SE417473B - Inverter system with at least two AC inverter parallel operating inverter - Google Patents

Description

10 15 20 25 50 55 TI14852'6 2 and / or by the central control unit. An error in an orthogonal belt or in it central control unit can therefore affect several inverters or all inverters should lead to loss of the entire facility and thus to interruptions in the feeding of the load object. This can, for example, if the load object is made by a computer system, few. very serious consequences. ' The present invention aims to provide an inverter installation in which the inverters can operate in parallel operation without any connections whatsoever between the inverters is required to ensure an even load distribution between the inverters.

What characterizes an inverter system according to the invention is clear from attached patentJQ-av.

An inverter of the current type is often equipped with a low-pass filter and output transformer, which means that it has a purely inductive output impedance. At parallel driving of at least two. inverters, an active power exchange occurs between the inverters, mainly determined by the phase differences between the inverters of the inverters, According to the invention, each inverter is then given. such a frequency - power - characteristic that its frequency decreases slowly with increasing active current emitted or effect. An increasing output then leads to a decrease in the frequency, which gives a phase correction of the emitter of the inverter and thus a reduction of the emitted the effect. One on. inverters made this way will therefore automatically adjust its emitted active power and can work stably in parallel operation with other inverters, with the desired distribution of the active power, without the need for some superior controls.

According to the invention, the frequency control gas of each inverter is supplied with a frequency led value, which is composed of a load-independent part and one of the active current but (effect) dependent part. In parallel operation, it is then set. preferably the load independent part so that it is substantially the same for all inverters. Advantage- The effect of the active effect will then be determined by the slope of the different ones the frequency characteristics of the inverters and their rated effects.

In addition, according to a preferred embodiment, the voltage of each inverter is made depending on its emitted reactive current or power. in the case of inverters with inductors positive output impedance, the inverters are then given such a voltage - power - characteristic that their voltage decreases slowly with increasing emitted reactive current or effect. In the same way as for the active effect, this is obtained at parallel operation a definite distribution also of the reactive power between O 10 15 20 25 EO 3 e 7714852-6 Conveniently, the voltage conduction axis of each inverter is also supplied with a voltage conductor. value, which is composed of a load-independent part and a part that depends on the the reactive current (power) of the inverter. In parallel operation, it is then set. preferably it the load-independent part to essentially the same value for all the gear directions. Advantage- the reactive power is then determined by the inclination of the inverters. characteristics and of their brand effects. at an inverter system with several parallel-operating inverters according to the invention achieves a desired, eg uniform, distribution of both active and reactive load between inverters at the same time as each inverter works completely separate from the other inverters. An inverter fault at, - seems thus. not the other rectangular ones. other than that they jointly takes over the load of the faulty inverter. Nor is any central needed control unit which in the event of a fault in it can affect several or all of the inverters. Through to use inverters according to the invention, an inverter can therefore be obtained, which exhibits an extremely high level of operational reliability.

A further advantage is that an inverter system according to the invention is simple can be extended by simply connecting one or to the load object several additional inverters. The only. action needed is thus to the load object (ooh to the supply network) connect the new supply (s) come the inverters. In the case of previously known facilities, on the other hand, necessary to reconnect or rebuild the system used for to achieve the desired load distribution between the gear directions.

The invention will be described in more detail below in connection with the appended claims Figures 1-5. Fig. 1 shows how two inverters at a plant according to the invention The connection can be connected in parallel for feeding a common 'unloading object'. Fig. 2 shows in more detail an example of. an inverter at a. facility according to the invention and its control system. Fig. 3 shows how the frequency of the inverter varies with its emitted active power and Fig. 4 how the output voltage of the inverter varies with its released reactive effect. Fig. 5 shows an alternative design of the gear rectifier frequency control system.

