CN102854465A - System and method for chaotic prediction of DFIG (doubly fed induction generator) running state based on phase-space reconstruction - Google Patents

Abstract

The invention discloses a system for chaotic prediction of a DFIG (doubly fed induction generator) running state based on phase-space reconstruction. The system is characterized by comprising a wind generation set, a test instrument and an upper computer with a chaotic prediction program. A working method of the system comprises the steps of acquiring a signal, filtering, correcting, calculating parameters, chaotically predicating, and predicating a wind turbine state. The system has the advantages as follows: (1) hardware devices are simple and practical; (2) data is automatically protected after power failure, chaotic real-time detection is continuous, and reliability is high; (3) measuring precision is high; (4) the real-time performance and the reliability of the system are high, requirements of monitoring a state of wind generation set, researching a transitional process, diagnosing and predicting a fault, and the like are met, and a high practical value is obtained.

Description

A kind of DFIG running status chaotic prediction system and method based on phase space reconfiguration

(1) technical field:

The invention belongs to the wind power system electric powder prediction, particularly English spelling---the double fed induction generators of a kind of DFIG(based on phase space reconfiguration) running status chaotic prediction system and method.

(2) background technology:

Along with the continuous increase of wind power generating set (mainly the being the double fed induction generators group) capacity that accesses electrical network, large such as extraneous difference variation because the service condition of wind-powered electricity generation unit is abominable, the wind speed variation is random etc.These uncertain extraneous factors cause the failure rate of wind-powered electricity generation unit high, make wind energy turbine set later stage operation expense high, and the peculiar intermittence of wind-power electricity generation itself and randomness have increased the instability with its interconnected network significantly.As the Coupling nonlinear system of a complexity, the safe and reliable operation of double fed induction generators group directly affects steady load reasonable distribution and the grid supply quality with its interconnected network.Therefore, improve the operational reliability of double fed induction generators group, safe and high quality operation and the raising system economy that ensures electrical network had significant role.For the safe and stable operation of realizing the double fed induction generators group judges that in time the performance condition of unit parts and development trend improve its operational efficiency, the present invention adopts several different methods and means to the vitals (gear case of unit, generator etc.) carry out on-line monitoring and analysis, have on the basis of chaos attribute in analysis double fed induction generators group running status, utilize State Space Reconstruction to set up the weighing first order local area forecast model of operating states of the units, its running status is predicted, in order to find in advance failure symptom, avoid and alleviate serious device damage, determine rational maintenance time and scheme, thereby reach the purpose that significantly reduces maintenance cost.

(3) summary of the invention:

The object of the present invention is to provide a kind of DFIG running status chaotic prediction system and method based on phase space reconfiguration, it can overcome the deficiencies in the prior art, is a kind of simple in structure, the system and method that can predict system's operation trend that can recover chaotic attractor.

Technical scheme of the present invention: a kind of DFIG running status chaotic prediction system based on phase space reconfiguration is characterized in that it comprises wind-powered electricity generation unit, tester and with the host computer of chaotic prediction program; Wherein, described tester gathers the signal of wind-powered electricity generation unit, is two-way with host computer with the chaotic prediction program and is connected.

Described tester is by signals collecting and modulate circuit unit, A/D conversion circuit unit, single-chip microcomputer, usb circuit unit, data storage circuitry unit and timing circuit cell formation; Wherein, described signals collecting and unit, modulate circuit unit gather the signal of wind-powered electricity generation unit, and its output terminal connects the input end of A/D conversion circuit unit; Described single-chip microcomputer is with A/D conversion circuit unit, usb circuit unit, data storage circuitry unit and be connected circuit unit and be respectively two-way the connection; Described usb circuit unit is two-way with host computer with the chaotic prediction program and is connected.

The signal of described collection wind-powered electricity generation unit is to gather wind-driven generator group wheel box inboard bearing temperature signal, aerogenerator winding maximum temperature signal, wind power generator rotor mean speed signal and aerogenerator active power parameter signal.

