This application is a 371 National Stage of International Application No. PCT/EP2015/001604, filed on Aug. 4, 2015, which claims the benefit of and priority to German Patent Application 10 2014 012 307.3, filed on Aug. 19, 2014. The entire disclosures of the above applications are incorporated by reference herein.
The disclosure relates to a method for the control of a reciprocating pump and an apparatus for using the method.
This section provides background information related to the present disclosure which is not necessarily prior art.
Electromagnetically driven reciprocating pumps are used to convey and meter fuels and reagents. They can be manufactured economically, and can, due to their pulsed mode of operation, be operated with adjustable conveyed quantities if the frequency of the pulse is changed.
Electromagnetically driven reciprocating pumps consist of an electromagnet and a fluid displacement unit into which the working fluid is sucked, from which it is ejected and subjected to pressure. The electromagnet and the displacement unit are in most cases inseparably connected together by common components, and if the structural form of a through-flow electromagnet is chosen, no rod seal is needed between the electromagnet and the displacement unit.
One disadvantage of the usual mode of operation of these displacement pumps is associated with current being fed through the magnetic coil until the movement of the magnetic armature has finished, or even for a longer period; this is necessary if a complete travel is to be achieved under all operating conditions without taking further measures.
The current feed described above results in a hard impact of the magnetic armature, with correspondingly high noise and low efficiency of the electromagnetic drive.
A hard impact also correspondingly results during a return of the magnetic armature to the rest position if the return springs move the magnetic armature back without braking when the electromagnet is switched off.
Various methods for overcoming the said disadvantages are known from the patent literature, but these are in some cases very expensive and in some cases unsatisfactory.
The known methods all exhibit at least the following disadvantages:
In the dissertation “Entwurf von magnetischen Mini- and Mikroaktoren mit stark nichtlinearem Magnetkreis” (Design of magnetic mini- and micro-actuators with highly non-linear magnetic circuits) by Dr. M. Kallenbach (TU Ilmenau), a method for estimating the position of a magnet on the basis of measured values of voltage and current and of calculated values for the linked magnetic flux is described; it also determines highly accurate values for the magnet position even for non-linear magnetic systems. An application of the method to the control of a reciprocating pump is not described.
This disclosure addresses the object of describing a controller of an electromagnetically driven reciprocating pump which, by switching the voltage applied to the electromagnet depending on the position of the magnetic armature, influences the velocity of the magnetic armature. The position of the magnetic armature is not measured here, but is determined from other measured or calculated state variables of the electromagnet. Knowledge of important properties of the electromagnet, in particular non-linear properties, is to be acquired prior to the intended operation, and stored in a suitable form in the controller.
The method according to the disclosure is based on a mathematical model of the driving electromagnet, wherein the behaviour of the electromagnet over time is described by the state variables of voltage, coil current, coil resistance, linked magnetic flux, magnetic armature velocity and magnetic armature displacement. These state variables are independent of one another when considered simultaneously, but do influence one another dynamically.
While the voltage and the coil current are measured by means of a measuring device, the coil resistance is calculated from the voltage and the coil current. The linked magnetic flux of the electromagnet cannot be calculated simultaneously from the other state variables; only the first time-derivative of the linked magnetic flux can be calculated simultaneously from the voltage, the coil current and the coil resistance. The linked magnetic flux refers to the integral over the penetration area of all the local magnetic flux densities at the conceptually cut-through magnetic circuit.
The linked magnetic flux is preferably calculated by numerical integration on the basis of an initial value and its first time-derivative, and this can be done in real time, i.e. during the magnet travel, by a sufficiently powerful processor.
As an alternative to numerical integration of the state variables, other mathematical methods for calculation of the temporal progression of these state variables on the basis of simplified linear models can be used; this requires less processing power, but is less precise, since linear models cannot adequately represent the important non-linearities of an electromagnet.
