US3626198A

US3626198A - Process and apparatus for optimizing the product of two physical magnitudes

Info

Publication number: US3626198A
Authority: US
Inventors: Andreas Boehringer
Original assignee: Dornier System GmbH
Current assignee: Dornier System GmbH
Priority date: 1968-10-19
Filing date: 1969-09-24
Publication date: 1971-12-07
Anticipated expiration: 1988-12-07
Also published as: CH512105A; JPS4836159B1; FR2021063A1; DE1804130C3; DE1804130B2; DE1804130A1; GB1232147A; CA928830A

Abstract

The product of the two interdependent physical magnitudes is regulated by comparing measured instantaneous values of each magnitude in alternating manner with a stored part value of a previously measured peak value of the same magnitude and of switching over the magnitude from a decreasing to an increasing value while reversing the other magnitude from an increasing to a decreasing value upon the instantaneous value algebraically equaling the stored value.

Description

United States Patent PROCESS AND APPARATUS FOR OPTIMIZING THE PRODUCT OF TWO PHYSICAL MAGNITUDES 9 Claims, 9 Drawing Figs.

U.S. Cl 307/52, 323/20 Int. Cl 1102j 1/10 Field of Search 323/20, 22

Primary ExaminerHerman .1. Hohauser Attorney-Kenyon & Kenyon Reilly Carr & Chapin ABSTRACT: The product of the two interdependent physical magnitudes is regulated by comparing measured instantaneous values of each magnitude in alternating manner with a stored part value of a previously measured peak value of the same magnitude and of switching over the magnitude from a decreasing to an increasing value while reversing the other magnitude from an increasing to a decreasing value upon the 153 instantaneous value algebraically equaling the stored value.

SOL/4 R CELL GE/vERATOA 3 MULT/PL/CANPS MA GNITUDE CONTROL MEANS l l l Tl l I +L ,0

IISTHQLE TILT/N6 I v e I La r a l l /2 w 4 l l I I 677115 +k I r! I l T I PATENTEI] DEE H971 SHEET 2 BF 4 FIG. 3

CELL GENERATOR 5 M01. T/PL ma N05 MA 45 Nl ru DE SpM1 II +L 6/5 TA 6L E T/L TING STAGE 45 00545 BOEHRINGER PATENIEUUEI: nan

SHEET 3 [IF 4 CONSUMER FIG. 4

1. oa/cxa L SWITCH/N6 0wcE INVENTOR. ANDREAS EOEHR/NGER r9770 A/HYS PATENTEnnEc 7197! 3.628198 sumuofa E FIG.6A U v H6368 F|G.6C

v In

FIG.'6D

INVENTOR.

ANDRE/76 BOEf/R/NGER PROCESS AND APPARATUS FOR OPTIMIZING THE PRODUCT OF TWO PHYSICAL MAGNITUDES This invention relates to a process for automatic optimizing of a product formed from two physical magnitudes. More particularly, this invention relates to a process and a switching apparatus for optimizing the product of two physical multiplicands.

In practice, in many cases it is necessary that a system, an apparatus, or an installation shall regulate itself so that the product of two multiplicands shall have an optimum value. Systems that meet this requirement belong to the class of selfoptimizing systems. However, various problems have arisen in such systems. For example, it has been a problem to maintain the electrical output of a generator at its optimum value, as the product of the amperage and voltage, in the case where the characteristic line of the generator, i.e. the relationship between amperage and voltage, becomes altered. As a special example, there may also be mentioned the energy direct-converters, with which for various reasons the amperage-voltage characteristic line may change considerably during operations. In this connection, where systems rely on the radiation from the sun as a source of energy, such as, is made use of in satellites, and with the so-called solar-cells, such sources of energy display a great change in performance, depending on the location and on the attitude of the satellite, which subject the amperage-voltage characteristic line to considerable changes.