Described signals collecting and modulate circuit unit are made of sensor, data collecting card, resistance R s, filtering circuit, voltage follower, modulate circuit and stabilivolt; It is connected to conventional the connection; Wherein said sensor adopts the isolation template, input signal all is converted to the standard voltage signal of 5V; Described data collecting card is 12 multifunctional data acquisition cards of PCI-1711 that grind magnificent company, have 16 tunnel single-ended analog inputs, 8 data signal channels are with an automatic channel/gain scan circuit, automatically control multi-channel gating switch during sampling, it is connected to conventional the connection.

Described A/D conversion circuit unit is made of conversion chip and peripheral circuit; Wherein said conversion chip be adopt CMOS technique, be that ternary data output latch is arranged in the sheet, input mode is single channel, be 100 μ s switching time, supply voltage is+8 conversion chip ADC0804 of successive approximation of 5V; Described conversion chip ADC0804 comprises pin DB0, pin DB1, pin DB2, pin DB3, pin DB4, pin DB5, pin DB6, pin DB7, pin/WR, pin/RD, pin/CS, pin VIN (+), pin VIN (-), pin C LK-IN, pin CLK-R and pin Vref/2, and is wait time-delay mode with singlechip chip and is connected according to pin DB0, pin DB1, pin DB2, pin DB3, pin DB4, pin DB5, pin DB6, pin DB7, pin/WR, pin/RD, pin/CS; Described peripheral circuit is comprised of capacitor C 28, resistance R 32, two resistance R 33, capacitor C 29, power supply VCC; The signal of described pin VIN (+) after capacitor C 28 and resistance R 33 receiving signal reason processing of circuit, capacitor C 28 is connected common ground with these resistance R 33 tie points with pin VIN (-), takes differential voltage analog input mode; Described pin CLK-R is through another resistance R 33 and capacitor C 29 ground connection, and pin CLK-IN connects the tie point of this resistance R 33 and capacitor C 29; Described pin Vref/2 meets power supply VCC through resistance R 32.

Described single-chip microcomputer adopts the single-chip microcomputer MC9S12DP256 of Freescale.

Described data storage circuitry unit adopts the DS1225 chip of Dallas company.

Described usb circuit unit adopts the ooze CH372 chip of permanent electronics of Nanjing.

Described timing circuit unit adopts the PIC16F716 device with house dog.

A kind of method of work of the chaotic prediction system for wind power system is characterized in that it may further comprise the steps:

⑴ arrange the acquisition interval timing by the timing circuit unit, comes Real-time Collection gearbox of wind turbine inboard bearing temperature, generator windings maximum temperature, rotor mean speed and generator active power signal by signals collecting and modulate circuit unit;

⑵ the signal that gathers among the step ⑴ is through signals collecting and modulate circuit unit and the A/D conversion circuit unit carries out filtering, self calibration is processed, and by single-chip microcomputer with set gear box inboard bearing temperature, generator windings maximum temperature, rotor mean speed and generator active power parameter are input in the data storage circuitry unit;

⑶ gearbox of wind turbine inboard bearing temperature, generator windings maximum temperature after process step ⑵ by usb circuit, rotor mean speed and generator active power data transmission are to being equipped with in the host computer of realizing the chaotic prediction program;

⑷ the Methods of Chaotic Forecasting that use in the host computer is carried out calculation of parameter and processing.

Methods of Chaotic Forecasting among the described step ⑷ is to adopt the method based on phase space reconfiguration to detect chaos, is made of following steps:

1. determine to embed dimension and time delay with the C-C method: during running of wind generating set, for a certain state parameter time series x={x _i, i=1,2 ..., N, if the embedding dimension is m, time delay is τ, then phase space reconstruction is X={X _i, X _iBe the phase point in the m dimension phase space:

X _i=[x _i,x _i+τ,…,x _i+(m-1)τ] ^T,i=1,2,…,M (1)

Then embedding the seasonal effect in time series correlation integral is

$C(m,N,r,\mathrm{\&tau;})=\frac{2}{M(M-1)}\underset{1\&le;i<j<M}{\mathrm{\&Sigma;}}\mathrm{\&theta;}(r-d_{\mathrm{ij}}),r>0---\left(2\right)$

M=N-(m-1) τ wherein, d _Ij=|| x _i-x _j|| _∞, be the ∞ norm; θ is the Heaviside function, and its expression formula is

