Time series data analysis method based on variable benchmark

Abstract

The invention relates to a variable benchmark-based time sequence data analysis method, which comprises the following steps of: (1) defining basic parameters required by time sequence data analysis, including a time sequence quantity, a descending value, an ascending value, an abnormal value, a rising edge, a descending edge, a time sequence state change mark and a time sequence data list value; (2) analyzing required variable parameters based on variable reference time sequence data; (3) determining structural body parameters required by analysis based on variable reference time sequence data; (4) defining a time sequence data analysis rule; and (5) analyzing the time series data. According to the method, the ascending value and the descending value of the time sequence data can be flexibly configured, meanwhile, an association rule between different time sequence data is supported to be formulated, the reference value is defined for each time sequence, the reference value serves as an analysis reference parameter of other time sequence variables, and time sequence data analysis logic under various complex conditions is met.

Description

Technical Field

[0001] This invention relates to a time-series data parsing method based on a variable benchmark, belonging to the field of data processing technology. Background Technology

[0002] Currently, time series data plays an increasingly important role in aerospace systems. Traditionally, time series data is defined and parsed using a single bit, where 0 represents a falling edge and 1 represents a rising edge. The value of this bit is used to determine the state of the time series data. However, as the types of time series data increase and processing methods become more complex, some time series data use state values ​​(e.g., 0xAAAA) to indicate the arrival of a rising or falling edge. Therefore, a single bit cannot meet the parsing requirements for this type of time series. A standard data type (e.g., uint32) is needed to represent a specific time series parameter, and the values ​​at the arrival times of the rising and falling edges (i.e., rising and falling values) need to be flexibly defined. Furthermore, for more complex time series, different time series parameters have correlations, with one time series parameter using certain states of another as a reference. Therefore, a method for representing the correlations between different time series needs to be studied. Summary of the Invention

[0003] The technical problem solved by this invention is to overcome the shortcomings of the prior art and provide a time series data parsing method based on a variable benchmark to meet the needs of time series data parsing under complex and special conditions. The technical solution of this invention is: A time-series data parsing method based on a variable baseline includes the following steps: (1) Define the basic parameters required for time series data parsing, including time series quantity, falling value, rising value, outlier, rising edge, falling edge, time series state change flag, and time series data list value; (2) Define the variable parameters required for parsing time series data based on the variable reference, including the time series variable reference value outVariable, the time series variable update flag updateFlag, and the expression; (3) Define the structure parameters required for parsing time series data based on variable reference, including the time series variable reference information structure and the time series variable information structure; (4) Define the time series data parsing rules, which include: the method for extracting the initial values ​​of time series variables from the original binary data, and the method for time series determination; Extracting the initial value of timing variables refers to extracting the field value corresponding to the timing data from the received raw binary data according to predetermined rules. The predetermined rules include the truncation of the nth byte, the length of the truncation byte, and whether it is big-endian or little-endian. The parameters related to timing determination include the timing rising edge reference value cUpSequenceValue and the timing falling edge reference value cLowSequenceValue. (5) Perform time series data parsing.

[0004] Furthermore, time series quantities are a class of physical quantities whose values ​​change between increasing and decreasing values ​​over time; the value of a time series quantity is 0 or 1, or can be configured as a positive integer in decimal / hexadecimal; when the value of a time series quantity is a positive integer, the value of the time series quantity is divided into: decreasing value, increasing value, and outlier value; the default value of a time series quantity is a decreasing value. Falling value: Represents the value at the falling edge of a timing quantity; the falling value can be configured as a positive integer in decimal / hexadecimal. Rising value: Represents the value at the rising edge of a timing quantity; rising values ​​can be specified as positive integers in decimal / hexadecimal. Outliers: Any positive integer other than increasing or decreasing values ​​is considered an outlier. Timing state change flag: indicates whether a timing value is a rising edge or a falling edge; Time series data list value: Used to record all time series data values; The rising edge is the start time. When a value changes from a falling value to a rising value, it is considered a rising edge. The original timestamp of the rising value is used as the rising edge. The first rising value of a time series is considered a rising edge. The falling edge, or end time, is defined as the transition from an increasing value to a decreasing value. The original timestamp of the decreasing value is used as the falling edge. The first decreasing value of a time series is not considered a falling edge.

