Station monitoring index fluctuation detection method and terminal

Abstract

The invention discloses a station monitoring index fluctuation detection method and a terminal, and the method comprises the steps: obtaining a historical monitoring index data sequence of a station, and dividing the historical monitoring index data sequence into a plurality of index data subsequences according to a time sequence; calculating an average value of trend values, a variance of residual values and a variance acceleration of the residual values of each index data sub-sequence, and further determining a trend value average value reference range, a residual value variance reference range and a residual value variance acceleration reference range; and acquiring the latest monitoring index data sequence of the station, and determining whether the latest monitoring index data sequence fluctuates or not by judging whether the average value of the trend values of the latest monitoring index data sequence, the variance of the residual values or the variance acceleration of the residual values exceed the corresponding reference range or not. According to the invention, whether the monitoring index data of the station fluctuates or not can be accurately judged from multiple angles, so that operation and maintenance personnel can carry out early warning processing in time, and reliable protection is provided for the safety of the station.

Description

Technical Field

[0001] The present invention relates to the field of electronic information technology, and in particular to a method and a terminal for detecting fluctuations of site monitoring indicators. Background Art

[0002] As one of the basic infrastructures for vehicle charging inspection, the existing photovoltaic storage charging inspection stations have an increasingly wide coverage. In order to ensure that the stations can provide stable power supply services for electric vehicles, it is currently necessary to improve the safety management level of the stations and ensure the safe operation of the stations.

[0003] However, the data distribution of existing site safety monitoring indicators is unimodal or multimodal, and the data changes are relatively unstable; therefore, how to effectively identify the waveform of site monitoring indicator data becomes the key to judging whether the site operation is abnormal. Summary of the invention

[0004] The technical problem to be solved by the present invention is to propose a method and terminal for detecting fluctuations in monitoring indexes of a station, so as to effectively identify fluctuations in monitoring index data of the station, so as to make early warning processing in time and provide protection for the safety of the station.

[0005] In order to solve the above technical problems, the technical solution adopted by the present invention is:

[0006] A method for detecting fluctuations in monitoring indicators of a station comprises the following steps:

[0007] S1. Obtain a historical monitoring indicator data sequence of the station within a first preset time period, and divide the historical monitoring indicator data sequence into a preset number of indicator data subsequences in chronological order;

[0008] S2. Calculate the average value of the trend value, the variance of the residual value, and the variance growth rate of the residual value of each of the indicator data subsequences according to the time series decomposition algorithm;

[0009] S3. Obtain a reference range of the mean trend value according to the average trend value of all the indicator data subsequences, obtain a reference range of the residual value variance according to the variance of the residual values ​​of all the indicator data subsequences, and obtain a reference range of the residual value variance growth rate according to the variance growth rate of the residual values ​​of all the indicator data subsequences;

[0010] S4. Obtain the latest monitoring indicator data sequence of the station from the current start to the second preset time length, and determine whether the average value of the trend value of the latest monitoring indicator data sequence exceeds the trend value mean reference range, or whether the variance of the residual value exceeds the residual value variance reference range, or whether the variance growth rate of the residual value exceeds the residual value variance growth rate reference range. If so, output that the latest monitoring indicator data sequence has fluctuations.

[0011] In order to solve the above technical problems, another technical solution adopted by the present invention is:

[0012] A station monitoring index fluctuation detection terminal includes a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the following steps when executing the computer program:

[0013] S1. Obtain a historical monitoring indicator data sequence of the station within a first preset time period, and divide the historical monitoring indicator data sequence into a preset number of indicator data subsequences in chronological order;

[0014] S2. Calculate the average value of the trend value, the variance of the residual value, and the variance growth rate of the residual value of each of the indicator data subsequences according to the time series decomposition algorithm;

[0015] S3. Obtain a reference range of the mean trend value according to the average trend value of all the indicator data subsequences, obtain a reference range of the residual value variance according to the variance of the residual values ​​of all the indicator data subsequences, and obtain a reference range of the residual value variance growth rate according to the variance growth rate of the residual values ​​of all the indicator data subsequences;

[0016] S4. Obtain the latest monitoring indicator data sequence of the station from the current start to the second preset time length, and determine whether the average value of the trend value of the latest monitoring indicator data sequence exceeds the trend value mean reference range, or whether the variance of the residual value exceeds the residual value variance reference range, or whether the variance growth rate of the residual value exceeds the residual value variance growth rate reference range. If so, output that the latest monitoring indicator data sequence has fluctuations.