$\mathrm{\&theta;}\left(x\right)=\left\{\begin{array}{cc}0,& x<0\\ 1,& x\&GreaterEqual;0\end{array}\right.$

Correlation integral is cumulative distribution function, in the expression phase space between any two points distance less than the probability of r.Define in addition x={x _iTest statistics:

S(m,N,r,τ)=C(m,N,r,τ)-C ^m(m,N,r,τ) （3）

S (m, N, r, τ) has reflected the seasonal effect in time series autocorrelation performance, and optimal time postpones to get the 1st zero point of S (m, N, r, τ), and the point in the phase space reconstruction is near even distribution at this moment, and reconstruct attractor track launches fully in phase space;

2. utilize maximum Lyapunov exponent identification DFIG chaotic characteristic: the time delay τ that is 1. calculated by step and embed dimension m, use the small data quantity method and calculate gearbox of wind turbine inboard bearing temperature, generator windings maximum temperature, rotor mean speed and generator active power one of four states parameter seasonal effect in time series maximum Lyapunov exponent, if four maximum Lyapunov exponent all greater than 0, then there is the DFIG chaotic characteristic in explanation, and carries out next step prediction;

3. utilize the weighing first order local area method that DFIG is predicted:

Delay time T and dimension m that step is tried to achieve in 1. carry out phase space reconfiguration, use the weighing first order local area method gearbox of wind turbine inboard bearing temperature, generator windings maximum temperature, rotor mean speed and the generator active power one of four states parameter time series of wind power generating set are predicted respectively;

If central point (i.e. the starting point of prediction) X _kNeighbor point be X _Ki, 2 distances are d _i, establish d _mD _iIn minimum value, the some X _KiWeights be:

$p_i=\frac{e^{-l(d_i-d_m)}}{\underset{i=1}{\overset{q}{\mathrm{\&Sigma;}}}{e}^{-l(d_i-d_m)}}$

Generally get l=1, then weighting 1 rank local linear fit is X _Ki+1=ae+bX _Ki, e=[1,1 ..., 1] ^TFind the solution according to weighted least-squares method

, obtain prediction type X _K+1=a+bX _k, construct next central point and neighbor point thereof, repeat 3. with further prediction, the abnormal conditions prediction occurs and stop until predicting unit.

Principle of work of the present invention:

The double fed induction generators group system is the multidimensional nonlinear system of a complexity, the variation audient multifactor impact of its running status, only rely on a certain factor and predict that operating states of the units has significant limitation, yet in production reality, hold facile operational factor time series and can only reflect a part of information, therefore consider to utilize the method for phase space reconstruction to predict, this method is constructed one group of coordinate components that characterizes the original system dynamics by single system's output time series, thus the chaotic attractor of recovery system approx.On this basis, utilize the phase space (real system approximate) of reconstruct to predict (mainly be utilize the track development of this system be that chaotic attractor is predicted).

According to the Takens theorem, for a time series, when m 〉=2d+1 (m embeds dimension, and d is the correlation dimension of power system), can recover attractor at this m dimension reconstruction attractor.Phase path in the reconstruction attractor and motive power system keep differomorphism.

Reconstructing method is as follows:

The C-C method is determined to embed dimension and time delay: during running of wind generating set, for a certain state parameter time series x={x _i, i=1,2 ..., N, if the embedding dimension is m, time delay is τ, then phase space reconstruction is X={X _i, X _iBe the phase point in the m dimension phase space:

X _i=[x _i,x _i+τ,…,x _i+(m-1)τ] ^T,i=1,2,…,M (1)

Then embedding the seasonal effect in time series correlation integral is

$C(m,N,r,\mathrm{\&tau;})=\frac{2}{M(M-1)}\underset{1\&le;i<j<M}{\mathrm{\&Sigma;}}\mathrm{\&theta;}(r-d_{\mathrm{ij}}),r>0---\left(2\right)$

M=N-(m-1) τ wherein, d _Ij=|| x _i-x _j|| _∞Be the ∞ norm; θ is the Heaviside function, and its expression formula is

$p_i=\frac{e^{-l(d_i-d_m)}}{\underset{i=1}{\overset{q}{\mathrm{\&Sigma;}}}{e}^{-l(d_i-d_m)}}$

Correlation integral is cumulative distribution function, in the expression phase space between any two points distance less than the probability of r.Define in addition x={x _iTest statistics:

S(m,N,r,τ)=C(m,N,r,τ)-C ^m(m,N,r,τ) （3）

S (m, N, r, τ) has reflected the seasonal effect in time series autocorrelation performance, and optimal time postpones to get the 1st zero point of S (m, N, r, τ), and the point in the phase space reconstruction is near even distribution at this moment, and reconstruct attractor track launches fully in phase space.

