A method for identifying step time of disturbed data based on neighboring average data

Abstract

The present invention provides a method for identifying step time of disturbed data based on neighboring average data, comprising the following steps: (1) reading step test data to obtain a set of disturbed data containing a time series; (2) calculating and extending neighboring average data; (3) comparing the values ​​of the neighboring average data, and taking the time corresponding to the maximum value as the step time. The present invention identifies the time when the step occurs by calculating the neighboring average data of the test data, thereby avoiding the error caused by manual judgment based on experience, and can eliminate the influence of the disturbed data while retaining all the information of the original data, and accurately identify the step time. The present invention is suitable for performance testing and parameter setting of various control systems.

Description

Technical Field

[0001] The invention relates to a method for identifying disturbed test data, in particular to a method for identifying step time of disturbed data based on adjacent average data. Background Art

[0002] When testing parameters of various automatic control systems, the step response characteristic is an important system characteristic. The acquisition of its step response indicators such as rise time and peak time requires accurate acquisition of the step start time. In actual experiments, the step start time is generally operated manually, and the exact time of the step response start time is uncertain. In addition, field tests are affected by various interference factors at the test site, and the test data contains disturbances. Directly using the change amplitude of the data volume as the criterion for the step time will bring large errors.

[0003] In order to eliminate the influence of disturbance, the common data preprocessing method is to filter or curve fit the data to obtain a smoother curve, and interpret the smooth curve to obtain the step response time. However, the phase delay of the filtering algorithm will cause distortion of the time domain data, resulting in errors; the first-order derivative of the original data is discontinuous at the step moment, and the curve fitting algorithm will lose the discontinuous characteristics, resulting in deviations in the results. Summary of the invention

[0004] The present invention provides a method for identifying step time of disturbed data based on neighboring average data, which directly identifies the step starting time for experimental data obtained from original experimental data, achieves accurate interpretation of the experimental data, and improves the accuracy of experimental data interpretation.

[0005] A method for identifying step time of disturbed step data based on adjacent average data comprises the following steps:

[0006] (1) Read the original test data to obtain a set of original data containing disturbances and time series:

[0007] X={t i ,x i},i=1,2,…,n;

[0008] where t i is a time series with a time series interval of Δt; x i is the original experimental data, n is a positive integer greater than or equal to 5;

[0009] (2) Calculate the time t according to the following formula i The average data of the neighborhood at:

[0010] e mi =(x i+m +x i+2m +...+x i+lm+x i-m +x i-2m +...+x i-lm -2lx i )×sign(x i+lm -x i-lm ), where m and l are

[0011] Positive integer, m*l<n, l>1;

[0012] (3) Compare the values ​​of adjacent average data and take the maximum value corresponding to time t s is the step time: t s =t j |(e mj =max{e mi}).

[0013] Furthermore, we calculate the time t i When the neighboring average data is taken at , the value range of m and l is: Where m, l are positive integers, δ m is the maximum disturbance value, and k is the step rise value at the time interval Δt.

[0014] The present invention has the following beneficial effects:

[0015] 1. Compared with the existing technology, more data points are used and more original data information can be obtained;

[0016] 2. Compared with the existing technology, when the m value is constant, higher anti-interference performance can be obtained by increasing the l value;

[0017] 3. With accurate identification results from the data smoothing algorithm, the neighboring average data of each data point utilizes the original data information rather than the processed data information, thus avoiding information loss caused by the data smoothing algorithm. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] Figure 1 It is a schematic diagram of original experimental data of the method for identifying step time of disturbed data based on neighboring average data of the present invention;

[0019] Figure 2 It is a schematic diagram of test data adopted by the present invention;

[0020] Figure 3 It is the adjacent average data curve calculated by the present invention. DETAILED DESCRIPTION

[0021] The technical solution of the present invention will be clearly and completely described below in conjunction with the accompanying drawings of the present invention.

[0022] Figure 1The figure shows the original experimental data schematic diagram of the method for identifying the step time of disturbed data based on the neighboring average data of the present invention, and the method comprises the following steps:

[0023] (1) Read the original test data to obtain a set of original data containing disturbances and time series:

[0024] X={t i ,x i},i=1,2,…,n;

[0025] where t i is an equally spaced time series with a time series interval of Δt; x i is the original experimental data; n is a positive integer greater than or equal to 5.