[0017] The beneficial effects of the present invention are: providing a method and terminal for detecting fluctuations in monitoring indicators of a station, calculating and generating a trend value mean reference range, a residual value variance reference range and a residual value variance growth rate reference range for fluctuation detection based on the historical monitoring indicator data sequence of the station, and then obtaining the latest monitoring indicator data sequence during detection, and according to whether the average value of the trend value, the variance of the residual value and the residual value variance growth rate of the latest monitoring indicator data sequence exceed the corresponding reference range, it is possible to accurately judge from multiple angles whether fluctuations occur in the monitoring indicator data of the station, so that operation and maintenance personnel can make early warning processing in a timely manner and provide reliable protection for station safety. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] Figure 1 A schematic diagram of the steps of a method for detecting fluctuations in site monitoring indicators of the present invention;

[0019] Figure 2A flow chart of an abnormal judgment of the average value of a trend value in a method for detecting fluctuations in monitoring indicators of a station according to the present invention;

[0020] Figure 3 A flow chart of performing variance abnormality judgment of residual values ​​in a method for detecting fluctuations of site monitoring indicators of the present invention;

[0021] Figure 4 A flow chart of abnormal judgment of variance growth rate of the variance of residual value by a method for detecting fluctuation of a station monitoring index of the present invention;

[0022] Figure 5 A flow chart of determining the variance homogeneity anomaly of residual values ​​in a method for detecting fluctuations in monitoring indicators of a station according to the present invention;

[0023] Figure 6 The present invention is a system block diagram of a station monitoring index fluctuation detection terminal.

[0024] Description of labels:

[0025] 1. A station monitoring index fluctuation detection terminal; 2. Memory; 3. Processor. DETAILED DESCRIPTION

[0026] In order to explain the technical content, achieved objectives and effects of the present invention in detail, the following is an explanation in conjunction with the implementation modes and the accompanying drawings.

[0027] Please refer to Figures 1 to 5 A method for detecting fluctuations in monitoring indicators of a station comprises the following steps:

[0028] S1. Obtain a historical monitoring indicator data sequence of the station within a first preset time period, and divide the historical monitoring indicator data sequence into a preset number of indicator data subsequences in chronological order;

[0029] S2. Calculate the average value of the trend value, the variance of the residual value, and the variance growth rate of the residual value of each of the indicator data subsequences according to the time series decomposition algorithm;

[0030] S3. Obtain a reference range of the mean trend value according to the average trend value of all the indicator data subsequences, obtain a reference range of the residual value variance according to the variance of the residual values ​​of all the indicator data subsequences, and obtain a reference range of the residual value variance growth rate according to the variance growth rate of the residual values ​​of all the indicator data subsequences;

[0031] S4. Obtain the latest monitoring indicator data sequence of the station from the current start to the second preset time length, and determine whether the average value of the trend value of the latest monitoring indicator data sequence exceeds the trend value mean reference range, or whether the variance of the residual value exceeds the residual value variance reference range, or whether the variance growth rate of the residual value exceeds the residual value variance growth rate reference range. If so, output that the latest monitoring indicator data sequence has fluctuations.