[0005] Furthermore, the time series variable reference value outVariable is used as a parameter for parsing and calculating other time series data; each time series variable reference value outVariable is related to the arrival time of a certain time series variable; Each time series variable baseline value outVariable corresponds to a time series variable update flag updateFlag. The content of the updateFlag field determines when the corresponding time series variable baseline value outVariable is updated, that is, whether the time series variable baseline value is updated when the time series data arrives for the first time, or when the time series variable baseline value is updated every time time series data is received. The time series data is parsed and calculated using the expression, which references the baseline values ​​of other time series variables.

[0006] Furthermore, the time-series variable reference information structure includes: strQName, a string type, represents the time series reference name; nValue, a floating-point number, represents the base value of a time-series variable.

[0007] Furthermore, the time-series variable information structure includes: strQName, a string type, represents the timing name; nValue, an integer type, represents a 0 / 1 type time series value, and its value is calculated from the original data according to the time series data parsing rules; dUpValue, a floating-point number, represents a non-zero / one time-series value, which is the calculated acquisition time. `updateFlag`, a boolean value, is a flag for updating timing variables. `true` means that the value is updated at the top edge of each frame, and `false` means that the value is updated only at the top edge of the first frame. cUpSequenceValue, an integer, is the timing edge reference value. It is optional and used for non-0 / 1 timing. cLowSequenceValue, an integer, is the timing lower edge reference value. It is optional and used for non-0 / 1 timing.

[0008] Furthermore, in the time-series variable information structure, the variables strQName, updateFlag, cUpSequenceValue, and cLowSequenceValue are defined in the time-series data parsing rules, while the variables nValue and dUpValue are calculated and assigned during the time-series data parsing process.

[0009] Furthermore, the step of performing time-series data parsing specifically includes: (1) Initialize the time series variable baseline value list lstSequenceBase, which is used to record the baseline value information of each time series variable; (2) Extract the initial values ​​of the time series variables; If the input data type is a floating-point number, the acquisition time is directly used as the initial value of the timing variable; otherwise, the initial value of the original timing data is extracted from the original binary data value: first, according to the starting byte and byte length required in the timing data parsing rules, the corresponding byte content is extracted; then, the data is converted to endianness according to the big-endian or little-endian mode defined in the parsing rules; finally, the data after the endianness conversion is converted according to the required data type to obtain the final required initial value of the timing variable. (3) Perform timing determination on the initial values ​​of the timing variables obtained in step (2): 3.1) If the current timing value is inconsistent with both the configured timing rising edge reference value cUpSequenceValue and the timing falling edge reference value cLowSequenceValue, then this timing value is considered invalid and an outlier, and is ignored. The timing rising edge reference value is the rising value, and the timing falling edge reference value is the falling value. 3.2) If the current timing value is the same as the previous timing value, since there is no transition for the same timing value, this timing value will not be processed. 3.3) Determine whether the current timing value is a timing rising edge reference value or a timing falling edge reference value, and update the timing status change flag and the timing data list value; (4) Calculate the acquisition time value of the current timing sequence and replace the timing value with the acquisition time for the calculation of the subsequent timing width value; (5) Update the baseline value of the time series variable outVariable: 5.1) The timing variable baseline value is updated only on the rising edge, and not on the falling edge; determine the current timing state change flag. If the current timing is a rising edge, proceed to step 5.2). 5.2) Traverse the time series variable baseline value list lstSequenceBase and search for the current time series variable baseline value outVariable; if found, proceed to step 5.3); otherwise, proceed to step 5.5. 5.3) Based on the timing variable update flag updateFlag, determine whether to update at every timing edge or in the first frame timing edge update mode. true means to update at every frame edge, false means to update only at the first frame edge. 5.4) If the time series variable update flag is true, update the baseline value of the time series variable with the current acquisition time; if it is false, do not update. 5.5) If the baseline value of the time series variable is not found, and this is the first time the time series variable has been found, insert the time series value into the time series variable baseline value list lstSequenceBase for use in the calculation of subsequent time series values; (6) Calculate the final result value of the timing variable according to the expression, where the value on the rising edge is the calculated value and the value on the falling edge is the width; 6.1) Traverse the time series variable baseline value list lstSequenceBase, and check whether the time series variable baseline parameter in the expression of the current time series has a time series variable baseline value outVariable in the time series variable baseline value list. If it exists, replace the time series variable baseline parameter in the expression with the time series variable baseline value outVariable of the other time series and perform calculation. An expression includes one or more time series variable baseline parameters. 6.2) After replacing all time-series variable baseline parameters, calculate the expression result based on the replaced numerical expression; 6.3) If the timing is a rising edge, update the timing variable information value dUpValue with the result calculated in step 6.2); if the timing is a falling edge, subtract the timing variable information value dUpValue of the previous rising edge from the expression result calculated in step 6.2) and use it as the width. (7) After steps (1) to (6), the final result of time series data parsing based on variable benchmark is obtained.