Maximum Lyapunov exponent identification DFIG chaotic characteristic: the present invention adopts has higher reliability and the less small data sets calculating wind-powered electricity generation unit state parameter seasonal effect in time series maximum Lyapunov exponent of calculated amount, behind phase space reconstruction, seek the nearest neighbor of each point on the given track, by

Estimate maximum Lyapunov exponent.

Chaotic prediction: the present invention uses weighing first order local area method forecast model the DFIG running status is predicted, it is a kind of improvement on original local method, owing to introducing weights so that it has the precision of prediction of better adaptive ability and Geng Gao.

If central point X _kThe neighbor point of (i.e. the starting point of prediction) is X _Ki, i=1,2 ..., q, 2 distances are d _i, establish d _mD _iIn minimum value, the some X _KiWeights be

$p_i=\frac{e^{-l(d_i-d_m)}}{\underset{i=1}{\overset{q}{\mathrm{\&Sigma;}}}{e}^{-l(d_i-d_m)}}$

Generally get l=1, then weighting 1 rank local linear fit is

X _ki+1=ae+bX _ki，e=[1,1,…,1] ^T （4）

Find the solution according to weighted least-squares method

, obtain prediction type X _K+1=a+bX _k, then construct next central point and neighbor point thereof, carry out next step prediction.

Single-chip microcomputer adopts the single-chip microcomputer MC9S12DP256 of Freescale, and its enhancing catches timer to have and compare more reliable tally function with other timers.The MC9S12DP256 single-chip microcomputer uses 16 MH z external crystal-controlled oscillation in the native system, and Bus Clock Rate is configured to 8 MHz.For improving measuring accuracy, the work clock of timer module does not carry out frequency division to bus clock to be processed, and namely the count frequency of timer also is 8MHz, adopts the mode that rising edge is caught, and allows input capture interruption and timer to overflow interruption.The signal of speed probe output become the digital signal of speed in frequency after through conditioning and be transferred to the input capture module of single-chip microcomputer MC9S12DP256, by catching effective hopping edge (such as rising edge) of measured signal, record the constantly value of timer of effective hopping edge arriving, calculate the value of current seizure moment timer and last time caught the poor of moment timer value, the counting number that this is timer in the measured signal one-period just can get thus the interval time on two effective edges and try to achieve according to this rotating speed.

Data storage circuitry adopts the DS1225 chip of Dallas company.A0-A12 is the address input end mouth, and DQ0-DQ7 is data input/outbound port,

Be the gating port,

Be exportable port, For can writing inbound port, NC is connectivity port not.

Usb circuit adopts the ooze CH372 chip of permanent electronics of Nanjing, and it is to select input pin CS#, interrupt output pin INT# and address input pin A0 to consist of by 8 BDB Bi-directional Data Bus D7～D0 that can be used as passive parallel interface, read gate input pin RD#, write gate input pin WR#, sheet; Described 8 BDB Bi-directional Data Bus D7～D0, read gate input pin RD#, write gate input pin WR#, sheet select input pin CS#, interrupt output pin INT# and address input pin A0 to be connected with single-chip microcomputer.

Superiority of the present invention is: 1. hardware unit is simple, practical; 2. data are protected automatically after the power down, and chaos detects uninterrupted in real time, and reliability is high; 3. high measurement accuracy; 4. the real time and reliability of system is high, can satisfy the requirement of the aspects such as wind-powered electricity generation unit status monitoring, transient process research, Fault diagnosis and forecast, has higher practical value.

(4) description of drawings:

Fig. 1 is the general structure schematic diagram of the related a kind of chaotic prediction system for the double fed induction generators group of the present invention.

Fig. 2 is the electrical block diagram of the related a kind of chaotic prediction system signal conditioning circuit unit for the double fed induction generators group of the present invention.

