CN111751721A - Inductive load starting protection method and device - Google Patents
Inductive load starting protection method and device Download PDFInfo
- Publication number
- CN111751721A CN111751721A CN202010510261.3A CN202010510261A CN111751721A CN 111751721 A CN111751721 A CN 111751721A CN 202010510261 A CN202010510261 A CN 202010510261A CN 111751721 A CN111751721 A CN 111751721A
- Authority
- CN
- China
- Prior art keywords
- current
- voltage
- inductive load
- circuit
- sampling
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Images
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01R—MEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES
- G01R31/00—Arrangements for testing electric properties; Arrangements for locating electric faults; Arrangements for electrical testing characterised by what is being tested not provided for elsewhere
- G01R31/34—Testing dynamo-electric machines
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01R—MEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES
- G01R19/00—Arrangements for measuring currents or voltages or for indicating presence or sign thereof
- G01R19/25—Arrangements for measuring currents or voltages or for indicating presence or sign thereof using digital measurement techniques
Abstract
A method and apparatus for protecting the start of inductive load, the said method obtains the real-time parameter of current, voltage when the inductive load starts through the sampling conditioning circuit of alternating current; extracting the fundamental wave of the inductive load starting current by using a fundamental wave analysis algorithm; obtaining a future predicted value of the current by using a grey prediction model, and calculating the relative offset of the predicted current; and comparing the predicted value with an actual index threshold value, if the index is met, determining that the engine is normally started, and if the index is not met, determining that the engine fails, and removing the inductive load. The device comprises a digital signal processor, a sampling module input circuit, a power supply module, a communication module and an output circuit. The invention obtains real-time parameters of current and voltage when the inductive load is started through the alternating current sampling conditioning circuit, obtains fundamental current of grid current by utilizing a fundamental decomposition method, and reduces the influence of harmonic current or sudden voltage on a protection threshold.
Description
Technical Field
The invention relates to an inductive load starting protection method and device, belonging to the technical field of motor starting protection.
Background
The motor, as the most common inductive load, plays an important role in our national life, and its application field is also becoming more and more extensive. According to statistics, the number of damaged motors in China reaches three million every year, huge loss is caused to national economy, and many potential safety hazards are caused due to abnormal operation of the motors. In order to reduce national economic loss and ensure the safety and stability of a power system, a motor comprehensive protector is additionally arranged on a plurality of motors to realize the removal of a failed motor. The current motor comprehensive protector is often set to be 5 to 8 times of rated current when the criterion of motor locked-rotor protection is designed, and the setting time is about 2S. The criterion of locked-rotor protection is partially overlapped with the starting current (about 4 to 7 times of rated current) of the motor, and in order to avoid the misoperation of the protector, a plurality of factories and families set the time delay of the protector for 8 to 16 seconds to avoid the starting current when the motor is started. However, setting a time delay of the protector to avoid the stalling protection by mistake at the time of starting the motor is equivalent to giving up the protection of the motor within 8 to 16 seconds of starting the motor, that is, if the motor has a fault such as stalling at the time of starting, the protector does not operate. The setting time of the locked rotor protection is about two seconds, which is far less than the starting delay of a conventional protector, and if the motor has a locked rotor fault during starting and cannot be cut off in time, considerable economic loss and potential safety hazard are likely to be caused.
Meanwhile, the power quality of the power system often causes that the current and voltage components of the inductive load system contain a large amount of harmonics or peaks due to the addition of a large amount of nonlinear loads and the conditions of harmonic content, voltage flicker and the like, so that the current detected by the protection monitoring device contains non-fundamental wave components, the protection malfunction is easily caused, the inductive loads, particularly the starting failure of the motor, are caused, and great loss is caused to enterprise production.
Disclosure of Invention
The invention aims to provide an inductive load starting protection method and device for improving the reliability of a motor comprehensive protector, ensuring safe and stable operation of a power grid and reducing possible economic loss.
The technical scheme of the invention is that the method for protecting the starting of the inductive load obtains real-time parameters of current and voltage when the inductive load is started through an alternating current sampling conditioning circuit; extracting the fundamental wave of the inductive load starting current by using a fundamental wave analysis algorithm; obtaining a future predicted value of the current by using a grey prediction model, and calculating the relative offset of the predicted current; and comparing the predicted value with an actual index threshold value, if the index is met, determining that the engine is normally started, and if the index is not met, determining that the engine fails, and removing the inductive load.
The fundamental wave extraction of the inductive load starting current adopts a fundamental wave analysis algorithm, and the algorithm is as follows:
three-phase power grid voltage u acquired by current sampling circuit and voltage sampling circuit for starting inductive loadabcWith three-phase load current iLabcObtaining the corresponding voltage u under the dq0 coordinate system through coordinate transformation from the three-phase static coordinate abc to the rotating coordinate system dqdqCurrent iLdq；udq、iLdqAfter passing through a low pass filter LPF becomes
T1、T2A current-voltage conversion transformation matrix is projected for a synchronous dq0 coordinate system, and theta is a power grid voltage phase acquired through a phase-locked loop; i.e. iLqdq、iLpdqIs passing through T2Converting the corresponding fundamental wave active and reactive currents obtained by the matrix; i.e. indqIs iLdqComponent i obtained after band-pass filteringhdqIs a harmonic component of the load current; p is an active instruction, Q is a reactive instruction, and an active compensation instruction current i is obtained after T1 matrix transformationgdq；iLqdqHarmonic component i of the same load currenthdqThe sum of which is obtained the control command current icdqControl command current icdqSame active power compensation command current igdqAdding to obtain a reference current i in dq coordinate systemref.dq(ii) a Reference current i in dq coordinate systemref.dqObtaining a reference current i under a three-phase coordinate system through the transformation from a rotating coordinate system dq to a three-phase static coordinate abcref.abc(ii) a The above process filters out the high-order component i of the starting currentref.abcThe instruction reference current of the inductive load starting current does not contain harmonic components and only contains fundamental wave components, and is more suitable for judging the starting current.