[0032] From the above description, it can be seen that the beneficial effects of the present invention are: based on the historical monitoring indicator data sequence of the station, a trend value mean reference range, a residual value variance reference range and a residual value variance growth rate reference range for fluctuation detection are calculated and generated, and then the latest monitoring indicator data sequence is obtained during detection, and the average value of the trend value, the variance of the residual value and the residual value variance growth rate of the latest monitoring indicator data sequence exceed the corresponding reference range, so as to achieve multi-angle and accurate judgment on whether the monitoring indicator data of the station fluctuates, so that the operation and maintenance personnel can make early warning processing in time and provide reliable protection for the safety of the station.

[0033] Furthermore, the step S4 further includes:

[0034] The latest monitoring indicator data sequence is subtracted from a preset calibration indicator data sequence to obtain an indicator residual sequence;

[0035] The indicator residual sequence is divided into two residual subsequences in chronological order, and the variance consistency results of the two residual subsequences are calculated and used to determine whether the latest monitoring indicator data sequence has fluctuations.

[0036] From the above description, it can be seen that in addition to the mean value of the trend value, the variance of the residual value and the variance growth rate of the residual value, the variance consistency judgment result of the residual is also introduced, so that the overall detection method not only uses historical data as a reference, but also combines and compares the ideal calibration indicator data series, so as to more comprehensively and accurately judge whether there is fluctuation in the monitoring indicator data.

[0037] Furthermore, the step S4 further includes:

[0038] Establishing an indicator calibration model, using the historical monitoring indicator data sequence as a model training set;

[0039] The model training set is used to train the indicator calibration model, and the calibration indicator data sequence is generated by the trained indicator calibration model.

[0040] From the above description, it can be seen that the indicator calibration model is applied to quickly generate a calibration indicator data sequence, and the indicator calibration model is trained in combination with the historical monitoring indicator data sequence, so as to improve the accuracy of the data provided by the indicator calibration model and its applicability relative to the site, so that the subsequent residual variance consistency judgment results are more reliable.

[0041] Furthermore, the step S2 comprises:

[0042] Selecting at least two indicator data blocks from the indicator data subsequence in chronological order;

[0043] Calculate and obtain the variance growth rate V of the residual value of the indicator data subsequence based on the variance of the residual value of the two adjacent indicator data blocks. std :

[0044] V std =(std cur -std pre ) / t;

[0045] Among them, std pre Indicates the variance of the residual value of the previous indicator data block, std cur represents the variance of the residual value of the latter indicator data block, and t represents the time interval corresponding to the indicator data recording time corresponding to the two adjacent indicator data blocks.

[0046] From the above description, it can be seen that when calculating the variance growth rate, the indicator data blocks contained in the indicator data subsequence are selected in chronological order, and the variance growth rate of the indicator data subsequence can be accurately calculated through the variance change of the residual values ​​of the two previous and subsequent indicator data blocks.

[0047] Furthermore, the step S1 further includes:

[0048] A sequence length threshold and a duration threshold are preset, and the indicator data subsequences whose sequence length is less than or equal to the sequence length threshold and whose sequence duration is less than or equal to the duration threshold are eliminated.

[0049] From the above description, it can be seen that before using the indicator data subsequence, the sequence length threshold and the time threshold are preset as filtering conditions, so as to filter out some indicator data subsequences that have no reference significance and improve the accuracy of fluctuation judgment.

[0050] Please refer to Figure 6 A station monitoring index fluctuation detection terminal 1 includes a memory 2, a processor 3, and a computer program stored in the memory 2 and executable on the processor 3. When the processor 3 executes the computer program, the following steps are implemented:

[0051] S1. Obtain a historical monitoring indicator data sequence of the station within a first preset time period, and divide the historical monitoring indicator data sequence into a preset number of indicator data subsequences in chronological order;

[0052] S2. Calculate the average value of the trend value, the variance of the residual value, and the variance growth rate of the residual value of each of the indicator data subsequences according to the time series decomposition algorithm;