Fig. 3 is the electrical block diagram of the related a kind of A/D of the chaotic prediction system conversion interface circuit unit for the double fed induction generators group of the present invention.

Fig. 4 is the structural representation of the data storage cell DS1225 chip of the related a kind of chaotic prediction system for the double fed induction generators group of the present invention.

Fig. 5 is the structural representation of the interface circuit CH372 chip of the related a kind of chaotic prediction system for the double fed induction generators group of the present invention.

Fig. 6 is the general flow chart of the method for work of the related a kind of chaotic prediction system for the double fed induction generators group of the present invention.

(5) embodiment:

Embodiment: a kind of DFIG running status chaotic prediction system (see figure 1) based on phase space reconfiguration is characterized in that it comprises wind-powered electricity generation unit, tester and with the host computer of chaotic prediction program; Wherein, described tester gathers the signal of wind-powered electricity generation unit, is two-way with host computer with the chaotic prediction program and is connected.

Described tester (see figure 1) is by signals collecting and modulate circuit unit, A/D conversion circuit unit, single-chip microcomputer, usb circuit unit, data storage circuitry unit and timing circuit cell formation; Wherein, described signals collecting and conditioning unit gather the signal of wind-powered electricity generation unit, and its output terminal connects the input end of A/D conversion circuit unit; Described single-chip microcomputer is with A/D conversion circuit unit, usb circuit unit, data storage circuitry unit and be connected circuit unit and be respectively two-way the connection; Described usb circuit unit is two-way with host computer with the chaotic prediction program and is connected.

The signal of described collection wind-powered electricity generation unit is to gather wind-driven generator group wheel box inboard bearing temperature signal, aerogenerator winding maximum temperature signal, wind power generator rotor mean speed signal and aerogenerator active power parameter signal.

Described signals collecting and modulate circuit unit (see figure 2) are made of sensor, data collecting card, resistance R s, filtering circuit, voltage follower, modulate circuit and stabilivolt; It is connected to conventional the connection; Wherein said sensor adopts the isolation template, input signal all is converted to the standard voltage signal of 5V; Described data collecting card is 12 multifunctional data acquisition cards of PCI-1711 that grind magnificent company, have 16 tunnel single-ended analog inputs, 8 data signal channels are with an automatic channel/gain scan circuit, automatically control multi-channel gating switch during sampling, it is connected to conventional the connection.

Described A/D conversion circuit unit (see figure 3) is made of conversion chip and peripheral circuit; Wherein said conversion chip be adopt CMOS technique, be that ternary data output latch is arranged in the sheet, input mode is single channel, be 100 μ s switching time, supply voltage is+8 conversion chip ADC0804 of successive approximation of 5V; Described conversion chip ADC0804 comprises pin DB0, pin DB1, pin DB2, pin DB3, pin DB4, pin DB5, pin DB6, pin DB7, pin/WR, pin/RD, pin/CS, pin VIN (+), pin VIN (-), pin C LK-IN, pin CLK-R and pin Vref/2, and is wait time-delay mode with singlechip chip and is connected according to pin DB0, pin DB1, pin DB2, pin DB3, pin DB4, pin DB5, pin DB6, pin DB7, pin/WR, pin/RD, pin/CS; Described peripheral circuit is comprised of capacitor C 28, resistance R 32, two resistance R 33, capacitor C 29, power supply VCC; The signal of described pin VIN (+) after capacitor C 28 and resistance R 33 receiving signal reason processing of circuit, capacitor C 28 is connected common ground with these resistance R 33 tie points with pin VIN (-), takes differential voltage analog input mode; Described pin CLK-R is through another resistance R 33 and capacitor C 29 ground connection, and pin CLK-IN connects the tie point of this resistance R 33 and capacitor C 29; Described pin Vref/2 meets power supply VCC through resistance R 32.

Described single-chip microcomputer adopts the single-chip microcomputer MC9S12DP256 of Freescale.

Described data storage circuitry unit adopts the DS1225 chip of Dallas company.

Described usb circuit unit adopts the ooze CH372 chip of permanent electronics of Nanjing.

Described timing circuit unit adopts the PIC16F716 device with house dog.

