CN201450472U - Magnetic suspension control unit based on non-linear tracing differentiator - Google Patents

Abstract

The utility model discloses a magnetic suspension control unit based on a non-linear tracing differentiator. The magnetic suspension control unit based on the non-linear tracing differentiator comprises a master control unit, a suspension electromagnet unit and a sensor unit for collecting clearance state signals, wherein an output end of the master control unit is connected with the suspension electromagnet unit, the output end of the sensor unit is connected with an input end of a tracing differentiator, the output end of the tracing differentiator is connected with the master control unit, the tracing differentiator filters real time clearance state signals sent from the sensor unit and then obtains the speed signals, the speed signals are output to the master control unit, and the master control unit generates control signals to the suspension electromagnet unit according to the speed signals. The magnetic suspension control unit is a magnetic suspension control unit based on the non-linear tracing differentiator with simple and compact structure, low cost, high control precision and wide adaptable range.

Description

Magnetic suspension control unit based on the Nonlinear Tracking differentiator

Technical field

The utility model is mainly concerned with and is used for maglev control appliance field, refers in particular to a kind of magnetic suspension control unit based on the Nonlinear Tracking differentiator.

Background technology

Magnetic suspension controller is the most important components of systems such as magnetic suspension bearing, magnetically levitated flywheel, magnetic suspention balance, magnetic-levitation train.In current magnetic suspension controller, in order to realize the stable suspersion of said system, levitation gap and differential thereof---state informations such as vertical velocity that must real-time acquisition system.Rate signal can't directly obtain, and is normally obtained by the acceleration transducer integration in real system.Because transducer particularly acceleration transducer costs an arm and a leg, and causes the cost of whole system high, is unfavorable for the design of whole system, and most critical is, in systems such as magnetically levitated flywheel, acceleration transducer can't be installed at all, make this scheme use and be very limited.

The utility model content

The technical problems to be solved in the utility model just is: at the technical problem that prior art exists, the utility model provide a kind of simple and compact for structure, with low cost, control precision is high, the magnetic suspension control unit based on the Nonlinear Tracking differentiator of wide accommodation.

For solving the problems of the technologies described above, the utility model by the following technical solutions.

A kind of magnetic suspension control unit based on the Nonlinear Tracking differentiator, it comprises main control unit, levitating electromagnet unit and the sensor unit that is used for gathering the gap state signal, the output of described main control unit links to each other with the levitating electromagnet unit, it is characterized in that: the output of described sensor unit links to each other with an input of following the tracks of differentiator, the output of following the tracks of differentiator links to each other with main control unit, following the tracks of differentiator carries out obtaining rate signal after the filtering to the real-time gap state signal that is sent by sensor unit, and this rate signal exported to main control unit, main control unit generates control signal to the levitating electromagnet unit according to rate signal.

As further improvement of the utility model:

Described main control unit links to each other with the levitating electromagnet unit by chopper.

Described main control unit links to each other with a CAN bus network control unit, and CAN bus network control unit comprises CAN controller, CAN transceiver and CAN bus.

Compared with prior art, advantage of the present utility model just is: the high-quality differential signal of the system that extracts that can be real-time, and realize optimal filter to signal, magnetic suspension control system does not need to use accelerometer, be particularly suitable for to install in the middle of the control system of accelerometer, have simple compact, with low cost, the control precision advantages of higher of system configuration, have a wide range of applications.

Description of drawings

Fig. 1 is a general structure schematic diagram of the present utility model;

Fig. 2 is the principle schematic of following the tracks of data processing unit in the differentiator in the utility model;

Fig. 3 is the circuit theory schematic diagram of main control unit in the specific embodiment;

Fig. 4 is the circuit theory schematic diagram of A/D converting unit in the specific embodiment;

Fig. 5 is the circuit theory schematic diagram that CAN bus network control unit is connected with main control unit in the specific embodiment;

Fig. 6 follows the tracks of the levitation gap signal curve filter effect figure that differentiator obtains experiment;

The rate signal comparison diagram that Fig. 7 rate signal that to be gap signal that experiment is gathered obtain by tracking filter and real integrated acceleration obtain;

Fig. 8 is the rate signal that adopts difference method to obtain to the gap signal that experiment obtains;

Fig. 9 is the circuit theory schematic diagram of logical transition chip in the specific embodiment;

Figure 10 is the circuit theory schematic diagram of decoder in the specific embodiment.