With knowledge of the linked magnetic flux, this numerical integration can also calculate the acceleration, the velocity and the travel of the magnetic armature. The travel of the magnetic armature can, however, be read more accurately and quickly from a previously determined, stored table, in which the travel of the magnetic armature is entered as a function of the coil current and of the linked magnetic flux. A table of this sort shows the strong yet non-linear dependency of the linked magnetic flux as a function of the coil current on the variable air gap, and therefore on the magnet travel.
It is true that the use of this table is an estimation method, and is therefore subject to inaccuracies, but the table does take the special non-linear properties of the electromagnet being used into account, as can be recorded for the general type of these electromagnets through measurements on a test bench, and therefore on the whole allows a significantly greater precision.
A further improvement in the estimation of the magnetic armature travel can be achieved if measurements are made on a test bench for different effective voltages and for both possible directions of the voltage changes at the magnetic coil, and if different tables are prepared from them and used. The non-linear effects of the saturation of the iron, the magnetic hysteresis and of the eddy currents are thus incorporated in the table, and thereby in the estimation method.
The precision of the calculation of the linked magnetic flux can, if necessary, be improved further if, in the calculation of the linked magnetic flux, the initial magnetization of the magnetic armature and of the iron return path from the previous history of the temporal progression of the linked magnetic flux is taken into account as an initial value for the numerical integration. The iron return path consists of the magnetic flux-carrying components of the magnetic pole, the housing and the yoke, and thus forms, together with the magnetic armature, an approximately closed circuit, broken only by the air gap between the magnetic armature and the magnetic pole.
With knowledge of the travel of the magnetic armature, the effective voltage at the magnetic coil can be changed by the controller, for example by switching on or off or by a suitable pulse-width modulation or pulse-length modulation, in such a way that the magnetic armature is braked in good time before striking the respective end stop, both during the working movement and during the return movement of the magnetic armature. The effective voltage refers to the mean DC voltage that would have the same effect as the voltage created by modulation.
The calculation and estimation method described can also be used, with small changes, for the return travel of the magnetic armature. Current flows through the magnetic coil even during the return travel, since the coil inductance only allows the current to decay slowly. The current can be measured, and a conclusion drawn as to the linked magnetic flux.
If the magnetic armature travel estimated by the method described reaches such a value that it would be appropriate to brake the return travel movement, the electrical controller increases the effective voltage to a value that generates suitable braking.
A table for the travel, the coil current and the linked magnetic flux, determined for correspondingly small voltages and with negative voltage changes, is advantageously used here in the estimation of the magnetic armature travel. This allows the non-linear behaviour of the magnetic materials to be appropriately taken into account.
In summary, the disclosure is characterized in that, as far as at all possible, existing knowledge about the electromagnet is used in order to perform the most accurate possible estimation of the magnetic armature travel on the basis of the temporal progressions of the coil current and voltage.
Reciprocating pumps of the type described and their electrical controllers are used to convey and/or meter fuels and reagents in vehicles and in mobile working machines.
The apparatus according to
The electromagnet is built from a magnetic coil (5), an iron return path (6) and a magnetic armature (7).
An electrical power supply (9) makes electrical power available to the apparatus, wherein the voltage can vary over a specified range, for example between 9 V and 16 V.
In an electrical controller (10), the electrical voltage is switched by means of a switching device (12), and the effective voltage and the resulting current are measured in a measuring device (13).
The magnetic coil is supplied with pulsed electrical power by the electrical controller (10), said electrical controller (10) also containing a processor (11) with programmable memory.
The processor (11) calculates
The position of the magnetic armature (7) is determined by means of the calculated value of the linked magnetic flux and the measured electrical current through the magnetic coil (5).
The electrical voltage at the magnetic coil (5) is switched by means of the switching device (12) depending on the position of the magnetic armature (7).
The present position of the magnetic armature is determined using an estimation process in the controller (11) from at least one table calculated prior to intended operation of the controller (10) and stored in the controller (11) with associated values of the electrical current, the linked magnetic flux and the position of the magnetic armature (7).