[0053] S3. Obtain a reference range of the mean trend value according to the average trend value of all the indicator data subsequences, obtain a reference range of the residual value variance according to the variance of the residual values ​​of all the indicator data subsequences, and obtain a reference range of the residual value variance growth rate according to the variance growth rate of the residual values ​​of all the indicator data subsequences;

[0054] S4. Obtain the latest monitoring indicator data sequence of the station from the current start to the second preset time length, and determine whether the average value of the trend value of the latest monitoring indicator data sequence exceeds the trend value mean reference range, or whether the variance of the residual value exceeds the residual value variance reference range, or whether the variance growth rate of the residual value exceeds the residual value variance growth rate reference range. If so, output that the latest monitoring indicator data sequence has fluctuations.

[0055] From the above description, it can be seen that the beneficial effects of the present invention are: based on the historical monitoring indicator data sequence of the station, a trend value mean reference range, a residual value variance reference range and a residual value variance growth rate reference range for fluctuation detection are calculated and generated, and then the latest monitoring indicator data sequence is obtained during detection, and the average value of the trend value, the variance of the residual value and the residual value variance growth rate of the latest monitoring indicator data sequence exceed the corresponding reference range, so as to achieve multi-angle and accurate judgment on whether the monitoring indicator data of the station fluctuates, so that the operation and maintenance personnel can make early warning processing in time and provide reliable protection for the safety of the station.

[0056] Furthermore, the step S4 further includes:

[0057] The latest monitoring indicator data sequence is subtracted from a preset calibration indicator data sequence to obtain an indicator residual sequence;

[0058] The indicator residual sequence is divided into two residual subsequences in chronological order, and the variance consistency results of the two residual subsequences are calculated and used to determine whether the latest monitoring indicator data sequence has fluctuations.

[0059] From the above description, it can be seen that in addition to the mean value of the trend value, the variance of the residual value and the variance growth rate of the residual value, the consistency judgment result of the residual is also introduced, so that the overall detection method not only uses historical data as a reference, but also combines and compares the ideal calibration indicator data series, so as to more comprehensively and accurately judge whether there is fluctuation in the monitoring indicator data.

[0060] Furthermore, the step S4 further includes:

[0061] Establishing an indicator calibration model, using the historical monitoring indicator data sequence as a model training set;

[0062] The model training set is used to train the indicator calibration model, and the calibration indicator data sequence is generated by the trained indicator calibration model.

[0063] From the above description, it can be seen that the indicator calibration model is applied to quickly generate a calibration indicator data sequence, and the indicator calibration model is trained in combination with the historical monitoring indicator data sequence, so as to improve the accuracy of the data provided by the indicator calibration model and its applicability relative to the site, so that the subsequent residual variance consistency judgment results are more reliable.

[0064] Furthermore, the step S2 comprises:

[0065] Selecting at least two indicator data blocks from the indicator data subsequence in chronological order;

[0066] Calculate and obtain the variance growth rate V of the residual value of the indicator data subsequence based on the variance of the residual value of the two adjacent indicator data blocks. std :

[0067] V std =(std cur -std pre ) / t;

[0068] Among them, std pre Indicates the variance of the residual value of the previous indicator data block, std cur represents the variance of the residual value of the latter indicator data block, and t represents the time interval corresponding to the indicator data recording time corresponding to the two adjacent indicator data blocks.

[0069] From the above description, it can be seen that when calculating the variance growth rate, the indicator data blocks contained in the indicator data subsequence are selected in chronological order, and the variance growth rate of the indicator data subsequence can be accurately calculated through the variance change of the residual values ​​of the two previous and subsequent indicator data blocks.

[0070] Furthermore, the step S1 further includes:

[0071] A sequence length threshold and a duration threshold are preset, and the indicator data subsequences whose sequence length is less than or equal to the sequence length threshold and whose sequence duration is less than or equal to the duration threshold are eliminated.

[0072] From the above description, it can be seen that before using the indicator data subsequence, the sequence length threshold and the time threshold are preset as filtering conditions, so as to filter out some indicator data subsequences that have no reference significance and improve the accuracy of fluctuation judgment.