The current sampling circuit and the voltage sampling circuit started by the inductive load work as follows:
the voltage sampling circuit directly adopts a resistance voltage division differential amplification circuit to measure the voltage of an alternating current lateral line and the voltage of a direct current bus; the first stage is a differential amplification circuit, and an RC filter circuit is arranged at the input end of the amplifier to realize the suppression of high-frequency interference signals; the second stage is a voltage follower, so that the load capacity of the input voltage is enhanced, and the voltage is more stable; because the AD sampling of the DSP can only input 0-3V voltage, a pull-up circuit is arranged at last to change all input voltage into positive voltage.
When the current sampling circuit collects current, the current transformer is used for reducing a current signal of a primary side, strong current isolation is realized, and a resistor is used for converting the current signal into a voltage signal; the first resistor R1 is a sampling resistor, the current on the secondary side generates voltage drop through the R1, and the collected current is converted into voltage; the first operational amplifier U1A is a voltage follower to increase the immunity to interference; the second resistor R2, the third resistor R3, the fifth resistor R5 and the second operational amplifier U2A form an inverting and adding circuit; the function is to add the voltage of the VF point and the voltage of the front stage and then invert the voltage; VF is a bias voltage, and is selected to be 1.65V; the sixth resistor R6, the eighth resistor R8 and the third operational amplifier U3A form an inverting circuit, and the inverted signal of the preceding-stage adder is inverted again and then is converted back to a positive value; the diodes D1, D2 act as clamps; VD1 and VD2 are forward voltage drops of the first diode D1 and the second diode D2 respectively; the ninth resistor R9 and the first capacitor C1 form a simple first-order passive RC low-pass filter, so that signals below the cut-off frequency normally pass through, signals above the cut-off frequency are greatly attenuated, the sampling and conditioning of the alternating current of the inductive load are finished, and finally the obtained analog signal of the inductive load monitoring device is sampled.
The method comprises the following steps of obtaining a future predicted value of the current by using a gray prediction model, and calculating the relative offset of the predicted current, wherein the algorithm flow of the gray prediction model is as follows:
(1) the definition of the gray GM (1,1) model is given first:
let X(0)＝(x(0)(1),x(0)(2),…x(0)(n))，X(1)＝(x(1)(1),x(1)(2),…,x(1)(n)), the X1 sequence is the cumulative generation of the original sequence X0;
x(0)(k)+ax(1)(k) b is the original form of the GM (1,1) model; a is called the development coefficient, b is called the ash contribution amount;
(2) is provided with Z(1)＝(z(1)(2),z(1)(3),…,z(1)(n)),
Generating a Z sequence as an adjacent mean value of the X1 sequence;
(3)x(0)(k)+az(1)(k) b is the basic form of the GM (1,1) model;
(4)
whitening equations called the GM (1,1) model base form;
the solution to the whitening equation, called the temporal response function:
(5) and discretely normalizing the time response function to obtain a time response sequence as follows:
reduction value:
the step of calculating the relative offset of the predicted current is as follows:
(1) the motor protector is set to delay 0.5S; the maximum value of the starting current of the motor appears before 0.2 second and then decays exponentially;
(2) the motor obtains the running real-time current of the motor through an alternating current sampling circuit every 0.1S, and sampling is carried out for ten times in one second; wherein the reciprocal of the current value obtained by the last sampling is recorded as i10；
(3) The result obtained by the grey prediction model is inevitably ascending in an exponential form, the reciprocal of the sampled data is put into the grey prediction model for prediction until the 30 th data is predicted, the reciprocal of the obtained data is taken again and returned to be the current prediction data after 3 seconds and is marked as i30；
Calculating the relative offset of the current
If Δ% is greater than the set index ΔzdIf the motor current is really attenuated according to the exponential law, namely the motor current is normally started, the step 4 is carried out; otherwise, the motor current is considered to be in a stable state, namely a locked rotor fault occurs, and the motor needs to be cut off from the power grid;
(4) sampling at equal intervals of 0.1 second next time, discarding the first of 10 data obtained in the previous round of sampling, sequentially advancing the remaining 9 data in corresponding storage units, and recording the newly sampled data as new i10And (4) repeatedly running the program in the step (3) until the time limit set by the user in the timer in the microprocessor is reached, and switching the motor protector to the normal working mode.