Advantageously the calculation of the linked magnetic flux is improved in that the calculation of the linked magnetic flux takes into account the initial magnetization of the magnetic armature (7) and of the iron return path (6) from the previous history of the temporal progression of the linked magnetic flux by means of the starting value.
A further improvement in the estimation of the position of the magnetic armature (7) is achieved in that, with different effective voltages and voltage changes at the magnetic coil (5), corresponding previously determined tables for different voltages and voltage changes, with respectively assigned values of the electrical current, the linked magnetic flux, and the position of the magnetic armature (7), are used. The effects of the non-linearity of the material properties, the magnetic hysteresis and the eddy currents are thus included in the estimation method.
The determination of the linked magnetic flux in the electromagnet (2) is advantageously carried out in the memory-programmable processor (11) through a calculation of the electrical and magnetic state variables of the electromagnet using a numerical integration running in real time.
Depending on the position of the magnetic armature (7) the electrical voltage at the magnetic coil (5) is if necessary switched off, or switched off and on a plurality of times, in the electrical controller (10) by means of the switching device (12), so that the effective voltage in the sense of a pulse-width modulation or of a pulse-length modulation has a time-average whose effect is reduced in comparison with the voltage of a voltage supply (9). When the magnetic armature (7) is moving forward against the force of the spring (4), the movement of the magnetic armature can in this way be braked to the extent that the magnetic armature only runs against its front stop at a very low residual velocity.
As the magnetic armature (7) is returned by the spring (4), the current through the magnetic coil only decays very slowly due to the inductance of the coil. Here again the current through the magnetic coil is measured by the measuring device (13) and is used in the calculation of the linked magnetic flux for determination of the position of the magnetic armature, wherein a previously calculated table for small voltages and negative voltage changes, also containing the coil current and the linked magnetic flux, is selected for the magnetic armature travel.
In this type of operation the information about the position of the magnetic armature is used in the electrical controller (10) in order to increase the effective mean voltage at the magnetic coil (5) depending on the position of the magnetic armature, and thus to brake the movement of the magnetic armature.
Number | Date | Country | Kind |
---|---|---|---|
10 2014 012 307 | Aug 2014 | DE | national |
Filing Document | Filing Date | Country | Kind |
---|---|---|---|
PCT/EP2015/001604 | 8/4/2015 | WO | 00 |
Publishing Document | Publishing Date | Country | Kind |
---|---|---|---|
WO2016/026551 | 2/25/2016 | WO | A |
Number | Name | Date | Kind |
---|---|---|---|
6208497 | Seale | Mar 2001 | B1 |
6945770 | Blaschke et al. | Sep 2005 | B1 |
20010043450 | Seale | Nov 2001 | A1 |
20150357107 | Fochtman | Dec 2015 | A1 |
Number | Date | Country |
---|---|---|
10020896 | Oct 2001 | DE |
10033923 | Jan 2002 | DE |
10127996 | Dec 2002 | DE |
102004002454 | Aug 2005 | DE |
19982757 | Jun 2009 | DE |
102011088699 | Jun 2013 | DE |
Entry |
---|
Kallenbach, Matthias,: “Design of magnetic mini-and micro-actuators with highly non-linear magnetic circuits”, Dissertation, Feb. 8, 1972, Leipzig, Germany (with English Abstract). |
International Search Report for PCT/EP2015/001604, ISA EP, Rijswijk, NL, dated Dec. 9, 2015. |
Written Opinion of the ISA for PCT/EP2015/001604, ISA EP, Rijswijk, NL, dated Dec. 9, 2015. |
Hayt, Engineering Electromagnetics, ISBN 0-07-027395-2, pp. 328-333, 352-353. © 1981. |
Number | Date | Country | |
---|---|---|---|
20170241413 A1 | Aug 2017 | US |