Fig. 1 shows a load object A, for example a computer or a telecommunications system, which is connected to a three-phase AC busbar rst. The rail is supplied with from a three-phase supply network ESI 'via two. parallel-working inverter VR1 and m2. 10 15 20 25 50 7714852-6 Each inverter is supplied from a DC intermediate which includes an accumulator. mulator, IB1 and 32 respectively. Normally the power supply of the inverter takes place from the network RST via a controllable rectifier, S111 and S122, respectively, which is controlled so that the aclezmxzlator is being charged. In the event of a power failure, the inverter is discharged the accumulator. Before the inverter bridge itself VXR1 and VXRZ, respectively, a voltage arranged control means, which controls the supply DC voltage of the bridge and thus the amplitude of its output voltage. As shown in the figure, the control means may consist of one DC converters, LSO1 and 1502, respectively, the output voltage of which is smoothed by a densator, C11 and C21, respectively. The inverter bridge itself, VXR1 and VXRZ, a self-assembled three-phase inverter bridge with controllable frequency. It emits a three-phase alternating voltage which is smoothed in a filter consisting of a series inductor, L1 and L2, respectively, and a parallel capacitor, C12 and C22, respectively, and supplied the rail rst via a transformer, TR1 and TB2 respectively. Each inverter is pre- seen with coupling means (not shown) for automatic disconnection of the inverter in case of internal error. ' Fig. 2 shows the inverter bridge VXR1 with the accumulator 31 and the converter 1301.

To control the frequency of the inverter, a potentiometer P1 is provided, from which a frequency reference value fref is obtained, which corresponds to the desired one the inverter frequency. This is applied to a summing circuit S1. From a strömmëït- means' IM1 is obtained single signal ID, which corresponds to that supplied to the inverter direct current. Because the output voltage of the inverter is controlled in this way. that it is essentially constant, and the amplitude of the output voltage is proportional to the DC DC voltage, even the latter one will be essentially constant. The signal ID therefore becomes i mainly proportional to the active power input to the inverter and thus with sufficient accuracy for the purpose even against the inverter's output active power P (inverter losses are relatively small). One against In pro- portion value is subtracted in S1 from fref whereby the output flee from S1 will decrease with increasing ID, ie with increasing active power P. The frequency link value f is supplied with a voltage-controlled osoillator V00, the frequency of which is proportional led mot fleà. The oscillator provides pulses to a control pulse device SPD1 which amplifies and distributes the pulses to the valves of the inverter, whereby the frequency f of the inverter the output voltage of the rectifier becomes proportional to the joint value fled.

Fig. 3 shows how the frequency varies with ID (P), ie the frequency - power - of the inverter mmsrlafm. the sum Ar of r when 1D am from mn m1 the commodity I; 10 15 20 25 STAY 55 vvåuesz-6 may be, for example, 1% of fref or less. The internal impedance of the alternator is essentially purely inductive and is largely determined by the inductor L1 and the leakage inductor. the ductance in transformer-n TR1. Em. 'increase in the output of the inverter output power (of ID and P) means that the control system now described reduces the the frequency of the rectifier slightly. This causes the inverter's "emf" EAC (see Fig. 2) is retarded relative to the späzmingen on. skenan rst. This means that the inverter delivered active effect decreases. This is where the inverters that feed the rail come from rst automatically to control. set their phase positions so that, if the inverters are identical, the active effect of the load object is evenly distributed between the directional axes. If the exchange The ratios have different rated effects but the same value of Af (see Fig. 3) and the same relative tive output impedance, the power will be distributed between gear rilctarzza in proportion to their brand effects.

The measurement of the active power can alternatively be done on the inverter. power side or on the battery side of the IS drive.

From a potentiometer P2 a voltage reference value Uref is obtained, which is applied a summing circuit S2, the output signal Ulma of which is the LED value of the converter voltage. which controls the amplitude of the inverter voltage with Ulea. A voltage measuring device UH1 (eg a voltage divider) is arranged to emit a room voltage signal proportional to the amplitude of the AC voltage Um and this signal is applied to a measuring rectifier IR whose output signal Um in summing-s- lccetsen S5 is compared to Ulm? The difference u u Uled - UÅO is applied to a voltage regulator UR with PI characteristics. The begulator's output signal US is applied to the IS converter control lever SPZDZ, which in dependence on Us controls the output voltage of the inverter and. thereby amp- the velocity Um of the output voltage of the inverter by varying to - from the ratio that of the voltage pulses emitted by the inverter.

