a computer program product implemented in one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) that include computer-usable program code.

[0144] The present application is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of the processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0145] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0146] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0147] In a typical configuration, a computing device includes one or more processors (CPU), input / output interfaces, network interfaces, and memory.

[0148] The memory may include non-permanent memory in a computer-readable medium, random access memory (RAM) and / or non-volatile memory in the form of read-only memory (ROM) or flash RAM. The memory is an example of a computer-readable medium.

[0149] Computer readable media include permanent and non-permanent, removable and non-removable media that can be implemented by any method or technology to store information. Information can be computer readable instructions, data structures, program modules or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technology, compact disk read-only memory (CD-ROM), digital versatile disk (DVD) or other optical storage, magnetic cassettes, magnetic tape disk storage or other magnetic storage devices or any other non-transmission media that can be used to store information that can be accessed by a computing device. As defined herein, computer readable media does not include temporary computer readable media (transitory media), such as modulated data signals and carrier waves.

[0150] It should also be noted that the terms "include", "comprises" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, commodity or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, commodity or device. In the absence of more restrictions, the elements defined by the sentence "comprises a ..." do not exclude the existence of other identical elements in the process, method, commodity or device including the elements.

[0151] The above are only embodiments of the present application and are not intended to limit the present application. For those skilled in the art, the present application may have various changes and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application should be included within the scope of the claims of the present application.

Claims

1. A method for evaluating the quality of digital supervision based on a big data model, characterized in that: include: Construct the evaluation level of the supervision model; Obtain the output timing characteristics and real timing characteristics of the current supervision model; Obtaining error timing characteristics of the current supervision model according to the output timing characteristics of the current supervision model and the real timing characteristics; Acquire an evaluation matrix according to the error timing characteristics and the evaluation level; Obtaining an evaluation value of the current supervision model according to the evaluation matrix; The quality of the current supervision model is evaluated according to the evaluation value of the current supervision model.

2. The evaluation method according to claim 1, characterized in that: The evaluation levels for building supervision models include: Obtaining multiple historical output time series features of the supervision model and corresponding historical real time series features; Acquire historical error timing characteristics according to the historical output timing characteristics and the historical real timing characteristics; Obtaining the mean, maximum and minimum values ​​of the historical error time series characteristics; The evaluation level of the supervision model is divided according to the mean value, maximum value and minimum value of the historical error time series characteristics.

3. The evaluation method according to claim 2, characterized in that: Acquiring the historical error timing characteristics according to the historical output timing characteristics and the historical real timing characteristics includes: According to formula (1), the historical error time series characteristics are obtained: in, is the value of the i-th historical error time series feature at time t, is the value of the i-th historical output time series feature at time t, is the value of the i-th historical real time series feature at time t, i is an integer number, and t is the time.

4. The evaluation method according to claim 3, characterized in that: The evaluation level of the supervision model is divided according to the mean, maximum and minimum values ​​of the historical error time series characteristics, including: According to formula (2), the level base value of the historical error time series feature is obtained: Among them, Δε is the level base value of the historical error time series feature, ε max is the maximum value of the historical error time series characteristic, ε min is the minimum value of the historical error time series characteristic, and σ is the interval base value; According to formula (3), the level value of the historical error time series feature is obtained: Among them, ε l is the level value of the historical error time series feature, l is the level sequence number of the historical error time series feature, l={1,2}, is the mean value of the historical error time series characteristics; Determine whether there is the error time series feature between the current level value of the historical error time series feature and its closest maximum or minimum value; In the case where it is determined that there is an error timing feature between the level value of the current historical error timing feature and its closest maximum value or minimum value, obtaining the level value of the current historical error timing feature; When it is determined that there is no error timing feature between the level value of the current historical error timing feature and its closest maximum or minimum value, the interval base value is updated according to formula (4): σ=σ+δ, (4) Among them, δ is the adjustment base value; Return to the step of obtaining the level base value of the historical error time series feature according to formula (2).

5. The evaluation method according to claim 4, characterized in that: Obtaining the error timing characteristics of the current supervision model according to the output timing characteristics of the current supervision model and the real timing characteristics includes: According to formula (5), the error time series characteristics of the current supervision model are obtained: in, is the value of the error time series characteristic of the current supervision model at time t, is the value of the output time series characteristic of the current supervision model at time t, is the value of the real time series characteristics of the current supervision model at time t.

6. The evaluation method according to claim 5, characterized in that: Acquiring an evaluation matrix according to the error timing characteristics and the evaluation level includes: Classifying the error time series characteristics of the current supervision model according to the evaluation level; According to formula (6), the evaluation probability of the error time series characteristics of the current supervision model at each level is obtained: Among them, p k is the evaluation probability of the error time series characteristics of the current supervision model at level k, k is the level interval, and k = {1, 2, 3}, T n2 k is the end time of the nth segment of the error time series characteristic of the current supervision model at level k, T n1 k is the start time of the nth segment of the error timing characteristics of the current supervision model at level k, n is an integer number, j is an integer number, N is the total number of segments of the error timing characteristics of the current supervision model at level k, and T is the total duration of the error timing characteristics of the current supervision model; The evaluation matrix of the current supervision model is obtained according to the evaluation probability of the error time series characteristics of the current supervision model at each level.

7. The evaluation method according to claim 6, characterized in that: Obtaining the evaluation value of the current supervision model according to the evaluation matrix includes: Obtain the decision value of the current supervision model at each level and construct a decision matrix; According to formula (7), the evaluation value of the current supervision model is obtained: S=P 3×1 ×Q 1×3 , (7) Among them, S is the evaluation value of the current supervision model, P 3×1 is the evaluation matrix of the current supervision model, Q 1×3 is the decision matrix.

8. The evaluation method according to claim 7, characterized in that: Obtaining the decision values ​​of the current supervision model at each level includes: Randomize the decision value of the current supervision model at each level and construct an initial population; According to formula (8), the fitness value of each individual in the population is obtained: Among them, F fitness is the fitness value, K is the number of levels of the supervision model, and K=3, q k is the decision value of the error time series characteristic of the supervision model at level k, p k·l is the evaluation probability of the historical error time series feature at level k; Selecting the best individual in the population according to the fitness value; Iterate the remaining individuals in the population according to the optimal individual; Determine whether the number of iterations of the current population is greater than or equal to a threshold number of iterations; When it is determined that the number of iterations of the current population is greater than or equal to the iteration number threshold, outputting the optimal solution of the decision value of the current supervision model; When it is determined that the current number of iterations of the population is less than the iteration number threshold, return to the step of obtaining the fitness value of each individual in the population according to formula (8).

9. The evaluation method according to claim 1, characterized in that: Evaluating the quality of the current supervision model according to the evaluation value of the current supervision model includes: Determine whether the evaluation value of the current supervision model is less than or equal to the evaluation threshold; In the case where it is determined that the evaluation value of the current supervision model is less than or equal to the evaluation threshold, determining that the supervision quality of the current supervision model is good; When it is determined that the evaluation value of the current supervision model is greater than the evaluation threshold, it is determined that the supervision quality of the current supervision model is poor.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores instructions, and the instructions are used to be read by a machine so that the machine executes the evaluation method according to any one of claims 1 to 9.

Images