Even in the case of other devices, it is often of great importance to regulate the' product of two physical magnitudes which are dependent on one another continuously and automatically to its optimum value. For example, in the case where through external or internal influences, such as pressure and temperature conditions, or chemical changes of the substances used, or as a result of damage to or failure of individual elements in a device composed of a great number of individual elements or structural parts, great fluctuations of the characteristic can be produced which affect the behavior of the system. If no additional influence or control is exerted on the system during such variations of the system's characteristic (in the behavior of the system), obtainment of the possible optimum of the product cannot be guaranteed under the prevailing conditions. As a more specific example, in the case of a generator made of solar-cells, operating directly on a buffer-battery of constant voltage, the constant voltage of the battery must be chosen so small that even under the most unfavorable conditions, that is, at the lowest cell voltages when energy positively must be fed from the generator to the battery, the generator voltage must be sufficiently greater than the battery voltage. However, with higher generator voltages (often with small short circuit currents) a great percentage of the actually available energy is not used. Under such circumstances, it is therefore advantageous to intervene in the system so that it becomes adapted to the new conditions and again transmits the optimum of energy possible.

The requirement made of a self-optimumizing system corresponds to the problem of variation computation where the extreme values of a function that depends on certain functions is sought. Such self-optimumizing processes are of themselves known. However, in order to be able to effect an automatic adaptation in the optimumizing of the product of two multiplicands, all the previously known processes require a measurement or a computation of the product itself. This often means (for example, in optimumizing electric output) that both multiplicands have to be separately determined and multiplied with one another. This multiplication however (regardless of how it is done) represents a considerable expenditure. These processes are complicated; and in addition a great amount of complicated apparatus is needed which leads to large expenditures as well as the need for added space and weight.

Further, it is not sufficient to analyze the condition of a selfregulating system at a certain point of time so as to determine in what direction the system must be moved to comply with the requirements for optimizing. lnstead, a trial operation is needed, which in general represents an essential characteristic of all self-optimumizing systems. It is only through this, that an analysis of the condition of thesystem is possible at all. This trial or seeking process is termed a trial movement and is produced through an intended alteration of the regulating parameter which produces the trial movement, and makes clear in what direction a further alteration must be made in order to arrive at the optimum. This further alteration is then termed the basic movement. The trial movement and the basic movement may occur simultaneously or alternately, or may even be combined.

Accordingly, it is an object of the invention to optimize the product of two magnitudes in a simple manner.

It is another object of the invention to optimize the product of two interdependent magnitudes in a relatively economical manner.

Briefly, the invention starts out from the above mentioned possibility of combining the trial movement and the basic movement, because this possibility in principle promises a lower expenditure for apparatus technology. In accordance with the invention, its purpose is achieved in that during continuous switching over operations the peak value of the one or the other multiplicand is measured each time by a suitable meter and a part k of this measured peak value is fed to a switching logic which controls the switching over process. The switching logic in its turn controls a regulating generator so that the generator steadily decreases, over one or more regulatory parameters, the instantaneous value of the one multiplicand whose peak value has just been measured and steadily increases the instantaneous value of the other multiplicand, whereby the next switching over process then becomes performed when the instantaneous value of the now decreasing multiplicand has reached the part value k of its previous peak value. This alteration in the value of the multiplicands represents the testing movement and basic movement.

In comparison with previously known processes, the process of the invention has the advantage that it is no longer necessary to measure the product itself, and that the seeking or selecting process regulates itself. The part value k, which represents the diminution of the individual multiplicands, is chosen in the range of from 0.8 to 0.9 of the measured peak values and is advantageously chosen in the approximate order of magnitude of k==0.9. In this way, the amplitude of the trial movement remains small.

As mentioned above, it is important for the regulating generator to be able, by an intervention into the system, to steadily alter the individual multiplicands. In accordance with a further development of the invention, this is made possible, for example, in that the regulatory generator is made as an integrator, to whose input there is alternately applied, from the switching logic a positive or a negative constant value, through whose action on the system alters the working point of the system.

The peak value meter needed to measure the multiplicands can be made in an extraordinarily simple way. in each case, an adjustable percent k is tapped off from and is fed to the switching logic. In accordance with a further development of with respect to how the component G controls the output of the component S,if the component S is a solar cell generator which produces a voltage and a current to produce a certain power output, the regulatory generator G changes the voltage and current produced by the solar cell generator S.

The essence of the process consists in that for the purpose of optimizing a product of two magnitudes, only two measuring organs and one switching logic are needed. The measuring organs are used to determine the peak values of the two multiplicands while the switching logic serves to influence the regulatory parameter or parameters in the system which is or are to be regulated. The switching logic produces two quasistable states. In one state, the characteristic line is run through in one direction; and in the other state, in the reverse direction.

