A method for eliminating outliers in satellite telemetry data based on statistics

Abstract

The present invention relates to a statistical-based method for eliminating outliers in satellite telemetry data. The steps include: storing the satellite telemetry data in the form of a time series to form a data table D; initializing the data position time t = M and the data table size L = |D|; when t < L - N, using the method of dividing the deviation by the mean based on a sliding window to calculate the category to which the telemetry variable x ，i at time t, x t，i belongs; when t >= L - N, end the operation. The present invention compares the telemetry variable data with the means of the subsequent M data and the previous N data at the same time, avoiding the problem of false detection of outliers caused by mutations of a certain telemetry variable due to reasons such as satellite equipment power on / off in a single comparison.

Description

Technical Field

[0001] The present invention relates to a method for removing outliers from satellite telemetry data based on statistics, and belongs to the technical field of satellite telemetry. Background Art

[0002] Nowadays, the number of satellites in orbit is increasing, the structure and functions of satellites are becoming more and more complex, and the application of various new technologies has also made the telemetry data of satellites more and more subject to interference and coupling. During the transmission of satellite telemetry data, due to the interference of sensors, transducers and radio transmission processes, abnormal jump points will occur in the telemetry data. Such noise data points that deviate from normal data are called outliers.

[0003] The telemetry data of satellites is time series data. Outlier data generally exists in satellite telemetry data and appears randomly. Outlier data is usually a maximum or minimum value, which is significantly different from normal telemetry data. Outlier data can be regarded as a kind of context abnormal data, and its value is significantly different from the data before and after it. The existence of outliers brings interference to the processing and analysis of satellite telemetry data. Removing outliers is one of the important tasks in the preprocessing of satellite telemetry parameters. Currently, there is a lack of an effective method for removing outliers from satellite telemetry data. Summary of the Invention

[0004] The technical problem solved by the present invention is: overcoming the deficiencies of the prior art, and proposing a method for removing outliers from satellite telemetry data based on statistics, effectively removing outliers in the telemetry data, and achieving the purpose of data cleaning.

[0005] The technical solution adopted by the present invention is:

[0006] A method for removing outliers from satellite telemetry data based on statistics, the steps are as follows:

[0007] (1) Store the satellite telemetry data in the form of a time series to form a data table D;

[0008] (2) Initialize the data position time t = M, and initialize the data table size L = |D|;

[0009] (3) When t < L - N, use the method of dividing the deviation by the mean based on a sliding window to calculate the category to which the telemetry variable x ，i of the data x t，i at time t belongs; when t >= L - N, end the operation;

[0010] (4) If the data point is classified as outlier data, then remove it, recalculate the data table size L = |D new |, and execute step (3);

[0011] (5) If the data point is classified as normal data, the number of position records t is increased by 1 and step (3) is executed.

[0012] Furthermore, the deviation divided by the mean method based on the sliding window in step (3) is:

[0013] Using the formula: Calculate telemetry variable x ，i The data x at time t t，i Categories t，i ,in:

[0014] For variable x t,i The absolute value of the deviation divided by the mean in the next m telemetry data;

[0015] For variable x t,i The absolute value of the deviation divided by the mean in the first n telemetry data;

[0016] H is the threshold.

[0017] Furthermore,

[0018] Where μ (subsequent) Represents telemetry parameter x t，i The mean of the next M data points, x t，i -μ (subsequent) is the deviation, and |·| represents the absolute value.

[0019] Furthermore, Where μ (previous) , represents the telemetry parameter x t，i The mean of the previous N data points, x t，i -μ (previous) is the deviation, and |·| represents the absolute value.

[0020] Furthermore,

[0021] Furthermore,

[0022] The beneficial effects of the present invention compared with the prior art are:

[0023] (1) The present invention compares the telemetry variable data with the mean of the subsequent M data and the previous N data at the same time, thereby avoiding the problem of false detection of outliers caused by a single comparison when a telemetry variable changes suddenly due to some reason, such as the switching on and off of satellite equipment;

[0024] (2) The present invention can effectively eliminate outliers in both continuous and discrete variables in satellite telemetry variables. Description of the Drawings

[0025] Figure 1 This is the flowchart of the present invention. Detailed Description of the Invention

[0026] The present invention will be further described below in conjunction with embodiments.