[0073] Please refer to Figures 1 to 4 , Embodiment 1 of the present invention is:

[0074] A method for detecting fluctuations in monitoring indicators of a station comprises the following steps:

[0075] S1. Obtain a historical monitoring indicator data sequence of the station within a first preset time period, and divide the historical monitoring indicator data sequence into a preset number of indicator data subsequences in chronological order;

[0076] In this embodiment, the first preset duration can be selected in hours, and the value is M; preferably, M is greater than or equal to 36; when dividing the indicator data subsequences, the preset time interval can be used as the basis for division, and the division is carried out in chronological order, and the preset time interval is preferably 2 hours; when the division is completed, the sequence length threshold and the duration threshold are preset, and the indicator data subsequences whose sequence length is less than or equal to the sequence length threshold and the sequence duration is less than or equal to the duration threshold are eliminated. Among them, combined with the aforementioned preferred scheme, the optional range of the sequence length threshold is 230 to 250 data points, preferably 240 data points, and the optional range of the duration threshold is 1.5 hours to 2.5 hours, preferably 2 hours.

[0077] S2. Calculate the average value of the trend value, the variance of the residual value, and the variance growth rate of the residual value of each indicator data subsequence according to the time series decomposition algorithm;

[0078] In this embodiment, the time series decomposition algorithm is preferably an STL algorithm, STL (Seasonal-Trend Decomposition Procedure based on Loess): In time series analysis, the STL method is used to decompose time series data into three parts: seasonality, trend, and residual.

[0079] S3. According to the average value of the trend value of all indicator data subsequences, a reference range of the mean value of the trend value is obtained; according to the variance of the residual value of all indicator data subsequences, a reference range of the variance of the residual value is obtained; according to the variance growth rate of the residual value of all indicator data subsequences, a reference range of the variance growth rate of the residual value is obtained;

[0080] In this embodiment, specifically in each indicator data subsequence, according to the sliding window length being the third preset time length and the number of data in the window being required to exceed the preset number of points, the data is slid and selected to calculate the average value of the trend value; the optional range of the third preset time length is 2 to 4 hours, preferably 3 hours, and the optional range of the preset number of points is 35 to 45 points, preferably 40 points; in other words, the average value of the trend value corresponding to each indicator data subsequence can be one or more; then, the average values ​​of the trend values ​​calculated from all sequences are merged into one sequence, and the reference range of the trend value mean is calculated based on the GESD algorithm;

[0081] Among them, the GESD algorithm (Generalized Extreme Studentized Deviate) is an existing statistical method used to detect outliers in univariate data sets that follow an approximately normal distribution.

[0082] In each indicator data subsequence, the sliding window length is also the third preset time length and the number of data in the window is required to exceed the preset number of points, and the data is slid and selected to calculate the variance of the residual value. In other words, the variance of the residual value corresponding to each indicator data subsequence can be one or more; then, the variances of the residual values ​​calculated from all sequences are merged into one sequence, and the reference range of the residual value variance is calculated based on the GESD algorithm;

[0083] In this embodiment, in each indicator data subsequence, according to the sliding window length being the third preset time length and the number of data in the window being required to exceed the preset number of points, at least two indicator data blocks are slid and selected; and the variance growth rate V of the residual value of the indicator data subsequence is obtained based on the variance of the residual value of the two adjacent indicator data blocks. std :

[0084] V std =(std cur -std pre ) / t;

[0085] Among them, std pre The variance of the residual value of the previous indicator data block, std cur It represents the variance of the residual value of the next indicator data block, and t represents the time interval corresponding to the indicator data recording time corresponding to the two adjacent indicator data blocks.

[0086] Finally, the variance growth rates of the residual values ​​calculated from all sequences are merged into one sequence, and the reference range of the variance growth rate of the residual values ​​is calculated based on the GESD algorithm.