At each switch over, the peak value of the one multiplicand which appears is measured and, as mentioned, an adjustable percentage k is tapped off. After one switch over by means of the logic, the characteristic line is run through in such a way that the multiplicand which had just before had its peak value travels along the characteristic line in the reverse direction. At this second switch over point, the peak value of the second multiplicand is automatically retained. In an entirely similar way, the same part value, the percentage k for example is tapped off from the peak value of the second multiplicand at this point and is used as a control value for the following switch over process. The two switch over points so characterized represent the two end points of the trial movement on the characteristic line. Z has been amended to indicate the signal from the regulatory generator G which is emitted so as to change the parameters such as the current or voltage produced by the generator S These and other objects and advantages of the invention will become more apparent from the following detailed description and appended claims taken in conjunction with the accompanying drawings in which:

FIG. 1 schematically illustrates a system which utilizes the process of the invention;

FIG. 2 graphically illustrates a characteristic line of two dependent multiplicands of a product having the end points of a trial movement plotted thereon;

FIG. 3 schematically illustrates a modified system similar to FIG. I which utilizes the process of the invention;

FIG. 4 schematically illustrates another system which utilizes the process of the invention;

FIG. 5 graphically illustrates a pair of characteristic lines obtained from a variance in the multiplicands of a product; and

FIGS. 6a to 6d graphically illustrate the relationships between the various multiplicands of a product and a characteristic of a regulatory generator and time.

Referring to FIG. 1, in order to regulate a system S, which has a product generated by two multiplicands, e.g. an electric generator such as a solar cell generator in which the two magnitudes determinitive thereto are interpreted as a voltage U and a current I which form a product (i.e. power) UXI so that the product is optimized, the instantaneous peak values of the two magnitudes U, I are alternately measured in peak value meters of a switching apparatus. For example, the voltage U is delivered via a line 1 to a peak value meter Sp M-l and measured therein as is known to obtain a peak value Ua while the current I is similarly delivered via a line 2 and measured in a peak value meter Sp M-Z to obtain a peak value Ia. These meters each are adapted, as is known, to obtain a part value kUa of the measured peak value so that the part value can be tapped from the respective meter and sent via lines 3, 4 to a switching logic such as a logical switching device SL for storage. In addition, the switching logic SL is connected to the lines I, 2 of the system S over suitable lines 5, 6 to receive the instantaneous values of the voltage V and current I for comparison with the stored k values. The switching logic SL is further constructed so that when the instantaneous value thereafter becomes equal to the stored k value, the switching logic SL emits a signal to a regulatory generator G such as a multiplicands magnitude control means via a line 7 (FIG. 3 shows one possible realization of a switching logic and regulatory generator G). The regulatory generator G receives the signal from the switching logic SL and in turn is constructed to emit a responsive signal 2 over one or more supply lines 8 to intervene in the system S for the purpose of altering the multiplicands, i.e. decreasing one multiplicand (U or I) while increasing the other multiplicand (I or U), in known manners using known equipment. It is noted that the multiplicands, i.e. parameters, which result in a particular product as well as the supply lines for the multiplicands are dependent on the construction of the system S. Hence, the manner in which the signal Z influences the supply line 8 depends on the system S.

Referring to FIG. 2, a characteristic line 9 for a currentvoltage product (i.e. the power) is shown with respect to coordinates U (voltage) and I (current and the optimum power for the system S occurs at point P on the characteristic line 9. With the system S momentarily at the point a and with the working point representing the actual power output just having been reversed, the operation of the process as applied to the system S will be described. First, the peak value Ua of the voltage is measured in the peak value meter Sp M-I and a part value kUa is tapped off and fed to the switching logic SL via line 3. Then, the regulatory generator G alters the system S by means of the trial movement described above via the signal Z through known equipment (not shown) so that the magnitude of the voltage U, i.e. the first multiplicand, is thereafter steadily decreased in accordance with the characteristic line while the magnitude of the current I, i.e. the second multiplicand, is steadily increased. The instantaneous voltage U then diminishes, while passing the point of optimum power position P on the characteristic line 9. When the voltage reaches the part value kUa, the power output of the system S is at the point B on the characteristic line 9 and the switching logic SL effects a switch over so that the working point is reversed. The voltage U is then made to increase along the characteristic line 9 while the current I is made to decrease via the known equipment. At the same time, the peak value IB of the current is measured in the peak value meter Sp M-2 and the part value MB is fed to the switching logic SL. When the instantaneous current I, i.e. the second multiplicand, reaches the part value kIB a second switch over occurs so that the working point again travels along the characteristic line toward point B. That is, the voltage U is made to decrease while the current increases.