[0010] Secondly, the present invention also proposes a processor for running a program, wherein the program executes the method during runtime.

[0011] Thirdly, the present invention also proposes a non-volatile storage medium comprising: a computer program product, wherein the method is executed when the computer program product is executed.

[0012] Fourthly, the present invention also proposes a computer program product, which includes a computer program that, when executed by a processor, implements the method described.

[0013] The advantages of this invention compared to the prior art are: (1) The present invention flexibly configures the values ​​(i.e. rising value and falling value) at the arrival time of the rising and falling edges of different time series data. (2) The present invention can flexibly configure the timing reference value of different timing data and can set multiple temporary variables of timing reference value. Each reference value can be updated according to the configuration method when the first frame of timing data is received, or updated every time the timing data is received (i.e., whether the reference value of each timing quantity is updated every time). (3) The present invention supports expression configuration for parsing time series data. In the expression configuration, the baseline values ​​of other time series variables are introduced to realize the correlation parsing between different time series. Attached Figure Description

[0014] Figure 1 This is a flowchart of the time series data parsing process based on a variable benchmark according to the present invention. Detailed Implementation

[0015] The specific embodiments of the present invention will now be described in further detail with reference to the accompanying drawings.

[0016] This invention relates to a time series data parsing method based on a variable benchmark, which can flexibly configure the rising and falling values ​​of time series data, and supports the formulation of association rules between different time series data. By defining a benchmark value for each time series and using the benchmark value as the parsing benchmark parameter for other time series variables, the time series data parsing logic is satisfied under various complex conditions.

[0017] This invention proposes a time-series data parsing method based on a variable benchmark, comprising the following steps: (1) Define the basic parameters required for time series data parsing, including time series quantity, falling value, rising value, outlier, rising edge, falling edge, time series state change flag, and time series data list value; (2) Define the variable parameters required for parsing time series data based on the variable reference, including the time series variable reference value outVariable, the time series variable update flag updateFlag, and the expression; (3) Define the structure parameters required for parsing time series data based on variable reference, including the time series variable reference information structure and the time series variable information structure; (4) Define the time series data parsing rules, which include: the method for extracting the initial values ​​of time series variables from the original binary data, and the method for time series determination; Extracting the initial value of timing variables refers to extracting the field value corresponding to the timing data from the received raw binary data according to predetermined rules. The predetermined rules include the truncation of the nth byte, the length of the truncation byte, and whether it is big-endian or little-endian. The parameters related to timing determination include the timing rising edge reference value cUpSequenceValue and the timing falling edge reference value cLowSequenceValue. (5) Perform time series data parsing.

[0018] The following is a detailed description.

[0019] 1. Define the basic parameter values ​​required for time series data parsing, as shown in the table below:

[0020] 2. Define the variable parameters required for parsing time series data based on a variable baseline: including the baseline value of the time series variable (outVariable), the update flag of the time series variable (updateFlag), and the expression.

[0021] 1) Define temporary variables for time series data, namely the time series variable baseline value outVariable, which serves as the baseline value for parsing and calculating other time series data; each time series variable baseline value outVariable is related to the arrival time of a certain time series variable; 2) Define the update rule for the baseline value of the timing variable by using the updateFlag flag, i.e., whether to update the baseline value of the timing variable when the first frame of timing data arrives, or to update the baseline value of the timing variable every time the timing data is received; Each time series variable baseline value outVariable corresponds to a time series variable update flag updateFlag. The content of the updateFlag field determines when the corresponding time series variable baseline value outVariable is updated, that is, whether the time series variable baseline value is updated when the time series data arrives for the first time, or when the time series variable baseline value is updated every time time series data is received. 3) Time series data is parsed using the expression, which references the baseline values ​​of other time series variables;

[0022] 3. Define the structure parameters required for parsing time series data based on variable reference: including the time series variable reference information structure and the time series variable information structure.