[0087] S4. Obtain the latest monitoring indicator data sequence of the station from the current start to the second preset time length, and determine whether the average value of the trend value of the latest monitoring indicator data sequence exceeds the trend value mean reference range, or whether the variance of the residual value exceeds the residual value variance reference range, or whether the variance growth rate of the residual value exceeds the residual value variance growth rate reference range. If so, output the latest monitoring indicator data sequence if there is fluctuation.

[0088] In this embodiment, the second preset duration can be selected in hours, and the value is N; preferably, N is greater than or equal to 4; based on the STL algorithm, the trend value of the latest monitoring indicator data sequence is extracted, and then the trend value of the last 2 hours in the second preset duration is intercepted, the average value of the trend value is calculated, and then the average value of the trend value is judged whether it exceeds the trend value mean reference range, and whether the variance of the residual value exceeds the residual value variance reference range is similar to the above content. When judging whether the variance growth rate of the residual value exceeds the reference range of the residual value variance growth rate, the data of the last three hours in the second preset duration is taken to calculate the variance values ​​of three residual values ​​in hours, and the variance growth rate of the variance values ​​of the residual values ​​between two adjacent data is calculated, and the two variance growth rates are averaged as the variance growth rate of the residual value of the latest monitoring indicator data sequence and used to judge whether it exceeds the reference range of the residual value variance growth rate. When any of the above three out-of-range conditions is met, the output of the latest monitoring indicator data sequence is fluctuating.

[0089] Please refer to Figure 5 , Embodiment 2 of the present invention is:

[0090] A method for detecting fluctuations in site monitoring indicators. Based on the above-mentioned embodiment 1, in order to improve the comprehensiveness and accuracy of fluctuation judgment, step S4 also includes: intercepting the original indicator sequence of a specified number of points from the latest monitoring indicator data sequence, subtracting a preset calibration indicator data sequence from the original indicator sequence, and obtaining an indicator residual sequence, wherein the specified number is preferably 60; dividing the indicator residual sequence into two residual subsequences in chronological order, calculating and judging whether the latest monitoring indicator data sequence has fluctuations based on the variance consistency results of the two residual subsequences.

[0091] In this embodiment, an indicator calibration model is established, and the historical monitoring indicator data sequence is used as a model training set, wherein the model training set is used to train the indicator calibration model, and the calibration indicator data sequence is generated by the trained indicator calibration model. The historical monitoring indicator data sequence is divided into multiple subsequences with continuous time and length of 60 points, the model feature is the station did, the subsequence his, the training target is the subsequence his, the input and output are consistent, and a model training set is formed. During training, the following is mainly used: Figure 5 The convolutional model shown is trained as follows:

[0092] Convolutional model: The convolutional model is a deep learning neural network structure that is mainly used to process tasks with grid-structured data, such as image and audio processing. Its core component is the convolutional layer, which uses filters (also called convolution kernels) to perform convolution operations on the input data.

[0093] Embedding: In the field of computer science, embedding refers to the technology of mapping high-dimensional data to low-dimensional space. In natural language processing (NLP), embedding usually refers to word embedding, which maps words to low-dimensional dense vectors in real vector space. These vectors have certain semantic information and can capture the semantic relationship between words.

[0094] Concatenate: In deep learning, "concatenate" usually refers to connecting two or more tensors together along a certain dimension.

[0095] Conv1d: Conv1D is a one-dimensional Convolutional Neural Network (CNN) layer, which is usually used to process data with sequence structure, such as time series data, text data, or audio data.

[0096] Dense: In deep learning, "Dense" refers to the fully connected layer, also known as the densely connected layer. Each neuron in the fully connected layer is connected to all neurons in the previous layer, and each connection has a weight. This means that each neuron in the fully connected layer receives the output of all neurons in the previous layer, and the weighted sum is performed through the weight, and then processed by the activation function.