So long as the working point is not at the optimum position P on the characteristic line, the two movements along the characteristic line are not of equal magnitude in the increasing or decreasing sense. Consequently, the trial movement is displaced, in a superposed fundamental movement, in a direction toward the optimum working point P. Should the working point be at the optimum position P, the amplitudes of the trial movement are constant. The working point then travels continuously between the two switch over points a, B forwardly and backwardly.

When the switch over point a, [3 are not situated too far apart, e.g. with k equal to approximately 0.9, the amplitude of the trial movement remains small and has practically no influence on the given product UXI of the system. Thus, the performance of the system is subjected to only small variations in the region of the trial movement.

If through some influence the characteristic line becomes altered so that the working point is no longer situated in the optimum region, then this alteration is immediately sensed by the trial movement. The trial movement then becomes relocated automatically through the maximum values found from the two multiplicands in a basic movement, until its two end points again lie on the characteristic line on different sides of the point where the product now has its optimum value. In this new location, the trialmovement is once more carried out continuously, that is, until a stationary state is established.

Referring to FIG. 3, wherein like reference characters indicate like parts as above, each of the peak value meters Sp M-l, Sp M-2 are constructed, for example, with simple diodecondenser combinations 10 to carry out the functions of the meters. The switching logic SL is constructed with a pair of bistable tilting stages 11, as is known, each of which has an input connected to one of the peak value meters Sp M-l, Sp M-2. The meters Sp M-l, Sp M-2 operate so as to deliver the part values +k of the measured multiplicands to the bistable tilting stages 11. In addition, the instantaneous value -I of the measured magnitudes are also fed directly to the respective bistable tilting stages 11. Each bistable tilting stage 11 serves to compare the instantaneous value of a multiplicand with the respective stored part value k so that a switching operation can be initiated when the compared values are equal in magnitude and opposite in sign.

The switching logic SL is further constructed with an electronic switching device 12 as is known which is connected to the outputs of the bistable tilting stages ll and to the input of a regulatory generator G.

The regulatory generator G is constructed in known manner to function in accordance with a characteristic line 13 whereby a regulatory signal Z for regulating a multiplicand function of the system S, as is known, is continuously varied over time r.

In operation, one of the bistable tilting stages receives an instantaneous multiplicand value of the voltage or current of the system S and compares the value to the respective stored k value. Should the instantaneous value be equal and opposite in sign so as to cancel out the respective stored k value, then a signal is emitted from that tilting stage ll comparing the values to the electronic switching stage 12 to initiate a switching operation. The switching stage 12 then, depending on the switched position of the tilting stage emitting the initiation signal, feeds a positive or negative constant value signal to the regulatory generator G. The regulatory generator then becomes activated and begins to regulate the regulatory signal Z continuously over a period of time t until the switching logic either emits a different constant value signal or becomes deactivated.

Referring to FIG. 4, wherein like reference characters as above are used, instead of using a separate regulatory generator, the switching apparatus can include a direct converter K of energy, such as a magnetic direct current converter, in which the dependence or relationship between the voltage U and current I is established by a characteristic line 14. This direct energy converter K is connected to a consumer V, sometimes through the intermediary of a storage battery, to supply its electrical output thereto.

In this case, a peak value meter Sp M-l, Sp M-2 is connected, as above, to the lines l6, 17 between the converter K and consumer V to tap off a predetermined k value of the peak values of the voltage U0: and current la for storage in a switch logic SL. Also, the instantaneous values of the voltage U and current I are tapped off and fed into the switching logic SL for comparison with the k values as above described. Further, a choke coil L and a diode D downstream of the coil L are incorporated in the connecting line 16 between the converter K and consumer V. Also, a switch T is interposed between the choke coil L and diode D so as to be able to short circuit the connecting line 16. This switch T can be constructed as a mechanical contact or as an electronic switch, for example, a transistor or a thyristor and, in any event, is controlled so as to open and close over a suitable line by the switching logic SL.