[0023] 1) Time series variable baseline information structure:

[0024] 2) Time-series variable information structure:

[0025] Among them, the variables strQName, updateFlag, cUpSequenceValue, and cLowSequenceValue are defined in the time series data parsing rules, while the variables nValue and dUpValue are calculated and assigned during the time series data parsing process.

[0026] 4. Time series data parsing process: (1) Initialize the time series variable baseline value list lstSequenceBase, which is used to record the baseline value information of each time series variable; (2) Extracting the initial values ​​of time series variables If the input data type is a floating-point number, the acquisition time is directly used as the initial value of the timing variable; otherwise, the initial value of the original timing data is extracted from the original binary data value: first, according to the starting byte and byte length required in the timing data parsing rules, the corresponding byte content is extracted; then, the data is converted to endianness according to the big-endian or little-endian mode defined in the parsing rules; finally, the data after the endianness conversion is converted according to the required data type to obtain the final required initial value of the timing variable.

[0027] (3) Perform timing determination on the initial values ​​of the timing variables obtained in step (2): 1) If the current timing value is inconsistent with both the configured timing rising edge reference value cUpSequenceValue (rising value) and timing falling edge reference value cLowSequenceValue (falling value), then this timing value is considered invalid and an outlier, and is ignored. 2) If the current timing value is the same as the previous timing value, since there is no transition for the same timing value, this timing value will not be processed. 3) Determine whether the current timing value is rising or falling, and update the timing state change flag (indicating whether the timing value is rising or falling) and update the timing data list value (used to record timing data information); (4) Calculate the acquisition time value of the current timing sequence and replace the timing value with the acquisition time for the calculation of the subsequent timing width value; (5) Update the baseline values ​​of time series variables: 5.1) The timing variable baseline value is updated only on the rising edge, and not on the falling edge; determine the current timing state change flag. If the current timing is a rising edge, proceed to step 5.2). 5.2) Traverse the time series variable baseline value list lstSequenceBase and search for the current time series variable baseline value outVariable; if found, proceed to step 5.3); otherwise, proceed to step 5.5. 5.3) Based on the timing variable update flag updateFlag, determine whether to update at every timing edge or in the first frame timing edge update mode. true means to update at every frame edge, false means to update only at the first frame edge. 5.4) If the time series variable update flag is true, update the baseline value of the time series variable with the current acquisition time; if it is false, do not update. 5.5) If the baseline value of the time series variable is not found, and this is the first time the time series variable has been found, insert the time series value into the time series variable baseline value list lstSequenceBase for use in the calculation of subsequent time series values; (6) Calculate the final result value of the time series variable according to the expression (the value on the rising edge is the calculated value, and the value on the falling edge is the width). 6.1) Traverse the time series variable baseline value list lstSequenceBase, and check whether the time series variable baseline parameter in the expression of the current time series has a time series variable baseline value outVariable in the time series variable baseline value list. If it exists, replace the time series variable baseline parameter in the expression with the time series variable baseline value outVariable of the other time series and perform calculation. An expression includes one or more time series variable baseline parameters. 6.2) After replacing all time-series variable baseline parameters, calculate the expression result based on the replaced numerical expression; 6.3) If the timing is a rising edge, update the timing variable information value dUpValue with the result calculated in step 6.2); (the value is only updated on the rising edge, for the purpose of calculating the timing width) If the timing is a falling edge, subtract the timing variable information value dUpValue of the previous rising edge from the expression result calculated in step 6.2), and use it as the width; (7) After the above steps, the final result of time series data parsing based on the variable baseline is obtained. This result can be used for subsequent processing operations on time series data, such as calculating the time series width and comparing the upper and lower limits of time series data.