[0097] Timedistributed: In deep learning, "TimeDistributed" is a technique for processing sequence data, usually in conjunction with a recurrent neural network (RNN) or a convolutional neural network (CNN). The technique applies a neural network layer to each time step of the sequence data so that the entire sequence can be processed in parallel.

[0098] In this embodiment, the variance analysis of the two residual subsequences is specifically verified based on the Levene test, and the p-value of the test is obtained as the variance consistency result; Levene test: Levene test is a statistical method for testing the homogeneity of variance of multiple groups of data samples. Its basic idea is to judge whether the difference between the data groups is significant by comparing the variance of each group of data. When judging, it is judged whether the p-value is less than 0.00001. If so, it means that the variance consistency is abnormal and the latest monitoring indicator data sequence has fluctuations; if not, it means that the variance consistency is normal and the latest monitoring indicator data sequence does not have fluctuations.

[0099] Please refer to Figure 6 , Embodiment 2 of the present invention is:

[0100] A site monitoring index fluctuation detection terminal 1 includes a memory 2, a processor 3, and a computer program stored in the memory 2 and executable on the processor 3. When the processor 3 executes the computer program, a site monitoring index fluctuation detection method of embodiment one or two is implemented.

[0101] In summary, the present invention provides a method and terminal for detecting fluctuations in station monitoring indicators. Based on the historical monitoring indicator data sequence of the station, a trend value mean reference range, a residual value variance reference range and a residual value variance growth rate reference range for fluctuation detection are calculated and generated. Then, the latest monitoring indicator data sequence is obtained during detection. The method and terminal are used to detect whether the average value of the trend value, the variance of the residual value and the variance growth rate of the residual value of the latest monitoring indicator data sequence exceed the corresponding reference range. At the same time, a calibration indicator data sequence is introduced to detect the variance consistency of the residual value of the latest monitoring indicator data sequence, so as to realize multi-angle and accurate judgment on whether the monitoring indicator data of the station fluctuates, so that the operation and maintenance personnel can make early warning processing in time, and provide reliable protection for the safety of the station.

[0102] The above descriptions are merely embodiments of the present invention and are not intended to limit the patent scope of the present invention. Any equivalent transformations made using the contents of the present invention's specification and drawings, or directly or indirectly applied in related technical fields, are also included in the patent protection scope of the present invention.

Claims

1. A method for detecting fluctuations in monitoring indicators of a station, characterized in that: The steps include: S1. Obtain a historical monitoring indicator data sequence of the station within a first preset time period, and divide the historical monitoring indicator data sequence into a preset number of indicator data subsequences in chronological order; S2. Calculate the average value of the trend value, the variance of the residual value, and the variance growth rate of the residual value of each of the indicator data subsequences according to the time series decomposition algorithm; S3. Obtain a reference range of the mean trend value according to the average trend value of all the indicator data subsequences, obtain a reference range of the residual value variance according to the variance of the residual values ​​of all the indicator data subsequences, and obtain a reference range of the residual value variance growth rate according to the variance growth rate of the residual values ​​of all the indicator data subsequences; S4. Obtain the latest monitoring indicator data sequence of the station from the current start to the second preset time length, and determine whether the average value of the trend value of the latest monitoring indicator data sequence exceeds the trend value mean reference range, or whether the variance of the residual value exceeds the residual value variance reference range, or whether the variance growth rate of the residual value exceeds the residual value variance growth rate reference range. If so, output that the latest monitoring indicator data sequence has fluctuations.

2. A method for detecting fluctuations in monitoring indicators of a station according to claim 1, characterized in that: The step S4 further comprises: The latest monitoring indicator data sequence is subtracted from a preset calibration indicator data sequence to obtain an indicator residual sequence; The indicator residual sequence is divided into two residual subsequences in chronological order, and the variance consistency results of the two residual subsequences are calculated and used to determine whether the latest monitoring indicator data sequence has fluctuations.