Marginal data:

1, main control unit 2, levitating electromagnet unit

3, sensor unit 4, chopper

5, CAN bus network control unit 6, A/D converting unit

7, follow the tracks of differentiator

Embodiment

Below with reference to specific embodiment and Figure of description the utility model is described in further details.

As shown in Figure 1, the utility model is based on the magnetic suspension control unit of Nonlinear Tracking differentiator, it comprises main control unit 1, levitating electromagnet unit 2 and the sensor unit 3 that is used for gathering the gap state signal, the output of main control unit 1 links to each other with levitating electromagnet unit 2, the output of sensor unit 3 links to each other with an input of following the tracks of differentiator 7 by A/D converting unit 6, the output of following the tracks of differentiator 7 links to each other with main control unit 1, following the tracks of 7 pairs of real-time gap state signals that sent by sensor unit 3 of differentiator carries out obtaining rate signal behind the tracking filter, and this gap signal and rate signal exported to main control unit 1, main control unit 1 generates control signal to levitating electromagnet unit 2 according to gap and rate signal.Wherein, main control unit 1 links to each other with levitating electromagnet unit 2 by chopper 4, follow the tracks of differentiator 7 and adopt embedded design, with main control unit 1 be integral structure, main control unit 1 links to each other with a CAN bus network control unit 5, and CAN bus network control unit 5 comprises CAN controller, CAN transceiver and CAN bus.Levitating electromagnet 2 provides power to magnetic suspension system, is used for the gap state of real-time detecting system as the sensor unit 3 of gap sensor, and gap state obtains filtered gap signal and gap differential---rate signal after following the tracks of differentiator 7.Main control unit 1 is used to calculate the required controlled quentity controlled variable of system suspension, and this controlled quentity controlled variable is input in the chopper 4 with the form of PWM ripple, and then controls the size of electric current in the levitating electromagnet unit 2, to produce buoyancy upwards, realizes the stable suspersion of system.

Referring to Fig. 2, the core algorithm that comprises in the data processing unit in the tracking differentiator 7 of the present utility model is:

$\left\{\begin{array}{c}\hat{u}\left(k\right)=\mathrm{fast}(x_1\left(k\right)-v\left(k\right),x_2\left(k\right),r,c_0\·h)\\ x_1(k+1)=x_1\left(k\right)+h{x}_2\left(k\right)+\frac{1}{2}\hat{u}\left(k\right)h^2\\ x_2(k+1)=x_2\left(k\right)+\hat{u}\left(k\right)h,k=\mathrm{0,1,2},...\\ y\left(k\right)=x_1\left(k\right)+(c_0-1)h\·x_2\left(k\right)\end{array}\right.$

In following formula, c ₀Be filtering factor, r is a Turbo Factor.x ₁(k) will follow the tracks of input signal v (k) fast, and x ₂(k) with the differential of sufficient approximation input signal.If signal packet contains noise, then above-mentioned tracking differentiator 7 is with filtering noise, and the amplitude of actual signal can not change, and has realized x ₁(k) phase-lag compensation of signal can play good filter effect.

The method of application of aforementioned can obtain the filtering output of signal, use difference method to obtain the approximate differential signal, this signal packet contains noise, continue to use the noise in the same filtering method filtering differential signal, will obtain not having phase lag, remove the approximate differential signal of noise.