[0028] The features and advantages of this invention are as follows: 1. Flexible configuration of the values ​​(i.e. rising and falling values) at the arrival times of the rising and falling edges of different time series data. 2. Flexible configuration of timing reference values ​​for different timing data. Multiple temporary variables of timing reference values ​​can be set. Each reference value can be updated based on the configuration method, whether it is updated when the first frame of timing data is received or every time the timing data is received (i.e., whether the reference value of each timing quantity is updated every time). 3. The parsing of time series data supports expression configuration, which allows the introduction of baseline values ​​for other time series variables to achieve correlation parsing between different time series.

[0029] The parts of this invention not described in detail are common knowledge to those skilled in the art.

Claims

1. A time-series data parsing method based on a variable benchmark, characterized in that... include: (1) Define the basic parameters required for time series data parsing, including time series quantity, falling value, rising value, outlier, rising edge, falling edge, time series state change flag, and time series data list value; (2) Define the variable parameters required for parsing time series data based on the variable reference, including the time series variable reference value outVariable, the time series variable update flag updateFlag, and the expression; (3) Define the structure parameters required for parsing time series data based on variable reference, including the time series variable reference information structure and the time series variable information structure; (4) Define the time series data parsing rules, which include: the method for extracting the initial values ​​of time series variables from the original binary data, and the method for time series determination; Extracting the initial value of timing variables refers to extracting the field value corresponding to the timing data from the received raw binary data according to predetermined rules. The predetermined rules include the truncation of the nth byte, the length of the truncation byte, and whether it is big-endian or little-endian. The parameters related to timing determination include the timing rising edge reference value cUpSequenceValue and the timing falling edge reference value cLowSequenceValue. (5) Perform time series data parsing.

2. The time series data parsing method based on a variable reference according to claim 1, characterized in that: Time series quantity: A type of physical quantity whose value changes between increasing and decreasing values ​​over time; the value of a time series quantity is 0 or 1, or can be configured as a positive integer in decimal / hexadecimal; when the value of a time series quantity is a positive integer, the value of the time series quantity is divided into: decreasing value, increasing value, and outlier value; the default value of a time series quantity is a decreasing value; Falling value: Represents the value at the falling edge of a timing quantity; the falling value can be configured as a positive integer in decimal / hexadecimal. Rising value: Represents the value at the rising edge of a timing quantity; rising values ​​can be specified as positive integers in decimal / hexadecimal. Outliers: Any positive integer other than increasing or decreasing values ​​is considered an outlier. Timing state change flag: indicates whether a timing value is a rising edge or a falling edge; Time series data list value: Used to record all time series data values; The rising edge is the start time. When a value changes from a falling value to a rising value, it is considered a rising edge. The original timestamp of the rising value is used as the rising edge. The first rising value of a time series is considered a rising edge. The falling edge, or end time, is defined as the transition from an increasing value to a decreasing value. The original timestamp of the decreasing value is used as the falling edge. The first decreasing value of a time series is not considered a falling edge.

3. The time series data parsing method based on a variable benchmark according to claim 1, characterized in that: The time series variable reference value outVariable is used as a parameter for parsing and calculating other time series data; each time series variable reference value outVariable is related to the arrival time of a certain time series variable; Each time series variable baseline value outVariable corresponds to a time series variable update flag updateFlag. The content of the updateFlag field determines when the corresponding time series variable baseline value outVariable is updated, that is, whether the time series variable baseline value is updated when the time series data arrives for the first time, or when the time series variable baseline value is updated every time time series data is received. The time series data is parsed and calculated using the expression, which references the baseline values ​​of other time series variables.

4. The time series data parsing method based on a variable benchmark according to claim 1, characterized in that: The time-series variable reference information structure includes: strQName, a string type, represents the time series reference name; nValue, a floating-point number, represents the base value of a time-series variable.

5. The time series data parsing method based on a variable benchmark according to claim 1, characterized in that: The time-series variable information structure includes: strQName, a string type, represents the timing name; nValue, an integer type, represents a 0 / 1 type time series value, and its value is calculated from the original data according to the time series data parsing rules; dUpValue, a floating-point number, represents a non-zero / one time-series value, which is the calculated acquisition time. updateFlag, a boolean type, is a flag for updating timing variables. true means that it is updated at the top edge of every frame, and false means that it is updated only at the top edge of the first frame. cUpSequenceValue, an integer, is the timing edge reference value. It is optional and used for non-0 / 1 timing. cLowSequenceValue, an integer, is the timing lower edge reference value. It is optional and used for non-0 / 1 timing.