3. A method for detecting fluctuations in monitoring indicators of a station according to claim 2, characterized in that: The step S4 further comprises: Establishing an indicator calibration model, using the historical monitoring indicator data sequence as a model training set; The model training set is used to train the indicator calibration model, and the calibration indicator data sequence is generated by the trained indicator calibration model.

4. A method for detecting fluctuations in monitoring indicators of a station according to claim 1, characterized in that: The step S2 comprises: Selecting at least two indicator data blocks from the indicator data subsequence in chronological order; Calculate and obtain the variance growth rate V of the residual value of the indicator data subsequence based on the variance of the residual value of the two adjacent indicator data blocks. std : V std =(std cur -std pre ) / t; Among them, std pre Indicates the variance of the residual value of the previous indicator data block, std cur represents the variance of the residual value of the latter indicator data block, and t represents the time interval corresponding to the indicator data recording time corresponding to the two adjacent indicator data blocks.

5. A method for detecting fluctuations in monitoring indicators of a station according to claim 1, characterized in that: The step S1 further comprises: A sequence length threshold and a duration threshold are preset, and the indicator data subsequences whose sequence length is less than or equal to the sequence length threshold and whose sequence duration is less than or equal to the duration threshold are eliminated.

6. A station monitoring index fluctuation detection terminal, 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, the following steps are implemented: S1. Obtain a historical monitoring indicator data sequence of the station within a first preset time period, and divide the historical monitoring indicator data sequence into a preset number of indicator data subsequences in chronological order; S2. Calculate the average value of the trend value, the variance of the residual value, and the variance growth rate of the residual value of each of the indicator data subsequences according to the time series decomposition algorithm; S3. Obtain a reference range of the mean trend value according to the average trend value of all the indicator data subsequences, obtain a reference range of the residual value variance according to the variance of the residual values ​​of all the indicator data subsequences, and obtain a reference range of the residual value variance growth rate according to the variance growth rate of the residual values ​​of all the indicator data subsequences; S4. Obtain the latest monitoring indicator data sequence of the station from the current start to the second preset time length, and determine whether the average value of the trend value of the latest monitoring indicator data sequence exceeds the trend value mean reference range, or whether the variance of the residual value exceeds the residual value variance reference range, or whether the variance growth rate of the residual value exceeds the residual value variance growth rate reference range. If so, output that the latest monitoring indicator data sequence has fluctuations.

7. A station monitoring index fluctuation detection terminal according to claim 6, characterized in that: The step S4 further comprises: The latest monitoring indicator data sequence is subtracted from a preset calibration indicator data sequence to obtain an indicator residual sequence; The indicator residual sequence is divided into two residual subsequences in chronological order, and the variance consistency results of the two residual subsequences are calculated and used to determine whether the latest monitoring indicator data sequence has fluctuations.

8. A station monitoring index fluctuation detection terminal according to claim 7, characterized in that: The step S4 further comprises: Establishing an indicator calibration model, using the historical monitoring indicator data sequence as a model training set; The model training set is used to train the indicator calibration model, and the calibration indicator data sequence is generated by the trained indicator calibration model.

9. A station monitoring index fluctuation detection terminal according to claim 6, characterized in that: The step S2 comprises: Selecting at least two indicator data blocks from the indicator data subsequence in chronological order; Calculate and obtain the variance growth rate V of the residual value of the indicator data subsequence based on the variance of the residual value of the two adjacent indicator data blocks. std : V std =(std cur -std pre ) / t; Among them, std pre Indicates the variance of the residual value of the previous indicator data block, std cur represents the variance of the residual value of the latter indicator data block, and t represents the time interval corresponding to the indicator data recording time corresponding to the two adjacent indicator data blocks.

10. A station monitoring index fluctuation detection terminal according to claim 6, characterized in that: The step S1 further comprises: A sequence length threshold and a duration threshold are preset, and the indicator data subsequences whose sequence length is less than or equal to the sequence length threshold and whose sequence duration is less than or equal to the duration threshold are eliminated.

Images