In operation, in order to carry out one complete trial movement, assuming the voltage U is at an operating value in the converter K, the switching logic SL closes the switch T so as to short the circuit. The voltage then changes in accordance with the law U=L-di/dt. The working point thus is caused to travel along the characteristic line 14 in a direction to make the voltage smaller. When the voltage reaches a k value, e.g. 0.9 U01,

the switching logic is actuated to cause a switch over and thereby causes the switch T to open. At this time, the peak value of the current 13 prevailing at this instant is measured. Because the energy stored in the inductance cannot jump, the current flows onward from the choke coil L over the diode to the consumer V and thus steadily decreases. When the current subsequently reaches its k value, e.g. 0.9 lfi, the switching logic SL again closes the switch T so that the working point again reaches and travels along the characteristic line 14 in the original direction. The voltage then decreases again. At the time of the second switching over, i.e. when the switch T closes, the peak value Ua of the voltage is measured and the k value kUa is fed to the switching logic SL.

This manner of carrying out the optimizing process requires no separate trial movement and no actual regulatory generator is needed. On the contrary, the displacement of the working point produced by the switching character of the magnetic direct current converter is used simultaneously as a trial and basic movement for the optimumizing process. Also, with this embodiment, the expenditure on technical apparatus is decreased still further. This form of carrying out the process is particularly valuable for generators of small output, for example, solar generators.

Referring to FIG. 5, the process can be used in those cases where there is an alteration of the current voltage characteristic line. Such as alteration occurs, for example, with solar cells when a satellite carrying the cells during the course of travel moves towards the sun or moves away from the sun. If, for example, in a first position of the satellite, the point of optimum performance is located at P, on the characteristic line K, then a, and B, are the switch over points of the corresponding trial movement and Ua, and I3, are the peak values of the voltage and current respectively in the switch over points. The corresponding part values are designated k-Ua, and M3,. However, if in a new position of the satellite, the characteristic line has become altered, then the trial movement becomes adapted to the altered characteristic line k and shifts the point of maximum performance in the second case at point P Here once more a, and B are the switch over points of the trial movement, while U0: and [B and the corresponding peak values of the voltage and current respectively. In each of the two characteristic lines the hatched areas in each case represent the optimum performance.

Referring to FIG. 6, for a nonlinear current voltage characteristic line (FIG. 6a) and a nonlinear characteristic (F IG. 6b) for the regulatory generator by means of which a regulatory signal Z alters the voltage U of a system in a steady but nonlinear manner, the voltage pattern plotted against time (FIG. 60) during an optimizing process becomes adjusted to a stationary state, after a few trial movements, with a pendular movement about the point of maximum performance. Correspondingly, the current pattern plotted against time (FIG. 6d) becomes adjusted to a stationary state. It is noted that the diagrams of FIG. 6 are intended to show in a purely schematic way the relationships of the various parameters and do not represent practical examples having data on magnitudes.

The optimizing process for the products of two magnitudes, explained by the example of electrical performance, is obviously also easily applicable to other systems. in each case, it has the advantage of simple construction and of a simple way of carrying out the process. Hereby, in each case, only knowledge or measurement of the two multiplicands, and not of the product itself, is needed. In each case, the optimumizing is effected through influencing the two multiplicands in opposite directions in a trial movement on the characteristic line of the system. Actually, it is noted that the regulatory generator G can be of any desired construction known, for example, electric or hydraulic. Also, the regulatory generator G for controlling the parameters influencing the multiplicands can provide the most different physical sizes, for example, electric tensions, streams, hydraulic or pneumatic pressures, temperatures, location motion, etc. The kind of regulatory generator used and the parameters depend therefore on the kind of system S to be regulated.