The algorithm flow of the gray prediction model in the step (3) is as follows:
(3.1) take the original ten sample data sets as I1 ═ I11,i12...i19,i110Generating an accumulation sequence I2 ═ I21,i22...i29,i210}, wherein:
(3.2) generating a new sequence I3 ═ I by performing a close-proximity average on the accumulated sequence I231,i32...i39Is in a relationship of i3k＝0.5(ik+ik+1),k＝1、2...9；
(3.3) constructing a data matrix B, B in the form of
(3.4) order
(3.5) solving the ash parameters a and b by using a least square method, wherein the calculation formula is
Wherein, a is a development coefficient, and b is an ash action amount;
(3.6) solution according to model
Solving the future change, and respectively taking k as 38 and 39 to obtain two data, namely the predicted value of I2 after 2.9sPredicted values of I2 after 3S are recorded as K1 and K2;
(3.7) subtracting K1 from K2, reducing to obtain expected data, and taking the reciprocal of the expected data, namely the predicted value of I1 after 3S; and sending the data into a set storage unit, and finishing the whole algorithm flow.
An inductive load starting current protection device comprises a digital signal processor, a sampling module input circuit, a power supply module, a communication module and an output circuit; the sampling circuit is connected with the digital signal processor through the A/D module; the power circuit is connected with the digital signal processor through the hour hand module of the system; the human-computer interaction communication circuit is interconnected with the digital signal processor through the SCI module; the output end of the digital signal processor is connected with the output circuit.
The digital signal processor adopts a DSP28335 main chip and is provided with a three-phase current and three-phase voltage detection circuit; the motor protection cut-off PWM pulse signal, the circuit breaker on-off switching signal and the actual test signal of the inductive load starting current; and the man-machine screen is interacted with the upper computer interface through 485 communication.
The invention has the advantages that the real-time parameters of current and voltage when the inductive load is started are obtained through the alternating current sampling conditioning circuit, the fundamental wave current of the power grid current is obtained by utilizing a fundamental wave decomposition method, and the influence of the harmonic current or the sudden change voltage on the protection threshold value is reduced. According to the characteristic that the locked-rotor current of the inductive load is stable and the starting current is gradually reduced, a future predicted value of the current is obtained by utilizing a grey prediction model, the relative offset of the predicted current is calculated, then the predicted value is compared with an actual index threshold value, if the predicted value meets the index, normal starting is considered, if the predicted value does not meet the index, a fault is considered to occur, and the inductive load is removed. The method is beneficial to improving the reliability of the comprehensive protector of the motor, and plays an important role in reducing possible economic loss for safe and stable operation of a power grid.
Drawings
FIG. 1 is a circuit diagram of a protection device according to the present invention;
FIG. 2 is a block diagram of current and voltage sampling;
FIG. 3 is a voltage sampling circuit;
FIG. 4 is a current sampling circuit;
FIG. 5 is a block diagram of inductive load startup reference current generation for a fundamental analysis algorithm;
FIG. 6 is a block diagram of a current collection process;
FIG. 7 is a flow chart of the inductive load startup current prediction and determination based on the gray prediction model.
Detailed Description
As shown in fig. 1, the inductive load startup current protection device of the present embodiment includes a digital signal processor, a sampling module input circuit, a power module, a communication module, and an output circuit; the sampling circuit is connected with the digital signal processor through the A/D module; the power circuit is connected with the digital signal processor through the hour hand module of the system; the human-computer interaction communication circuit is interconnected with the digital signal processor through the SCI module; the output end of the digital signal processor is connected with an output circuit and sends out a PWM pulse signal for motor protection cutting, a switch signal for switching on and off of a circuit breaker and an actual test signal of inductive load starting current; the digital signal processor adopts a DSP28335 main chip.
Fig. 2 is a block diagram of a current and voltage sampling circuit for starting an inductive load. The method comprises the steps of firstly, carrying out strong current and weak current conversion and isolation by a voltage transformer and a current transformer to obtain a low-voltage signal which can be received by a digital signal processor, then filtering clutter interference by a secondary conditioning circuit, adjusting a corresponding measurement range to an operation voltage range of the digital signal processor, and finally entering an AD sampling port of the digital signal processor.
Fig. 3 is a voltage sampling circuit, which directly adopts a resistance voltage-dividing differential amplifying circuit to measure the ac side line voltage and the dc bus voltage when sampling the voltage in consideration of the economical problem. The first stage is a differential amplification circuit which comprises a fourth amplifier U11A, an eleventh resistor R11, a twelfth resistor R12 and a capacitor C, wherein an RC filter circuit is formed by the resistor R11 and the capacitor C at the input end of the amplifier, so that the suppression of high-frequency interference signals is realized; the second stage is a voltage follower which comprises a fifth amplifier U12A and a tenth resistor R10, so that the load capacity of the input voltage is enhanced, and the voltage is more stable; since only 0-3V voltage can be input for the AD sampling of the DSP, a pull-up circuit including a thirteenth resistor R13, a fourteenth resistor R14, an eleventh diode D11, and a twelfth diode is finally provided to convert all input voltages to positive voltages.