A kind of method of work (see figure 6) of the chaotic prediction system for wind power system is characterized in that it may further comprise the steps:

⑴ arrange the acquisition interval timing by the timing circuit unit, comes Real-time Collection gearbox of wind turbine inboard bearing temperature, generator windings maximum temperature, rotor mean speed and generator active power signal by signals collecting and modulate circuit unit;

⑵ the signal that gathers among the step ⑴ is through signals collecting and modulate circuit unit and the A/D conversion circuit unit carries out filtering, self calibration is processed, and by single-chip microcomputer with set gear box inboard bearing temperature, generator windings maximum temperature, rotor mean speed and generator active power parameter are input in the data storage circuitry unit;

⑶ gearbox of wind turbine inboard bearing temperature, generator windings maximum temperature after process step ⑵ by usb circuit, rotor mean speed and generator active power data transmission are to being equipped with in the host computer of realizing the chaotic prediction program;

⑷ the Methods of Chaotic Forecasting that use in the host computer is carried out calculation of parameter and processing.

Methods of Chaotic Forecasting among the described step ⑷ is to adopt the method based on phase space reconfiguration to detect the chaos (see figure 6), is made of following steps:

1. determine to embed dimension and time delay with the C-C method: during running of wind generating set, for a certain state parameter time series x={x _i, i=1,2 ..., N, if the embedding dimension is m, time delay is τ, then phase space reconstruction is X={X _i, X _iBe the phase point in the m dimension phase space:

X _i=[x _i,x _i+τ,…,x _i+(m-1)τ] ^T,i=1,2,…,M (1)

Then embedding the seasonal effect in time series correlation integral is

$C(m,N,r,\mathrm{\&tau;})=\frac{2}{M(M-1)}\underset{1\&le;i<j<M}{\mathrm{\&Sigma;}}\mathrm{\&theta;}(r-d_{\mathrm{ij}}),r>0---\left(2\right)$

M=N-(m-1) τ wherein, d _Ij=|| x _i-x _j|| _∞, be the ∞ norm; θ is the Heaviside function, and its expression formula is

$\mathrm{\&theta;}\left(x\right)=\left\{\begin{array}{cc}0,& x<0\\ 1,& x\&GreaterEqual;0\end{array}\right.$

Correlation integral is cumulative distribution function, in the expression phase space between any two points distance less than the probability of r.Define in addition x={x _iTest statistics:

S(m,N,r,τ)=C(m,N,r,τ)-C ^m(m,N,r,τ) （3）

S (m, N, r, τ) has reflected the seasonal effect in time series autocorrelation performance, and optimal time postpones to get the 1st zero point of S (m, N, r, τ), and the point in the phase space reconstruction is near even distribution at this moment, and reconstruct attractor track launches fully in phase space;

2. utilize maximum Lyapunov exponent identification DFIG chaotic characteristic: the time delay τ that is 1. calculated by step and embed dimension m, use the small data quantity method and calculate gearbox of wind turbine inboard bearing temperature, generator windings maximum temperature, rotor mean speed and generator active power one of four states parameter seasonal effect in time series maximum Lyapunov exponent, if four maximum Lyapunov exponent all greater than 0, then there is the DFIG chaotic characteristic in explanation, and carries out next step prediction;

3. utilize the weighing first order local area method that DFIG is predicted:

Delay time T and dimension m that step is tried to achieve in 1. carry out phase space reconfiguration, use the weighing first order local area method gearbox of wind turbine inboard bearing temperature, generator windings maximum temperature, rotor mean speed and the generator active power one of four states parameter time series of wind power generating set are predicted respectively;

If central point (i.e. the starting point of prediction) X _kNeighbor point be X _Ki, 2 distances are d _i, establish d _mD _iIn minimum value, the some X _KiWeights be:

$p_i=\frac{e^{-l(d_i-d_m)}}{\underset{i=1}{\overset{q}{\mathrm{\&Sigma;}}}{e}^{-l(d_i-d_m)}}$

Generally get l=1, then weighting 1 rank local linear fit is X _Ki+1=ae+bX _Ki, e=[1,1 ..., 1] _TFind the solution according to weighted least-squares method

, obtain prediction type X _K+1=a+bX _k, construct next central point and neighbor point thereof, repeat 3. with further prediction, the abnormal conditions prediction occurs and stop until predicting unit.