Wherein, fast (x ₁, x ₂, r, c ₀H) being expressed as of function:

Make s=sign (x ₁+ x ₂| x ₂|/2r) (1)

$t_A={\mathrm{<figure-callout\; id="2"\; label="sx"\; filenames="H2009200635894E00011.png,H2009200635894E00022.png"\; state="\{\{state\}\}">sx</figure-callout>}}_2+\sqrt{{\mathrm{<figure-callout\; id="1"\; label="sx"\; filenames="H2009200635894E00011.png,H2009200635894E00022.png"\; state="\{\{state\}\}">sx</figure-callout>}}_1+\frac{x_2^2}{2r}}---\left(2\right)$

Then:

$\tilde{u}\left(k\right)=\left\{\begin{array}{c}-\mathrm{rs},t_A\&GreaterEqual;h\\ -\mathrm{rs}(-\frac{1}{2}+\frac{x_{2}s}{\mathrm{rh}}+\frac{1}{2}\sqrt{1+\frac{4}{r}(\frac{x_2}{h}+\frac{{2x}_1}{h^2})s}),t_A<h\end{array}\right.---\left(3\right)$

Wherein, h is the sampling time step-length, t _AArrive the time of switching curve for the point on the phase plane.

When gap signal being input in the middle of the tracking differentiator 7 of the present utility model, just can well filter out the noise that wherein comprises, and what obtain is the gap signal of no amplitude and phase error, in Fig. 6, fluctuating stage gap variation diagram for Suspension Control, as can be seen, filtering method of the present utility model can play good filter effect to gap signal.

In the utility model, adopt the reliability of high-speed ADC chip as sensor unit 3 transmission, the CAN bus network links to each other with the CAN bus transceiver after photoelectricity is isolated.Adopt the control compute chip of the TMS320F240 of TI company as main control unit 1, the DSP main control chip is finished control and communication function by address wire, data wire and control line and CAN bus control unit.

Referring to Fig. 3, be input, the output pin circuit diagram of the utility model main control unit 1.Consider that the on-the-spot electromagnetic environment of Suspension Control is complicated abominable, in order to improve the antijamming capability of system, master control DSP unit adopts the operating voltage of TI company be+the TMS320F240 processor of 5V, and its operating frequency is 20Mhz, can satisfy the demand of control calculating.The main interface of DSP is address bus, data/address bus, memory control pin, the output of PWM ripple, jtag interface, power supply and clock drive signal etc.The external crystal oscillator of system is 8M, and through behind the inner PLL phase-locked loop circuit, the CPUClock of system is 20Mhz.The 37th pin is the MP/MC pin, and when it was logical one when setting, DSP was operated in microprocessor model, the external memory storage of using system, and during for logical zero, DSP is operated in the microcomputer pattern, the internal storage of using system.Address bus, data/address bus and peripheral read-write control logic are the interfaces of the peripheral components of system.

Referring to Fig. 4, be the pin and the logic level transition figure of A/D converting unit 6 in the present embodiment.A/D converting unit 6 adopts the Max196 chip of Maxim company.Max196 is 12 a multrirange data acquisition chip, possesses the input of 6 tunnel analog signals, can programme separately in each road.A certain road fault does not influence the work on other roads.The changing voltage scope of Max196 be can for 0～10V ,-5～+ 5V, 0～5V etc.When Max196 taked normal running, internal clocking pattern, the startup conversion instruction of 0～10V range of voltages was ADC=0x5x.Wherein ADC is the address of Max196, the x road analog input channel of the x among the 5x for selecting, and x belongs to 1～6.DSP of the present utility model can not directly link to each other with the ADC chip, must pass through logic level transition.As shown in Figure 9, the G16V8 chip is the logical transition chip, and wherein, ADC is the chip selection signal of Max196, and ADWE and ADRD are respectively the control signals of its write and read.In GAL16V8, following logic is arranged:

ADC＝A0&A1&A2&IS

ADRD＝RW

ADWE＝WE

Referring to Fig. 5, the circuit diagram that is connected with main control unit 1 for CAN bus network control unit 5 in the present embodiment.Itself does not have the CAN bus control unit this DSP, needs external relevant chip.Here select independent CAN bus control unit SJA1000 and 82C250 as the interface of bus transceiver as network.In order to strengthen the antijamming capability of CAN bus node, SJA1000 links to each other with 82C250 by high speed photo coupling 6N137 again, and power supply that optocoupler two lateral circuits adopt and ground isolate fully.The build-out resistor by one 62 Ω is reducing the end points reflection interference that signal of communication is transferred to lead separately for the CANH of 82C250 and CANL pin, and the little electric capacity of 2 100pF in parallel can filter out the High-frequency Interference on the bus between CANH and CANL and the ground.

The interface signal that SJA1000 provides mainly contains data and address time-sharing multiplex line AD0～AD7 and ALE address latch signal, CS, RD, signals such as WE, INT.The inner Harvard structure that adopts of DSP, address wire is separated parallel with data wire, and the ALE address valid signal in order to guarantee the normal access sequential beat of SJA1000, does not need to add the combination decoding circuit.

As shown in figure 10, adopt GAL16V8 as decoder, its inside programming is as follows:

CANCS＝IS+A2

CANRD＝RW&A0&A2&IS

CANWR＝WE&A0&A2&IS

ALE＝A0&A2&IS

So the CAN address is 101b (5) and 100b (4), the address during gating can data signal line is 101b, and the address during gating CAN address signal line is 100b.Operation to SJA1000 will produce ale signal, so the read/write operation of each SJA1000 all comprised for two steps: at first write the SJA1000 address that will operate, and then read or write data, can finish the operation of once reading or writing SJA1000.

Fig. 6 adopts algorithm described in the utility model, and the experiment gap data that is obtained when Suspension Control is tested can be seen, follows the tracks of 7 pairs of gap signals of differentiator and has realized good filter effect.

The contrast of the rate signal that Fig. 7 is obtained for the gap differential that obtained by algorithm described in the utility model and the rate signal of integrated acceleration gained, as can be seen, the match preferably of gap differential the integrated acceleration signal.Error occurring to the second half section is because drift has appearred in the integration of acceleration.

Fig. 8 is the common approximate differential signal that calculus of finite differences obtained, and this signal has been submerged in the middle of the noise fully, can't be applied to Suspension Control.By contrast as can be seen, the utility model can effectively extract rate signal.

Through experimental verification, this method can replace accelerometer, be applied in the middle of the Suspension Control, and respond well.

The above only is a preferred implementation of the present utility model, and protection range of the present utility model also not only is confined to the foregoing description, and all technical schemes that belongs under the utility model thinking all belong to protection range of the present utility model.Should be pointed out that for those skilled in the art in the some improvements and modifications that do not break away under the utility model principle prerequisite, these improvements and modifications also should be considered as protection range of the present utility model.

Claims (3)

1. magnetic suspension control unit based on the Nonlinear Tracking differentiator, it comprises main control unit (1), levitating electromagnet unit (2) and the sensor unit (3) that is used for gathering the gap state signal, the output of described main control unit (1) links to each other with levitating electromagnet unit (2), it is characterized in that: the output of described sensor unit (3) links to each other with an input of following the tracks of differentiator (7), and the output of following the tracks of differentiator (7) links to each other with main control unit (1).

2. the magnetic suspension control unit based on the Nonlinear Tracking differentiator according to claim 1 is characterized in that: described main control unit (1) links to each other with levitating electromagnet unit (2) by chopper (4).

3. the magnetic suspension control unit based on the Nonlinear Tracking differentiator according to claim 1 and 2, it is characterized in that: described main control unit (1) links to each other with a CAN bus network control unit (5), and CAN bus network control unit (5) comprises CAN controller, CAN transceiver and CAN bus.

Images