Fig. 4 is a current sampling circuit, which uses a current transformer to reduce a current signal on a primary side when collecting current, so as to achieve strong current isolation, and simultaneously needs a resistor to convert the current signal into a voltage signal. The first resistor R1 is a sampling resistor, and the current of the secondary side generates voltage drop through the R1 to convert the collected current into voltage. The first operational amplifier U1A is a voltage follower, and the purpose of this part is to utilize the higher equivalent resistance of the amplifier input, so that the anti-interference capability can be increased, and at the same time, the output impedance is smaller, so that the sampling time can be reduced gradually, and the accuracy of converting into data volume is improved. The second resistor R2, the third resistor R3, the fifth resistor R5 and the second operational amplifier U2A form an inverting and adding circuit. Its function is to add the voltage at VF point to the voltage at the previous stage and then invert it. VF is the bias voltage, since the digital signal processor only accepts single polarity voltage signals of 0 to 3.3V, the negative half cycle of the sampled ac signal is not allowed to be less than 0, and the original ac signal amplitude should be between-1.65V and + 1.65V. The addition circuit is used to raise the AC signal as a whole, so that the negative part of the signal can be avoided. VF is generally chosen to be 1.65V. The sixth resistor R6, the eighth resistor R8 and the third operational amplifier U3A form an inverter circuit, which can invert the inverted signal of the preceding adder back to positive. The first diode D1 and the second diode D2 play a clamping role. The front-stage voltage is clamped by the diodes and can be ensured to be between-VD 2 and (VMF + VD1), and VD1 and VD2 are forward voltage drops of the first diode D1 and the second diode D2 respectively. VMF may take 3.3V. The ninth resistor R9 and the first capacitor C1 form a simple first-order passive RC low-pass filter, so that signals below the cut-off frequency can normally pass through, and signals above the cut-off frequency can be greatly attenuated. And finally, sending the analog signal of the inductive load monitoring device to the AD port of the sampling after the sampling and the conditioning of the inductive load alternating current are finished.
In the inductive load start protection method of the embodiment, real-time parameters of current and voltage when an inductive load is started are obtained through an alternating current sampling conditioning circuit; extracting the fundamental wave of the inductive load starting current by using a fundamental wave analysis algorithm; obtaining a future predicted value of the current by using a grey prediction model, and calculating the relative offset of the predicted current; and comparing the predicted value with an actual index threshold value, if the index is met, determining that the engine is normally started, and if the index is not met, determining that the engine fails, and removing the inductive load.
1. The fundamental wave extraction algorithm of the starting current of the inductive load is as follows:
the electric energy quality of the power system often causes that the current and voltage components of an inductive load system contain a large amount of harmonic waves or peak waves due to the fact that a large amount of nonlinear loads are added, the harmonic content, the voltage flicker and the like of the nonlinear loads, so that the current detected by the protection monitoring device contains non-fundamental wave components, the protection misoperation is easily caused, the inductive loads are caused, particularly the starting failure of the motor is caused, and great loss is caused to enterprise production. Therefore, the invention adopts a fundamental wave extraction algorithm to extract the fundamental wave of the inductive load starting current, so that the numerical value of the inductive load starting current is not influenced by the bad electric energy quality such as harmonic wave, flicker and the like, and a block diagram of the inductive load starting current is shown in figure 5.
As shown in FIG. 5, uabcFor the collected three-phase grid voltage iLabcIs the three-phase load current collected; collected three-phase grid voltage uabcWith the collected three-phase load current iLabcBy transformation (T) of the three-phase stationary coordinate abc into a rotating coordinate system dqabc/dq) Obtain the corresponding voltage udqCurrent iLdqNamely, the voltage and the current under the dq0 coordinate system; the LPF is a low-pass filter and,
is udq、idqThe corresponding numerical value after passing through a low pass filter LPF; t1, T2A current-voltage conversion transformation matrix is projected for a synchronous dq0 coordinate system, and theta is a power grid voltage phase acquired through a phase-locked loop; i.e. iLqdqiLpdqTo be transformed intoObtaining corresponding fundamental wave active and reactive currents by the matrix; i.e. indqIs iLdqComponent i obtained after band-pass filteringhdqThe harmonic component of the load current is P, Q are active and reactive commands, and the active compensation command current i is obtained after T1 matrix transformationgdq，iLqdqHarmonic component i of the same load currenthdqThe sum of which is obtained the control command current icdqControl command current icdqAdding the same active power compensation command current to obtain a reference current i under a dq coordinate systemref.dq(ii) a Reference current i in dq coordinate systemref.dqBy transformation (T) of the rotating coordinate system dq into the three-phase stationary coordinates abcdq/abc) Then the reference current i under the three-phase coordinate system can be obtainedref.abc. Due to the above process, the higher order component i of the starting current is filteredref.abcThe instruction reference current of the inductive load starting current does not contain harmonic components and only contains fundamental wave components, and is more suitable for judging the starting current.
Fig. 6 is a block diagram of a current collection process, in which decomposed and extracted fundamental wave sampling currents are subjected to flat superposition every 0.1S, and rolling superposition averaging is performed after ten times of sampling to obtain calculated currents.
2. Inductive load starting current prediction and judgment based on grey prediction model
According to the characteristic that the locked-rotor current of the inductive load is stable and the starting current is gradually reduced, a future predicted value of the current is obtained by utilizing a grey prediction model, the relative offset of the predicted current is calculated, then the predicted value is compared with an actual index threshold value, if the predicted value meets the index, normal starting is considered, if the predicted value does not meet the index, a fault is considered to occur, and the inductive load is removed. Fig. 7 is a flowchart illustrating the inductive load starting current prediction and determination based on the gray prediction model according to this embodiment.