A measuring means m2 forms a signal IAC which is proportional to the inverter delivered alternating current. Im supplied with a phase sensitive lilcciktare PSR, where the resktiva component IQ of IAc is formed. The phase sensitive PSR is controlled by pulses p from the control pulse device SPD1. Because UAC is going to be close to constant is IQ with good accuracy proportional to the inverted reactive power of the inverter Q. In S2, a value proportional to IQ is subtracted from the Clock, whereby Tila and thus Um will decrease with increasing IQ, i.e. with increasing emitted reactive effect Q. Fig. 4 shows Um (actually Ulecl) as a function of IQ, ie the voltage power characteristic of the cleaner. When IQ increases from zero to its mark value decreases UAC by the amount AU1 ¿which for example can be 19% of Uni .. If the emitted reactive power Q of the inverter increases, the control system thus reduces the the voltage of the rectifier, which results in a lowering of Q. The inverters which supply the 10 15 20 25 30 55 ? 71¿§852 ~ 6 nan rst will hereby automatically adjust their. voltages so that a some distribution of the reactive effect is obtained. If the inverters are identical the reactive power consumption of the load object is distributed equally between the If they have different brand effects but the same. value of. A111 and Sami. relative output impedance is distributed over the reactive. the effect in proportion to the inverters' effects.

The LS converter can be of any type known per se, and is in Fig. 2 schematically shown with an extinguished thyristor T, which is periodically controlled between conductors and non-conductive state, as well as a diode D and an inductor L.

An alternative way to control the gearbox operation. maintained if instead of UAC (voltage after inauræenn m) mm (epännsngen before snauktem) is compared but Oulu. The voltage regulator will then strive to control E AC to equality with Oulu. For UAC to get. the desired dependence on the reactive effect is given Uled (een aännea EAC) en eåann knnelefenietik ett hänsyn tee m1 epänningefellet i inaustem m. on dette epänningefell (4113) vin nänkefnesnnen: å än 1; ex 10% ev Uref gen Uled een aan-nen Em en men IQ akenae materiens: (se fig 4, kurvan EAC) but eåann luming an A H2 than 9%, which gene of e 1%.

Normally the rail is supplied only by the switches VBJ and VRZ in Fig. 1 even when voltage is on. network EST. In order to achieve a further increase in operational reliability, a connection (streaked in Fig. 1) and an electrical switch SW1 are arranged directly between the mains EST and the rail rst. If all the inverters fall off, the clutch SW, which is normally open, is closed, whereby the rail rst receives feed directly from network EST. To enable such a connection without transients in the voltage on. ekenan rst it is necessary that this voltage whole. time is in phase with the voltage in the network EST. Fig. 5 shows a circuit which accomplishes this. and which constitute a supplement to the frequency control circuit shown in Fig. 2. A phase match don PC, such as a phase-sensitive rectifier, is connected to the mains EST and the rail rst and emits a signal A? which is proportional to the phase difference between the network and the stresses of the rail. AWare positive if the mains voltage is before the voltage at skenan rst. Alf ”is supplied via a humming circuit S4 with a phase control size FR with PI characteristic. llegulatorns utsigzsl fia (1 is supplied to a coupling organaxx SW2. This is controlled by a voltage measuring device UMZ, which senses the voltage in the network EST. At normal voltage in the mains, the S142 forwards the signal to the solar circuit S1. In the event of a failure of the voltage in the network RST, SW2 disconnects the connection from the phase latorn to S1. The rest of the control system, ie the units P1, S1, VCO and SPD1 fully corresponds in structure and function to the corresponding units in Fig. 2. 10 15 20 25 30 35 771Å852 ~ 6 In standard operation, ie the network EST is voltage set, the fiber rectifiers feed the rail rst and the switch SW1 is open, the phase controller FR will adjust its output signal so. that the inverter (if ID-O) is controlled in this way. that the voltage on. the rail rst to its fasiäge nnnnanfaller nes voltage on nam nsæ. signal ID (see ng 2) m1- the numbering unit S4 is moved with the opposite sign to the signal A 5 ”. If variable active power and thus ID increases, the phase regulator reduces the rectifier frequency, ooh thus delays the phase position of the inverter, which gives minsloiing of the active power of the inverter. Hereby the desired distribution of it is obtained active load between the inverters on. true. methods described in connection Fig. 2. The signals, fref and ID applied to the summing circuit S1, does not affect the fact that s is controlled against phase similarity (thanks to tower integrative effect). In this way is thus achieved. during normal operation partly that the voltage on the rail is first kept synchronized to the mains voltage, partly a fomig load distribution between the inverters. Would all. the inverters fall off therefore, the switch SW1 in Fig. 1 can be momentarily closed without transients occurring in the voltage applied to the belt object.