[0026] (2) Calculate the neighboring average data according to the following formula:

[0027] e mi =(x i+m +x i+2m +...+x i+lm +x i-m +x i-2m +...+x i-lm -2lx i )×sign(x i+lm -x i-lm ), where m, l are positive integers, m*l<n, l>1;

[0028] At a disturbance level no greater than δ m When , since the first-order derivative of the original data without disturbance before and after the step is unchanged, it is easy to know that the maximum absolute value of the neighboring average data of the original data with disturbance before and after the step satisfies:

[0029] max{e mi}<4lδ m

[0030] At the moment of step occurrence, the first-order derivative of the original data is discontinuous. The first-order derivative of the original data before the step is 0, and the first-order derivative of the data after the step is not zero. Taking the above step as an example, assuming that the step rise value at the sampling time interval Δt is k, then the original data at the moment of step occurrence t j The average data near the location is:

[0031]

[0032] when Right now When the step occurs, the average data of the neighborhood at the time of the step is guaranteed to be greater than the maximum value of the average data of the neighborhood at the non-step time 8δ m The same conclusion can be drawn for the next step.

[0033] (3) Compare the values ​​of adjacent average data and take the maximum value corresponding to time t s The moment when the upper step (lower step) occurs:

[0034] t s =t j |(e mj =max{e mi})

[0035] The method is further described below with reference to the accompanying drawings and a specific example:

[0036] A step response test is conducted on a control system. After a step signal is input, the output data is measured as follows: Figure 2 As shown. Read the output data and get a set of original perturbation data containing time series as follows:

[0037] Table 1

[0038]

[0039]

[0040] Calculate the neighboring average data according to the formula:

[0041] e mi =(x i+m +x i+2m +...+x i+lm +x i-m +x i-2m +...+x i-lm -2lx i )×sign(x i+lm -x i-lm )

[0042] Let m be 2 and l be 2, then we have:

[0043] e mi =(x i+2 +x i+4 +x i-2 +x i-4 -4x i )×sign(x i+4 -x i-4 )

[0044] After calculation, the average data of the neighborhood at each moment is as follows:

[0045] Table 2

[0046]

[0047]

[0048] The adjacent average data curve is shown in Figure 3 .

[0049] Compare the values ​​of adjacent average data and take the maximum value:

[0050] max{e mi}=278

[0051] The time t corresponding to this maximum value s If it is 0.21s, then take 0.21s as the starting time of the next step.

[0052] After obtaining the step starting time, we can further calculate indicators such as the rise time and peak time of the step response.

[0053] The present invention proposes a method for identifying step time of disturbed data based on neighboring average data. By calculating the neighboring average data, the influence of the disturbed data can be eliminated while retaining all the information of the original data, and the step moment can be accurately identified.

[0054] The above is only a specific implementation of the present invention, but the protection scope of the present invention is not limited to the step response test of the generator excitation system. The step response test of any automatic control system is included in the technical field of the present invention. The changes or replacements that can be easily thought of by technicians in this technical field within the technical scope disclosed by the present invention should be included in the protection scope of the present invention. Therefore, the protection scope of the present invention should be based on the protection scope of the claims.

Claims

1. A method for identifying step time of disturbed data based on neighboring average data, characterized in that The steps include: (1) Read the output data obtained when performing a step response test on a control system to obtain a set of original disturbance data containing a time series: X={t i ,x i },i=1,2,…,n; where t i is a time series with a time series interval of Δt; x i is the original experimental data, n is a positive integer greater than or equal to 5; (2) Calculate the time t according to the following formula i The average data of the neighborhood at: e mi =(x i+m +x i+2m +...+x i+lm +x i-m +x i-2m +...+x i-lm -2lx i )×sign(x i+lm -x i-lm ), where m and l are Positive integer, m*l<n, l>1; (3) Compare the values ​​of adjacent average data and take the maximum value corresponding to time t s is the step time: t s =t j |(e mj =max{e mi }).

Images