The inductive load starting current prediction and judgment algorithm flow based on the gray prediction model is as follows:
the definition of the gray GM (1,1) model is given first:
let X(0)＝(x(0)(1),x(0)(2),…x(0)(n))，X(1)＝(x(1)(1),x(1)(2),…,x(1)(n)), the X1 sequence is the accumulated generation of the original sequence X0.
x(0)(k)+ax(1)(k) B is the original form of the GM (1,1) model. a is called the coefficient of development and b is called the amount of ash contribution.
Is provided with Z(1)＝(z(1)(2),z(1)(3),…,z(1)(n)),
The Z sequence is generated for the close-spaced mean of the X1 sequence.
x(0)(k)+az(1)(k) B is the basic form of the GM (1,1) model.
and discretely normalizing the time response function to obtain a time response sequence as follows:
reduction number
The method comprises the following specific steps
Step 1: the motor protector is set to a 0.5S delay. The maximum value of the motor starting current occurs before about 0.2 seconds and then decays exponentially.
Step 2: the motor obtains the running real-time current of the motor through an alternating current sampling circuit every 0.1S, and the sampling is carried out for ten times in one second. Wherein the reciprocal of the current value obtained by the last sampling is recorded as i10。
And step 3: grey predictionThe result obtained by the model is inevitably ascending in an exponential form, the reciprocal of the sampled data is put into a gray prediction model for prediction until the 30 th data is predicted, the reciprocal of the obtained data is returned to be the current prediction data after 3 seconds and is marked as i30。
The normal starting current of the motor decays according to an exponential law, the decay speed of the normal starting current of the motor is related to the characteristic motor constant of the motor, and the current of the motor is stable when a locked-rotor fault occurs.
Calculating the relative offset of the current
If Δ% is greater than the set index ΔzdIf the motor current really attenuates according to the exponential law, namely the motor is normally started, the step 4 is carried out, otherwise, the motor current is in a stable state, namely a locked rotor fault occurs, and the motor needs to be cut off from the power grid. Considering that the current decays exponentially, i.e. the current tends more smoothly, deltazdMay also be multiplied by a factor which decays exponentially with time
Where the setting of τ may be set by the user based on actual motor constants, capacity, etc.
And 4, step 4: sampling at equal intervals of 0.1 second next time, discarding the first of 10 data obtained in the previous round of sampling, sequentially advancing the remaining 9 data in corresponding storage units, and recording the newly sampled data as new i10The procedure of step 3 is repeated until the time limit set by the user in the microprocessor internal timer (approximately 5 to 6 seconds) is reached and the motor protector switches to the normal operating mode.
In step 3, the algorithm flow of the gray prediction model is as follows:
step 3.1: the original ten sample data sets are recorded as I1 ═ I11,i12...i19,i110Generating an accumulation sequence I2 ═ I21,i22...i29,i210}, wherein:
step 3.2: the accumulated sequence I2 is subjected to close-proximity average to generate a new sequence I3 ═ I31,i32...i39Is in a relationship of i3k＝0.5(ik+ik+1),k＝1、2...9。
Step 3.3: constructing a data matrix B, B in the form of
Step 3.4: order to
Step 3.5: solving the ash parameters a (development coefficient) and b (ash action amount) by using a least square method, wherein the calculation formula is
Step 3.6: solution according to model
And solving future changes, and respectively taking K as 38 and 39 to obtain two data, namely a predicted value of I2 after 2.9S and a predicted value of I2 after 3S, which are recorded as K1 and K2.
Step 3.7: subtracting K1 from K2 to obtain the expected data, and taking the reciprocal of the expected data, namely the predicted value of I1 after 3S. And sending the data into a set storage unit, and finishing the whole algorithm flow.
So far, through steps 1,2,3 and 4, a gray model obtained according to current sampling values when the motor is started is established.
According to the model, the defect that the starting time delay is set because the locked-rotor current and the starting current cannot be distinguished when the motor is started in the traditional motor protector is overcome. The start delay of the new scheme is only 0.5S, which is far less than the setting time of locked-rotor protection 2S, so that the reliability of the motor protector is improved, and the method has important significance for reducing economic loss caused by motor faults.
Claims (7)
1. The inductive load starting protection method is characterized in that real-time parameters of current and voltage when an inductive load is started are obtained through an alternating current sampling conditioning circuit; extracting the fundamental wave of the inductive load starting current by using a fundamental wave analysis algorithm; obtaining a future predicted value of the current by using a grey prediction model, and calculating the relative offset of the predicted current; and comparing the predicted value with an actual index threshold value, if the index is met, determining that the engine is normally started, and if the index is not met, determining that the engine fails, and removing the inductive load.
