CN101702612A

CN101702612A - Static state modeling method of excitation regulator system of electric generator

Info

Publication number: CN101702612A
Application number: CN200910311214A
Authority: CN
Inventors: 林呈辉
Original assignee: Guizhou Electric Power Test and Research Institute
Current assignee: Guizhou Electric Power Test and Research Institute
Priority date: 2009-12-10
Filing date: 2009-12-10
Publication date: 2010-05-05

Abstract

The invention discloses a static state modeling method of an excitation regulator system of an electric generator. The static state modeling method is characterized by comprising the following steps of: firstly, adding a frequency sweeping sinusoidal signal at the input end of a link to be measured of an excitation regulator by using a dynamic signal analyzer, and measuring the frequency response characteristic of the output end; and secondly, carrying out curve fitting on the actually-measured frequency response characteristic by using an MATLAB emulator program and combining a minimum square law, and establishing an actually-measuring model. With the method, the actually-measuring model of various links of an excitation system is established, is and more accurate than a prototype model provided by manufacturers and better reflects the actual condition of the excitation system. As the number of units of the tested electric generators is more, the method can greatly save the calculation time. The static state modeling method is suitable for static state modeling calculation of the excitation system in wide range and provides scientific basis for establishing excitation system databases for grids.

Description

Static state modeling method of excitation regulator system of electric generator

Technical field

The present invention relates to a kind of static state modeling method in excitation regulator system of electric generator.

Background technology

According to " actual measurement of CHINA SOUTHERN POWER synchronous generator excited system parameter and modeling administrative provisions (trying) " requirement, need survey and modeling work the unit excitation system parameter of generating electricity by way of merging two or more grid systems of south electric network, test comprises static test and dynamic test.Each link in the static test reply field regulator model of generator excited system model parameter is measured and identification.Because excitation system is subjected to factor affecting such as sampling error, Algorithm Error, hardware non-ideal characteristic, excitation system archetype that producer provides and the model that obtains by the frequency domain method actual measurement exist bigger deviation.South electric network excitation system database needs a comparatively accurate computation model, is applied in the engineering practice.

Summary of the invention

The technical problem to be solved in the present invention is, a kind of static state modeling method of excitation regulator system of electric generator is provided, and sets up the computation model of a realistic excitation system, with overcome deviation that prior art exists big wait not enough.

Technical scheme of the present invention is: it comprises following process: at first, utilize dynamic signal analyzer to add sweep sine at field regulator link input to be measured, measure the frequency response characteristic of output; Then, utilize the MATLAB simulated program to carry out curve fitting, set up Model Measured in conjunction with the frequency response characteristic of least square method to actual measurement.

Least square method utilizes the quadratic sum minimum of Model Measured and measured data deviation can guarantee that each deviation can be very not big, match practical frequency response well.And MATLAB self integrated least square method, only need a function call can realize curve fit, very easy to use.

Beneficial effect of the present invention: set up the Model Measured of each link of excitation system by the method, more accurate than the archetype that producer provides, more can reflect the actual conditions of excitation system.Because the generator unit quantity of test is more, adopt the method to save computing time greatly.Be applicable to the static Modeling Calculation of large-scale excitation system, for south electric network is set up the foundation that the excitation system database provides science.

Description of drawings

Fig. 1 is frequency response measured result and the archetype frequency response schematic diagram of pure integral element Kd=2;

Fig. 2 is frequency response measured result and the Model Measured frequency response schematic diagram of pure integral element Kd=2.

In the accompanying

drawing

1,2 and in the table 1,2, P and Q represent the gain response and the phase response of frequency response measured result respectively, P1 and Q1 represent the gain response and the phase response of the archetype parameter frequency response that producer provides respectively, and P2 and Q2 represent the gain response and the phase response of Model Measured parameter frequency response respectively.The unit of P, P1 and P2 is (V/V), is equivalent to a numerical value; The unit of Q, Q1 and Q2 for the degree (°).The frequency values of the corresponding swept-frequency signal in frequency hurdle, unit is Hz; (P-P1)/and P and (P-P2)/P is the gain relative error, unit is percentage value (%).

Embodiment

Embodiments of the invention: the pure differential link with certain power plant's excitation system is an example.The archetype transfer function that producer provides is KdS, wherein definite value Kd=2.Adopt dynamic signal analyzer that pure integral element is carried out 0-8Hz sinusoidal signal frequency sweep, obtain the frequency response measurement result, the archetype parameter frequency response that provides with producer compares.

Actual frequency response measurement result and archetype result of calculation such as table 1 and shown in Figure 1.As can be seen from Table 1, the maximum gain relative error is 12.79%, occurs in 0.09Hz; The maximum phase error is-30.24 °, occurs in 8.01Hz.The frequency response that the archetype parameter is described is compared with measured data, and gain relative error and phase error are bigger, can not be used for Practical Calculation.

The practical frequency response of the pure integral element Kd=2 of table 1 and archetype frequency response are relatively

For the archetype that provides with producer compares, the actual frequency response measurement result of table 2 and Fig. 2 is identical with table 1 and Fig. 1, obtains by the dynamic signal analyzer actual measurement.Model Measured is by the matlab simulated program measured data to be carried out curve fitting to obtain.Actual frequency response measurement result and Model Measured result of calculation such as table 2 and shown in Figure 2.Through calculating the actual measurement computation model is KdS/ (1+T1S), Kd=2.22 wherein, T1=0.01.The maximum gain relative error is 6.12% as can be seen from Table 2, occurs in 0.50Hz; The maximum phase error is 2.89 °, occurs in frequency 1.01Hz.Gain relative error and phase error that Model Measured is described are less, all in the error allowed band, can be used for the engineering practical application.

The practical frequency response of the pure integral element Kd=2 of table 2 and Model Measured frequency response are relatively

Claims

1. static state modeling method of excitation regulator system of electric generator, it is characterized in that: it comprises following process: at first, utilize dynamic signal analyzer to add sweep sine at field regulator link input to be measured, measure the frequency response characteristic of output; Then, utilize the MATLAB simulated program to carry out curve fitting, set up Model Measured in conjunction with the frequency response characteristic of least square method to actual measurement.