Claims (10)

1. DFIG running status chaotic prediction system based on phase space reconfiguration is characterized in that it comprises wind-powered electricity generation unit, tester and with the host computer of chaotic prediction program; Wherein, described tester gathers the signal of wind-powered electricity generation unit, is two-way with host computer with the chaotic prediction program and is connected.

2. described a kind of DFIG running status chaotic prediction system based on phase space reconfiguration according to claim 1 is characterized in that described tester is by signals collecting and modulate circuit unit, A/D conversion circuit unit, single-chip microcomputer, usb circuit unit, data storage circuitry unit and timing circuit cell formation; Wherein, described signals collecting and conditioning unit gather the signal of wind-powered electricity generation unit, and its output terminal connects the input end of A/D conversion circuit unit; Described single-chip microcomputer is with A/D conversion circuit unit, usb circuit unit, data storage circuitry unit and be connected circuit unit and be respectively two-way the connection; Described usb circuit unit is two-way with host computer with the chaotic prediction program and is connected.

3. described a kind of DFIG running status chaotic prediction system based on phase space reconfiguration according to claim 1, the signal that it is characterized in that described collection wind-powered electricity generation unit is to gather wind-driven generator group wheel box inboard bearing temperature signal, aerogenerator winding maximum temperature signal, wind power generator rotor mean speed signal and aerogenerator active power parameter signal.

4. described a kind of DFIG running status chaotic prediction system based on phase space reconfiguration according to claim 2 is characterized in that described A/D conversion circuit unit is made of conversion chip and peripheral circuit; Wherein said conversion chip be adopt CMOS technique, be that ternary data output latch is arranged in the sheet, input mode is single channel, be 100 μ s switching time, supply voltage is+8 conversion chip ADC0804 of successive approximation of 5V; Described conversion chip ADC0804 comprises pin DB0, pin DB1, pin DB2, pin DB3, pin DB4, pin DB5, pin DB6, pin DB7, pin/WR, pin/RD, pin/CS, pin VIN (+), pin VIN (-), pin C LK-IN, pin CLK-R and pin Vref/2, and is wait time-delay mode with singlechip chip and is connected according to pin DB0, pin DB1, pin DB2, pin DB3, pin DB4, pin DB5, pin DB6, pin DB7, pin/WR, pin/RD, pin/CS; Described peripheral circuit is comprised of capacitor C 28, resistance R 32, two resistance R 33, capacitor C 29, power supply VCC; The signal of described pin VIN (+) after capacitor C 28 and resistance R 33 receiving signal reason processing of circuit, capacitor C 28 is connected common ground with these resistance R 33 tie points with pin VIN (-), takes differential voltage analog input mode; Described pin CLK-R is through another resistance R 33 and capacitor C 29 ground connection, and pin CLK-IN connects the tie point of this resistance R 33 and capacitor C 29; Described pin Vref/2 meets power supply VCC through resistance R 32.

5. described a kind of DFIG running status chaotic prediction system based on phase space reconfiguration according to claim 2 is characterized in that described single-chip microcomputer adopts the single-chip microcomputer MC9S12DP256 of Freescale.

6. described a kind of DFIG running status chaotic prediction system based on phase space reconfiguration according to claim 2 is characterized in that described data storage circuitry unit adopts the DS1225 chip of Dallas company.

7. described a kind of DFIG running status chaotic prediction system based on phase space reconfiguration according to claim 2 is characterized in that described usb circuit unit adopts the ooze CH372 chip of permanent electronics of Nanjing.

8. described a kind of DFIG running status chaotic prediction system based on phase space reconfiguration according to claim 2 is characterized in that described timing circuit unit adopts the PIC16F716 device with house dog.