2. The method of claim 1, wherein the fundamental wave extraction of the start-up current of the inductive load adopts a fundamental wave analysis algorithm, the algorithm is as follows:
three-phase power grid voltage u acquired by current sampling circuit and voltage sampling circuit for starting inductive loadabcWith three-phase load current iLabcObtaining the corresponding voltage u under the dq0 coordinate system through coordinate transformation from the three-phase static coordinate abc to the rotating coordinate system dqdqCurrent iLdq；udq、iLdqAfter passing through a low pass filter LPF becomes
T1、T2A current-voltage conversion transformation matrix is projected for a synchronous dq0 coordinate system, and theta is a power grid voltage phase acquired through a phase-locked loop; i.e. iLqdq、iLpdqIs passing through T2Converting the corresponding fundamental wave active and reactive currents obtained by the matrix; i.e. indqIs iLdqComponent i obtained after band-pass filteringhdqIs a harmonic component of the load current; p is an active instruction, Q is a reactive instruction, and the active compensation instruction power is obtained after T1 matrix transformationStream igdq；iLqdqHarmonic component i of the same load currenthdqThe sum of which is obtained the control command current icdqControl command current icdqSame active power compensation command current igdqAdding to obtain a reference current i in dq coordinate systemref.dq(ii) a Reference current i in dq coordinate systemref.dqObtaining a reference current i under a three-phase coordinate system through the transformation from a rotating coordinate system dq to a three-phase static coordinate abcref.abc(ii) a The above process filters out the high-order component i of the starting currentref.abcThe instruction reference current of the inductive load starting current does not contain harmonic components and only contains fundamental wave components, and is more suitable for judging the starting current.
3. The method according to claim 1, wherein the gray prediction model is used to obtain a predicted value of the future current and calculate the relative offset of the predicted current, and the algorithm flow of the gray prediction model is as follows:
(1) the definition of the gray GM (1,1) model is given first:
let X(0)＝(x(0)(1),x(0)(2),…x(0)(n))，X(1)＝(x(1)(1),x(1)(2),…,x(1)(n)), the X1 sequence is the cumulative generation of the original sequence X0;
x(0)(k)+ax(1)(k) b is the original form of the GM (1,1) model; a is called the development coefficient, b is called the ash contribution amount;
(2) is provided with Z(1)＝(z(1)(2),z(1)(3),…,z(1)(n)),
Generating a Z sequence as an adjacent mean value of the X1 sequence;
(3)x(0)(k)+az(1)(k) b is the basic form of the GM (1,1) model;
(4)
whitening equations called the GM (1,1) model base form;
the solution to the whitening equation is called the temporal response function:
(5) discrete normalization of the time response function, the obtained time response sequence is:
reduction value:
4. the method of claim 1, wherein the step of calculating the relative offset of the predicted current comprises:
(1) the motor protector is set to delay 0.5S; the maximum value of the starting current of the motor appears before 0.2 second and then decays exponentially;
(2) the motor obtains the running real-time current of the motor through an alternating current sampling circuit every 0.1S, and sampling is carried out for ten times in one second; wherein the reciprocal of the current value obtained by the last sampling is recorded as i10；
(3) The result obtained by the grey prediction model is inevitably ascending in an exponential form, the reciprocal of the sampled data is put into the grey prediction model for prediction until the 30 th data is predicted, the reciprocal of the obtained data is taken again and returned to be the current prediction data after 3 seconds and is marked as i30；
Calculating the relative offset of the current
If Δ% is greater than the set index ΔzdIf the motor current is really attenuated according to the exponential law, namely the motor current is normally started, the step 4 is carried out; otherwise, the reverse is carried outThe motor current is considered to be in a stable state, namely, a locked rotor fault occurs, and the motor needs to be cut off from the power grid;
(4) sampling at equal intervals of 0.1 second next time, discarding the first of 10 data obtained in the previous round of sampling, sequentially advancing the remaining 9 data in corresponding storage units, and recording the newly sampled data as new i10And (4) repeatedly running the program in the step (3) until the time limit set by the user in the timer in the microprocessor is reached, and switching the motor protector to the normal working mode.
5. The method of claim 4, wherein the gray prediction model in step (3) has the following algorithm flow:
(3.1) take the original ten sample data sets as I1 ═ I11,i12...i19,i110Generating an accumulation sequence I2 ═ I21,i22...i29,i210}, wherein:
(3.2) generating a new sequence I3 ═ I by performing a close-proximity average on the accumulated sequence I231,i32...i39Is in a relationship of i3k＝0.5(ik+ik+1),k＝1、2...9；
(3.3) constructing a data matrix B, B in the form of
(3.4) order
(3.5) solving the ash parameters a and b by using a least square method, wherein the calculation formula is
Wherein, a is a development coefficient, and b is an ash action amount; b isConstructing a data matrix in the step (3.3); y is the data matrix in the step (3.4); b isTIs a transposed matrix of the matrix B;
(3.6) solution according to model
Solving future changes, and respectively taking K as 38 and 39 to obtain two data, namely a predicted value of I2 after 2.9S and a predicted value of I2 after 3S, which are recorded as K1 and K2;
(3.7) subtracting K1 from K2, reducing to obtain expected data, and taking the reciprocal of the expected data, namely the predicted value of I1 after 3S; and sending the data into a set storage unit, and finishing the whole algorithm flow.