[0027] A method for removing outliers from satellite telemetry data based on a statistical method according to this embodiment is as follows Figure 1 shown, and specifically includes:[[]]

[0028] (1) Store the satellite telemetry data in the form of a time series to form a data table D;

[0029] (2) Initialize the data position time t = M and the data table size L = |D|;

[0030] (3) When t < L - N, use the method of dividing the deviation by the mean based on a sliding window to calculate the category (normal data or outlier data) of the telemetry variable x ，i at time t, which is x t，i ; when t >= L - N, end the operation;

[0031] (4) If the data point is classified as outlier data, then remove it, recalculate the data table size L = |D new |, and execute step (3);

[0032] (5) If the data point is classified as normal data, increment the position record number t by 1 and execute step (3);

[0033] (6) The specific method for detecting outliers in satellite telemetry data using the method of dividing the deviation by the mean based on a sliding window is as follows:

[0034] Use the formula:

[0035] to calculate the classification s ，i of the telemetry variable x t，i at time t, where: t，i Among them:

[0036] In the formula represents the mean of the M data points after the telemetry parameter x t，i , x t，i - μ (subsequent) is the deviation, and |·| represents the absolute value.

[0037] In the formula represents the telemetry parameter x t，iThe mean of the previous N data points, x t，i -μ (previous) is the deviation, and |·| represents the absolute value.

[0038] H represents the threshold, and the symbol & represents the AND operation.

[0039] If s t，i =1, that is, both and Then the telemetry data x t，i For outlier data, if s t，i =0, then the telemetry data x t，i It is normal data.

[0040] The method contains parameters for time window size M and N. If the time window size is set too large, New noise data may be introduced into the calculation. If the time window is set too small, the calculated deviation and mean may be inaccurate. When M=10, N=10 and the threshold H is set to H=2, the effect of outlier removal is better.

[0041] The present invention compares the telemetry variable data with the mean of the subsequent M data and the previous N data at the same time, thus avoiding the problem of wrong detection of wild values ​​caused by a single comparison when a telemetry variable changes suddenly due to some reason, such as the switching on and off of satellite equipment;

[0042] The present invention can effectively eliminate wild values ​​in both continuous variables and discrete variables in satellite telemetry variables.

[0043] Although the present invention has been disclosed as above in the form of a preferred embodiment, it is not intended to limit the present invention. Any person skilled in the art may make possible changes and modifications to the technical solution of the present invention by using the methods and technical contents disclosed above without departing from the spirit and scope of the present invention. Therefore, any simple modifications, equivalent changes and modifications made to the above embodiments based on the technical essence of the present invention without departing from the content of the technical solution of the present invention shall fall within the protection scope of the technical solution of the present invention.

Claims

1. A method for eliminating outliers in satellite telemetry data based on statistics, characterized in that: Here are the steps: (1) Store satellite telemetry data in time series to form data table D; (2) Data location time initialization t=M , data table size initialization L =|D|; (3) When t <L-N When , the deviation divided by the mean method based on the sliding window is used to calculate the telemetry variables in the data table. At the moment t Data The category to which it belongs is calculated; when t>=LN When , the operation ends; (4) If the data point is classified as outlier data, it is removed and the data table size is recalculated. L =| |, execute step (3); (5) If the data point is classified as normal data, the number of position records t Add 1 and execute step (3). The deviation divided by the mean method based on the sliding window in step (3) is: Using the formula: > H ) & > H ) Calculating telemetry variables At the moment t Data Categories ,in: For variable x t,i The absolute value of the deviation divided by the mean in the next m telemetry data; For variable x t,i The absolute value of the deviation divided by the mean in the first n telemetry data; H is the threshold; ， In the formula Indicates telemetry parameters After M The mean of the data points, is the deviation, Indicates absolute value; , where , indicating telemetry parameters Previous N The mean of the data points, is the deviation, Indicates absolute value.

2. The method for eliminating outliers in satellite telemetry data based on statistics according to claim 1, characterized in that: 。 3. The method for eliminating outliers in satellite telemetry data based on statistics according to claim 1, characterized in that: 。

Images