If, while the alternating direction at the above-described number of operating meters meters the rail first, the mains voltage drops off, the electrical switch SW2 disconnects the connection from regxzlatom FR to the summing circuit S1, and controlling the frequency of the gear unit comes then. to happen on. the same way as described above in connection with Fig. 2. a gear nozzle according to the invention may have any number inverters. "The inverters may be self-commutating inverters of any kind with controllable frequency and amplitude. Above has be- written how the amplitude is controlled by controlling themselves. inverter maimings- rectifier, but alternatively can inverters with built-in variation of the amplitude be used, eg so-called pulse inverters. Above has been described how each inverter has a separate simulator battery and a separate DC power supply (S111, S112). Alternatively, of course, the alternating rectum can be fed from a common power supply ooh and even sokxnnxzlator.

The ones described above. filter circuits. (eg L1, G12) for the output voltage of the inverter can of course be designed on. other than as described and can for example be so designed that the internal impedance of the inverter becomes substantially capacitive in snanet for snank fl v. styrngnrnn rön frequency / voltage non function of nktiv / / reactive effect then gives. modified with regard to the nature of the output impedance.

Also. can be the signals that give the desired dependence on the frequency converter ooh amplitude of the active and resective power or current is obtained at. other than as described above.

Claims (5)

7714352-6 PATENT REQUIREMENTS An inverter system with at least two AC-parallel inverters (VXIH, VXRZ), each inverter having a substantially reactive output impedance and provided with means for controlling the frequency (S1, vco) and amplitude of the wax output voltage (sz, un) ), characterized in that each inverter is provided with a separate frequency controlling means (S1, VCO), which is arranged to control the output frequency of the inverter (fled) so. that the frequency constitutes the sum. on the one hand a load-independent part (fšef), which is substantially equal in size to the various inverter axes, and on the other hand a part dependent on the active load (ID) of the inverter. 2. A transducer system according to claim 1, characterized in that each inverter is provided with means (IM2, PSR, S2, UR) for controlling the amplitude of the inverter output voltage (UAC) depending on its reactive current (IQ). An inverter arrangement according to claim 1, the output impedance of which is inductive, characterized in that the frequency controlling means of each inverter (S1, V00) is so. designed, that an increase in the active current (ID) of the inverter results in a decrease in its frequency. An inverter system according to claim 2, the output impedance of which is inductive, characterized in that the voltage controlling means of each inverter is designed so that an increase in the reactive current emitted by the inverter reduces its output voltage. 5. An inverter system according to claim 1, characterized in that each inverter is provided with a phase sensing means (PC) arranged to sense. the phase difference (A70) between the inverter output voltage (UAC) and the voltage in an AC voltage supply (Rss), the output signal (A4) from the phase-tuning means being arranged to be supplied with a phase regulator for actuating the inverter frequency controller (V00) said phase difference is controlled against a joint value dependent on the active current. STATED PUBLICATIONS: Sweden potentonsökon 10 047/69, 400 423 (HOZJ 3/18) Schweiz 496 360 (H02P 13/22), 566 671 (H02P 13/18), 593 585 (H02P 13/18) us 4 o19116 (a2 _45)