6. The method of claim 2, wherein the current sampling circuit and the voltage sampling circuit for inductive load startup operate as follows:
the voltage sampling circuit directly adopts a resistance voltage division differential amplification circuit to measure the voltage of an alternating current lateral line and the voltage of a direct current bus; the first stage is a differential amplification circuit, and an RC filter circuit is arranged at the input end of the amplifier to realize the suppression of high-frequency interference signals; the second stage is a voltage follower, so that the load capacity of the input voltage is enhanced, and the voltage is more stable; because AD sampling of the DSP can only input 0-3V voltage, a pull-up circuit is arranged at last to change all input voltage into positive voltage;
when the current sampling circuit collects current, the current transformer is used for reducing a current signal of a primary side, strong current isolation is realized, and a resistor is used for converting the current signal into a voltage signal; the first resistor R1 is a sampling resistor, the current on the secondary side generates voltage drop through the R1, and the collected current is converted into voltage; the first operational amplifier U1A is a voltage follower to increase the immunity to interference; the second resistor R2, the third resistor R3, the fifth resistor R5 and the second operational amplifier U2A form an inverting and adding circuit; the function is to add the voltage of the VF point and the voltage of the front stage and then invert the voltage; VF is a bias voltage, and is selected to be 1.65V; the sixth resistor R6, the eighth resistor R8 and the third operational amplifier U3A form an inverting circuit, and the inverted signal of the preceding-stage adder is inverted again and then is converted back to a positive value; the diodes D1, D2 act as clamps; VD1 and VD2 are forward voltage drops of the first diode D1 and the second diode D2 respectively; the ninth resistor R9 and the first capacitor C1 form a simple first-order passive RC low-pass filter, so that signals below the cut-off frequency normally pass through, signals above the cut-off frequency are greatly attenuated, the sampling and conditioning of the alternating current of the inductive load are finished, and finally the obtained analog signal of the inductive load monitoring device is sampled.
7. An inductive load starting current protection device for implementing the method according to claims 1-6, wherein said device comprises a digital signal processor, a sampling module input circuit, a power supply module, a communication module, and an output circuit; the sampling circuit is connected with the digital signal processor through the A/D module; the power circuit is connected with the digital signal processor through the hour hand module of the system; the human-computer interaction communication circuit is interconnected with the digital signal processor through the SCI module; the output end of the digital signal processor is connected with an output circuit and sends out a PWM pulse signal for motor protection cutting, a switch signal for switching on and off of a circuit breaker and an actual test signal of inductive load starting current; the digital signal processor adopts a DSP28335 main chip.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010510261.3A CN111751721A (en) | 2020-06-08 | 2020-06-08 | Inductive load starting protection method and device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010510261.3A CN111751721A (en) | 2020-06-08 | 2020-06-08 | Inductive load starting protection method and device |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN111751721A true CN111751721A (en) | 2020-10-09 |
Family
ID=72675947
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202010510261.3A Pending CN111751721A (en) | 2020-06-08 | 2020-06-08 | Inductive load starting protection method and device |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111751721A (en) |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112630524A (en) * | 2021-03-05 | 2021-04-09 | 江苏瑞电智芯信息科技有限公司 | Low-current signal acquisition processing circuit and acquisition processing method |
| CN112783031A (en) * | 2020-12-30 | 2021-05-11 | 中国科学院长春光学精密机械与物理研究所 | Soft instruction starting control and level signal isolation circuit system |
| CN113848456A (en) * | 2021-09-26 | 2021-12-28 | 湖南科比特电气技术有限公司 | Circuit detection method and related assembly |
| CN116859289A (en) * | 2023-09-04 | 2023-10-10 | 国网江苏省电力有限公司常州供电分公司 | Ground fault detection device for direct current system of transformer substation |
Citations (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1645708A (en) * | 2005-01-21 | 2005-07-27 | 浙江大学 | Active electric filter by mixed harmonic detection |
| CN103066602A (en) * | 2012-12-31 | 2013-04-24 | 湖南工业大学 | Microgrid mixed type mixed type power filter harmonic current prediction method based on back-direction optimum linear prediction theory |
| CN103199534A (en) * | 2013-03-18 | 2013-07-10 | 武汉大学 | Improved beat control method of active filter |
| CN103986133A (en) * | 2014-05-30 | 2014-08-13 | 上海交通大学 | Fixed time-lag over-current protection method based on reactive power break variable |
| CN104393575A (en) * | 2014-10-24 | 2015-03-04 | 上海山源电子科技股份有限公司 | Phase-sensitive protection method and system for coal mine low-voltage distribution protector |
| CN107769259A (en) * | 2017-12-01 | 2018-03-06 | 北京航空航天大学 | A kind of inverter current forecast Control Algorithm based on discrete averaging model |
| CN109802600A (en) * | 2019-03-20 | 2019-05-24 | 合肥为民电源有限公司 | A kind of method and control device of power frequency inverter startup motor |
| CN109888733A (en) * | 2019-03-02 | 2019-06-14 | 安徽云汉智能科技有限公司 | A kind of early warning type motor protection method and device |