What is claimed is:

l. A process of automatically optimizing the product formed by two physical interdependent multiplicand magnitudes comprising the steps of measuring the peak value of one of said multiplicands,

obtaining a predetennined part value of the peak value of said measured peak value,-

subsequently steadily decreasing the instantaneous value of the one multiplicand while simultaneously steadily increasing the instantaneous value of the other of the multiplicands,

thereafter reversing the decreasing of the instantaneous value of said one multiplicand to a steadily increasing value while simultaneously reversing the increasing of the instantaneous value of said other multiplicand to a steadily decreasing value upon said one multiplicand reaching said predetermined part value,

measuring the peak value of said other of said multiplicands upon reversal of the value thereof,

obtaining a predetermined part value of the peak value of the measured peak value of said other multiplicand, and

subsequently reversing the values of said multiplicands upon said other multiplicand reaching said predetermined part value thereof and simultaneously repeating said steps of measuring the peak value and obtaining a part value of said one multiplicand upon reversal of the values of said multiplicands until the optimum product of said multiplicands is obtained.

2. A process as set forth in claim 1 wherein the predetermined part values are between 0.8 and 0.9 of the measured peak values.

3. A switching apparatus for the automatic regulation of a product formed by two physical interdependent multiplicand magnitudes of a system to an optimum value comprising a switching logic connected to the system to receive a measured peak value of each of the multiplicands,

a pair of peak value meters connected to the system and said switching logic, each said meter being connected to the system for measuring the peak value of a respective one of the multiplicands and to said switching logic for emitting a part value of the peak value to said switching logic, and

means connected between said switching logic and the system for steadily increasing or decreasing one of the multiplicands while simultaneously decreasing or increasing, respectively, the other of the multiplicands in respect to a measured value of one of the multiplicands reaching a part value of said one multiplicand.

4. A switching apparatus as set forth in claim 3 wherein said means includes a regulatory generator between said switching logic and the system for receiving a signal from said switching logic and emitting a signal over a regulatory parameter to the system for respectively and alternately increasing and decreas ing the values of the multiplicands.

5. A switching apparatus as set forth in claim 4 wherein said generator is an integrator and is connected to said switching logic to alternately receive a signal therefrom of constant positive or negative value.

6. A switching apparatus as set forth in claim 3 wherein each said peak value meter includes a pair of diode condenser means for measuring the peak values and of emitting the part values.

7. A switching apparatus as set forth in claim 6 wherein said switching logic includes a pair of bistable tilting stages, each stage being connected to a diode condenser means of a respective one of said peak value meters to compare the instantaneous measured value of a multiplicand with a stored part value of said multiplicand to initiate the actuation of said means.

8. A switching apparatus as set forth in claim 3 wherein each said peak value meter is a mean value meter having a time lag whereby at the output of each said means value meter the change in the measured value of a respective multiplicand is immaterial.

9. The combination of a direct energy converter having an electrical output, a consumer, a connecting line between said converter and consumer for conducting the electrical output from said converter to said consumer, and a switching apparatus for automatically optimizing the electrical output of said converter; said switching apparatus including a switching logic connected to said connecting line for receiving a measured instantaneous value of the voltage and the current respectively of the electrical output, a pair of peak value meters connected between said connecting line and said switching logic for measuring the peak value of the voltage and current respectively and for emitting a part value of each measured peak value to said switching logic, a choke coil and a diode downstream of said choke coil connected in said connecting line, and a switch connected in said connecting line between said choke coil and diode and connected to said switching logic for actuation thereby to selectively short CH- cuit said connecting line whereby the values of the voltage and current are respectively steadily increased and decreased in alternation in response to the opening and closing of said switch by said switching logic upon a measured instantaneous value becoming equal to a respective part value in said switching logic.

UNITED STATES PATENT-OFFICE CERTIFICATE OF CORRECTION Patent No. 3: 6s 9 D t d December 7 1971 lnventofls) Andreas Boehringer It is certified that error appears in the above-identified patent and that said Letters Patent are hereby corrected as shown below:

Column 2, lines 7 61 to 65, delete "with respect to how... .cell generator S. and insert --the invention, instead of using a peak value meter, a mean .value meter may be used provided that this mean value meter has such a time lag that as its output the change in the measured value p appearing as a result of the trial movement is immaterial.-

Column *3, between lines 5 and 6, insert "taken, becomes steadily decreased in value If the multiplicand has now assumed a value which corresponds to the above described tapped off percentage k of the peak value, then the output of the logic changes its state, and the working point, of the system-- I Column 3, lines 14 to 17, cancel "Z has been. generator, S"

Signed and sealed this 19th day of September 1972.