6. The time series data parsing method based on a variable benchmark according to claim 5, characterized in that: In the time-series variable information structure, the variables strQName, updateFlag, cUpSequenceValue, and cLowSequenceValue are defined in the time-series data parsing rules, while the variables nValue and dUpValue are calculated and assigned during the time-series data parsing process.

7. The time series data parsing method based on a variable benchmark according to claim 5, characterized in that: The process of parsing time-series data specifically includes: (1) Initialize the time series variable baseline value list lstSequenceBase, which is used to record the baseline value information of each time series variable; (2) Extract the initial values ​​of the time series variables; If the input data type is a floating-point number, the acquisition time is directly used as the initial value of the timing variable; otherwise, the initial value of the original timing data is extracted from the original binary data value: first, according to the starting byte and byte length required in the timing data parsing rules, the corresponding byte content is extracted; then, the data is converted to endianness according to the big-endian or little-endian mode defined in the parsing rules; finally, the data after the endianness conversion is converted according to the required data type to obtain the final required initial value of the timing variable. (3) Perform timing determination on the initial values ​​of the timing variables obtained in step (2): 3.1) If the current timing value is inconsistent with both the configured timing rising edge reference value cUpSequenceValue and the timing falling edge reference value cLowSequenceValue, then this timing value is considered invalid and an outlier, and is ignored. The timing rising edge reference value is the rising value, and the timing falling edge reference value is the falling value. 3.2) If the current timing value is the same as the previous timing value, since there is no transition for the same timing value, this timing value will not be processed. 3.3) Determine whether the current timing value is a timing rising edge reference value or a timing falling edge reference value, and update the timing status change flag and the timing data list value; (4) Calculate the acquisition time value of the current timing sequence and replace the timing value with the acquisition time for the calculation of the subsequent timing width value; (5) Update the baseline value of the time series variable outVariable: 5.1) The timing variable baseline value is updated only on the rising edge, and not on the falling edge; determine the current timing state change flag. If the current timing is a rising edge, proceed to step 5.2). 5.2) Traverse the time series variable baseline value list lstSequenceBase and search for the current time series variable baseline value outVariable; if found, proceed to step 5.3); otherwise, proceed to step 5.

5. 5.3) Based on the timing variable update flag updateFlag, determine whether to update at every timing edge or in the first frame timing edge update mode. true means to update at every frame edge, false means to update only at the first frame edge. 5.4) If the time series variable update flag is true, update the baseline value of the time series variable with the current acquisition time; if it is false, do not update. 5.5) If the baseline value of the time series variable is not found, and this is the first time the time series variable has been found, insert the time series value into the time series variable baseline value list lstSequenceBase for use in the calculation of subsequent time series values; (6) Calculate the final result value of the timing variable according to the expression, where the value on the rising edge is the calculated value and the value on the falling edge is the width; 6.1) Traverse the time series variable baseline value list lstSequenceBase, and check whether the time series variable baseline parameter in the expression of the current time series has a time series variable baseline value outVariable in the time series variable baseline value list. If it exists, replace the time series variable baseline parameter in the expression with the time series variable baseline value outVariable of the other time series and perform calculation. An expression includes one or more time series variable baseline parameters. 6.2) After replacing all time-series variable baseline parameters, calculate the expression result based on the replaced numerical expression; 6.3) If the timing is a rising edge, update the timing variable information value dUpValue with the result calculated in step 6.2); if the timing is a falling edge, subtract the timing variable information value dUpValue of the previous rising edge from the expression result calculated in step 6.2) and use it as the width. (7) After steps (1) to (6), the final result of time series data parsing based on variable benchmark is obtained.

8. A processor, characterized in that, The processor is used to run a program, wherein the program executes the method according to any one of claims 1 to 7 when it runs.

9. A non-volatile storage medium, characterized in that, include: A computer program product that, when executed, performs the method described in any one of claims 1 to 7.

10. A computer program product, characterized in that, The computer program product includes a computer program that, when executed by a processor, implements the steps of the method according to any one of claims 1 to 7.

Images