| CN209559972U (en) * | 2018-11-12 | 2019-10-29 | 国网江西省电力有限公司电力科学研究院 | A kind of high-frequency harmonic measuring device based on fundamental frequency prediction |
| CN110610226A (en) * | 2018-06-14 | 2019-12-24 | 北京德知航创科技有限责任公司 | Generator fault prediction method and device |
-
2020
- 2020-06-08 CN CN202010510261.3A patent/CN111751721A/en active Pending
Patent Citations (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1645708A (en) * | 2005-01-21 | 2005-07-27 | 浙江大学 | Active electric filter by mixed harmonic detection |
| CN103066602A (en) * | 2012-12-31 | 2013-04-24 | 湖南工业大学 | Microgrid mixed type mixed type power filter harmonic current prediction method based on back-direction optimum linear prediction theory |
| CN103199534A (en) * | 2013-03-18 | 2013-07-10 | 武汉大学 | Improved beat control method of active filter |
| CN103986133A (en) * | 2014-05-30 | 2014-08-13 | 上海交通大学 | Fixed time-lag over-current protection method based on reactive power break variable |
| CN104393575A (en) * | 2014-10-24 | 2015-03-04 | 上海山源电子科技股份有限公司 | Phase-sensitive protection method and system for coal mine low-voltage distribution protector |
| CN107769259A (en) * | 2017-12-01 | 2018-03-06 | 北京航空航天大学 | A kind of inverter current forecast Control Algorithm based on discrete averaging model |
| CN110610226A (en) * | 2018-06-14 | 2019-12-24 | 北京德知航创科技有限责任公司 | Generator fault prediction method and device |
| CN209559972U (en) * | 2018-11-12 | 2019-10-29 | 国网江西省电力有限公司电力科学研究院 | A kind of high-frequency harmonic measuring device based on fundamental frequency prediction |
| CN109888733A (en) * | 2019-03-02 | 2019-06-14 | 安徽云汉智能科技有限公司 | A kind of early warning type motor protection method and device |
| CN109802600A (en) * | 2019-03-20 | 2019-05-24 | 合肥为民电源有限公司 | A kind of method and control device of power frequency inverter startup motor |
Non-Patent Citations (4)
| Title |
|---|
| 李晓娓等: "基于粒子群动态灰色模型的短期风电功率预测", 《电气传动自动化》 * |
| 王枫等: "基于灰色模型的锂电池充电技术研究", 《电源学报》 * |
| 胡海江等: "一种任意次谐波电流检测算法的仿真研究", 《低压电器》 * |
| 贾聪: "基于灰色系统模型的变配电监控预警", 《集成电路应用》 * |
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112783031A (en) * | 2020-12-30 | 2021-05-11 | 中国科学院长春光学精密机械与物理研究所 | Soft instruction starting control and level signal isolation circuit system |
| CN112783031B (en) * | 2020-12-30 | 2021-12-17 | 中国科学院长春光学精密机械与物理研究所 | Soft instruction starting control and level signal isolation circuit system |
| CN112630524A (en) * | 2021-03-05 | 2021-04-09 | 江苏瑞电智芯信息科技有限公司 | Low-current signal acquisition processing circuit and acquisition processing method |
| CN113848456A (en) * | 2021-09-26 | 2021-12-28 | 湖南科比特电气技术有限公司 | Circuit detection method and related assembly |
| CN116859289A (en) * | 2023-09-04 | 2023-10-10 | 国网江苏省电力有限公司常州供电分公司 | Ground fault detection device for direct current system of transformer substation |
| CN116859289B (en) * | 2023-09-04 | 2023-11-14 | 国网江苏省电力有限公司常州供电分公司 | Ground fault detection device for direct current system of transformer substation |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111751721A (en) | Inductive load starting protection method and device | |
| CN102843060B (en) | Two-level two-direction current transformer and control method thereof | |
| WO2015165286A1 (en) | Relay protection method and device against lc parallel circuit detuning faults | |
| CN108123486B (en) | Control method and device for low voltage ride through of wind power converter | |
| CN109459618A (en) | The online capacitance detection method of the standard of electric drive system for electric vehicles dc-link capacitance | |
| CN104459389B (en) | It is a kind of to input determination methods and system that electricity is shaken in power down | |
| CN108627731A (en) | A kind of rapid detection method of single-phase power-off | |
| CN103022976A (en) | Frequency converter output short circuit protection circuit and protection method | |
| CN110518557B (en) | Fault current limiter input control method based on short-circuit current comprehensive information | |
| CN105305498A (en) | Low voltage ride through control method for high-power photovoltaic grid-connected inverter | |
| CN105301426A (en) | Multi-level selective low-voltage short circuit protection experiment system | |
| CN110596455B (en) | Power frequency electrical parameter extraction method, system and computer readable storage medium | |
| CN107658852A (en) | A kind of guard method of AC excitation synchronous machinery excitation winding phase fault and system | |
| CN115144700A (en) | New energy grid-connected system fault protection method based on injection signal amplitude | |
| CN214011368U (en) | Double-fed converter direct current bus capacitance capacity detection device | |
| CN113933752B (en) | IGBT open-circuit fault detection method and device for cascaded H-bridge converter | |
| CN113437732B (en) | Pilot protection method and system for photovoltaic power generation grid-connected tie line | |
| CN112865117B (en) | Reactive emergency control method, device and system after new energy collection direct current blocking | |
| Xia et al. | Fault diagnosis of frequency control system based on FFT | |
| CN109245049B (en) | Over-current protection method and device for rotor winding of alternating-current excitation synchronous motor | |
| CN103066602A (en) | Microgrid mixed type mixed type power filter harmonic current prediction method based on back-direction optimum linear prediction theory | |
| Babaa et al. | Combined electrical faults detection and diagnosis using current signature analysis | |
| JP5452259B2 (en) | Power regeneration converter | |
| CN110939542A (en) | Mechanical brake input method and device for pumped storage unit | |
| CN114524048B (en) | Control method for preventing spontaneous combustion of electric bicycle |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| WD01 | Invention patent application deemed withdrawn after publication | ||
| WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20201009 |