Attest:

EDWARD M.FLETCHER,JR. ROBERT GOTTSCHALK Attesting Officer v Commissioner of Patents FORM PO-105O (10-69) uscoMM-Dc 60376-P69 UTS. GOVERNMENT PRINTING OFFICE i959 O 35(-334 Patent No. 39 9 9 D d December 7 1971.

Andreas Boehringer Inventor(s) It is certified that error appears in the above-identified patent and that said Letters Patent are hereby corrected as shown below:

Column 2, lines 61 to 65, delete "with respect to how ocell generator S. and insert the invention, in stead of using a peak value meter, 8. mean value meter may be used provided that this mean value meter has such a time lag that as its output the change in the measured value appearing as a result of the trial movement is irmnaterialw- Column 3, between lines 5 and 6 insert "taken, becomes steadily decreased in value, If the multiplicand has now assumed a value which corresponds to the above described tapped off percentage k of the peak value, then the output of the logic changes its state, and the working point of the system Column 3, lines 14 to 17, cancel Z has been... generator S" Signed and sealed this 19th day of September 1972.

(SEAL) Attest:

EDWARD II.FLETCHER,JR. ROBERT GOTTSCHALK Attesting Officer 7 Commissioner of Patents FORM po'wso (10439) USCOMM-DC 60376-P69 9 U.S. GOVERNMENT PRINTING OFFICE: 1969 0*366-334

Claims

1. A process of automatically optimizing the product formed by two physical interdependent multiplicand magnitudes comprising the steps of measuring the peak value of one of said multiplicands, obtaining a predetermined part value of the peak value of said measured peak value, subsequently steadily decreasing the instantaneous value of the one multiplicand while simultaneously steadily increasing the instantaneous value of the other of the multiplicands, thereafter reversing the decreasing of the instantaneous value of said one multiplicand to a steadily increasing value while simultaneously reversing the increasing of the instantaneous value of said other multiplicand to a steadily decreasing value upon said one multiplicand reaching said predetermined part value, measuring the peak value of said other of said multiplicands upon reversal of the value thereof, obtaining a predetermined part value of the peak value of the measured peak value of said other multiplicand, and subsequently reversing the values of said multiplicands upon said other multiplicand reaching said predetermined part value thereof and simultaneously repeating said steps of measuring the peak value and obtaining a part value of said one multiplicand upon reversal of the values of said multiplicands until the optimum product of said multiplicands is obtained.

3. A switching apparatus for the automatic regulation of a product formed by two physical interdependent multiplicand magnitudes of a system to an optimum value comprising a switching logic connected to the system to receive a measured peak value of each of the multiplicands, a pair of peak value meters connected to the system and said switching logic, each said meter being connected to the system for measuring the peak value of a respective one of the multiplicands and to said switching logic for emitting a part value of the peak value to said switching logic, and means connected between said switching logic and the system for steadily increasing or decreasing one of the multiplicands while simultaneously decreasing or increasing, respectively, the other of the multiplicands in respect to a measured value of one of the multiplicands reaching a part value of said one multiplicand.

4. A switching apparatus as set forth in claim 3 wherein said means includes a regulatory generator between said switching logic and the system for receiving a signal from said switching logic and emitting a signal over a regulatory parameter to the system for respectively and alternately increasing and decreasing the values of the multiplicands.

9. The combination of a direct energy converter having an electrical output, a consumer, a connecting line between said converter and consumer for conducting the electrical output from said converter to said consumer, and a switching apparatus for automatically optimizing the electrical output of said converter; said switching apparatus including a switching logic connected to said connecting line for receiving a measured instantaneous value of the voltage and the current respectively of the electrical output, a pair of peak value meters connected between said connecting line and said switching logic for measuring the peak value of the voltage and current respectively and for emitting a part value of each measured peak value to said switching logic, a choke coil and a diode downstream of said choke coil connected in said connecting line, and a switch connected in said connecting line between said choke coil and diode and connected to said switching logic for actuation thereby to selectively short circuit said connecting line whereby the values of the voltage and current are respectively steadily increased and decreased in alternation in response to the opening and closing of said switch by said switching logic upon a measured instantaneous value becoming equal to a respective part value in said switching logic.