9. method of work that is used for the chaotic prediction system of wind power system is characterized in that it may further comprise the steps institute:

⑴ arrange the acquisition interval timing by the timing circuit unit, comes Real-time Collection gearbox of wind turbine inboard bearing temperature, generator windings maximum temperature, rotor mean speed and generator active power signal by signals collecting and modulate circuit unit;

⑵ the signal that gathers among the step ⑴ is through signals collecting and modulate circuit unit and the A/D conversion circuit unit carries out filtering, self calibration is processed, and by single-chip microcomputer with set gear box inboard bearing temperature, generator windings maximum temperature, rotor mean speed and generator active power parameter are input in the data storage circuitry unit;

⑶ gearbox of wind turbine inboard bearing temperature, generator windings maximum temperature after process step ⑵ by usb circuit, rotor mean speed and generator active power data transmission are to being equipped with in the host computer of realizing the chaotic prediction program;

⑷ the Methods of Chaotic Forecasting that use in the host computer is carried out calculation of parameter and processing.

10. the method for work of described a kind of chaotic prediction system for wind power system according to claim 9 is characterized in that the Methods of Chaotic Forecasting among the described step ⑷ is to adopt the method based on phase space reconfiguration to detect chaos, is made of following steps:

1. determine to embed dimension and time delay with the C-C method: during running of wind generating set, for a certain state parameter time series x={x _i, i=1,2 ..., N, if the embedding dimension is m, time delay is τ, then phase space reconstruction is X={X _i, X _iBe the phase point in the m dimension phase space:

$X_i={[x_i,x_{i+\mathrm{\&tau;}},\&CenterDot;\&CenterDot;\&CenterDot;,x_{i+(m-1)\mathrm{\&tau;}}]}^T,i=\mathrm{1,2},\&CenterDot;\&CenterDot;\&CenterDot;,M---\left(1\right)$

Then embedding the seasonal effect in time series correlation integral is

$C(m,N,r,\mathrm{\&tau;})=\frac{2}{M(M-1)}\underset{1\&le;j<M}{\mathrm{\&Sigma;}}\mathrm{\&theta;}(r-d_{\mathrm{ij}}),r>0---\left(2\right)$

M=N-(m-1) τ wherein, d _Ij=|| x _i-x _j|| _∞, be the ∞ norm; θ is the Heaviside function, and its expression formula is

$\mathrm{\&theta;}\left(x\right)=\left\{\begin{array}{c}0,x<0\\ 1,x\&GreaterEqual;0\end{array}\right.$

Correlation integral is cumulative distribution function, in the expression phase space between any two points distance less than the probability of r.Define in addition x={x _iTest statistics:

$S(m,N,r,\mathrm{\&tau;})=C(m,N,r,\mathrm{\&tau;})-C^m(m,N,r,\mathrm{\&tau;})---\left(3\right)$

S (m, N, r, τ) has reflected the seasonal effect in time series autocorrelation performance, and optimal time postpones to get the 1st zero point of S (m, N, r, τ), and the point in the phase space reconstruction is near even distribution at this moment, and reconstruct attractor track launches fully in phase space;

2. utilize maximum Lyapunov exponent identification DFIG chaotic characteristic: the time delay τ that is 1. calculated by step and embed dimension m, use the small data quantity method and calculate gearbox of wind turbine inboard bearing temperature, generator windings maximum temperature, rotor mean speed and generator active power one of four states parameter seasonal effect in time series maximum Lyapunov exponent, if four maximum Lyapunov exponent all greater than 0, then there is the DFIG chaotic characteristic in explanation, and carries out next step prediction;

3. utilize the weighing first order local area method that DFIG is predicted:

Delay time T and dimension m that step is tried to achieve in 1. carry out phase space reconfiguration, use the weighing first order local area method gearbox of wind turbine inboard bearing temperature, generator windings maximum temperature, rotor mean speed and the generator active power one of four states parameter time series of wind power generating set are predicted respectively;

If central point (i.e. the starting point of prediction) X _kNeighbor point be X _Ki, 2 distances are d _i, establish d _mD _iIn minimum value, the some X _KiWeights be:

$p_i=\frac{e^{-l(d_i-d_m)}}{\underset{i=1}{\overset{q}{\mathrm{\&Sigma;}}}{e}^{-l(d_i-d_m)}}$

Generally get l=1, then weighting 1 rank local linear fit is X _Ki+1=ae+bX _Ki, e=[1,1 ..., 1] ^T

Find the solution according to weighted least-squares method , obtain prediction type X _K+1=a+bX _k, construct next central point and neighbor point thereof, repeat 3. with further prediction, the abnormal conditions prediction occurs and stop until predicting unit.

Images