T 0430/88 () of 14.8.1990

Summary of Facts and Submissions

I. European patent No. 0 083 945 was granted to the Appellant on 7 May 1986 in response to European patent application No. 83 100 177.1, filed on 11 January 1983 and claiming priority of 11 January 1982.

II. A Notice of Opposition was filed on 5 February 1987 by the Respondent requesting revocation of the patent on the ground of lack of inventive step. In support of his request, the Opponent referred in particular to the prior art document D1: DE-A-3 023 135.

III. In response to the opposition the Patentee amended with letter of 21 September 1987 the two independent Claims 1 and 2 of the patent, but otherwise objected to the Opponent's submissions.

IV. After considering the facts and submissions, the Opposition Division revoked the patent by a decision dated 23 June 1988.

V. On 1 September 1988 the Patentee lodged an appeal against the decision and paid the appeal fee on the same day.

A Statement of Grounds of appeal was filed on 18 October 1988. The Respondent (Opponent) replied with a letter dated 12 December 1988. After having received a further letter from the Appellant dated 9 March 1989, the Rapporteur sent a communication dated 13 July 1989 with an analysis of the claimed method for controlling an induction motor in the light of the prior art represented by D1. In the result, the Rapporteur voiced doubts in respect of inventive step.

VI. With letter of 22 January 1990 the Appellant filed a slightly amended Claim 1 and submitted analytical considerations of his own. The Respondent maintained his objections in a letter dated 29 March 1990.

VII. The independent Claims 1 and 2 read as follows:

"1. A method for controlling an induction motor by slip-frequency control, wherein the primary current of the induction motor (5) is divided into an excitation component (Im) and a torque component (I2) which are independently controlled, the difference ( ; Im) between an instruction value ( *; Im*) indicative of the desired magnetic flux and a value ( , Im) indicative of the actual magnetic flux ( ) of said induction motor (5) is detected, a slip frequency correction value ( fs*) is determined on the basis of said difference ( , Im), and the value of the slip frequency (fs) of said induction motor (5) is controlled such that it is equal to the sum of a slip frequency instruction value (fs*) evaluated from the torque component (I2*) and said slip frequency correction value ( fs*), characterized in that the slip frequency correction value ( fs*) is determined by multiplying an instruction value (fs* or I2*) derived from the torque component (I2*) and the value of said difference ( ;Im).

2. An apparatus for controlling induction motors, comprising: means (1,2) for providing electric energy, an inverter (4) for generating an a-c voltage from the output of said means (1,2) for providing electric energy wherein the phase and frequency of the inverter (4) are controlled by an instruction signal and the output of the inverter (4) drives an induction motor (5), speed setting means (7) for said induction motor (5), speed detecting means (6) for said induction motor (5), a speed control unit (8) for generating a torque current instruction signal (I2*) corresponding to the difference between the output of said speed setting means (7) and the output of said speed detecting means (6), means (23;12) for setting an instruction value ( *; Im*) indicative of the desired magnetic flux of said induction motor (5), means (21,22; 26) for detecting a value ( , Im) indicative of the actual magnetic flux of said induction motor (5), means (24) for detecting the difference ( ; Im) between the instruction value ( *; Im*) indicative of the desired magnetic flux and the value indicative of the actual magnetic flux ( ; Im), control means (11-19) generating a control signal for the induction motor (5) on the basis of the output of said speed control unit (8), multiplying means (25) for generating a slip frequency signal ( fs*) by multiplying a signal (fs*; I2*) derived from the output of the speed control unit (8) by a signal derived on the basis of the output ( ; Im) of said means (24) for detecting the difference, characterized by a slip frequency operation unit (9) for generating a slip frequency instruction signal (fs*) on the basis of the output (I2*) of said speed control unit (8), wherein the resulting slip frequency instruction signal (fs*) and the output of the means (24) for detecting the difference ( ; Im) are used for generating a slip frequency correction signal ( fs*), means (10) for generating a primary frequency instruction value (f1*) of the induction motor (5) by adding said outputs of the slip frequency operation unit (9), multiplying means (25) and speed detecting means (6)."

VIII. Thus, the Appellant requests the maintenance of the patent as amended by the Claims 1 and 2 cited above.

The Respondent has maintained his request to dismiss the appeal.

IX. The Appellant's (Patentee's) submissions can be summarised as follows:

The present invention is an improvement of the method and apparatus for controlling an induction motor as disclosed in D1, in particular Figure 9. Although in the present invention the control circuit of Figure 6 has substantially the same function as the control circuit of D1, the ways of achieving the same result and function are different.

In the reasons of the revocation decision, the Opposition Division made a mathematical analysis of both the control method of D1 and that of the present patent. It is no more shown by this analysis that the method and apparatus of the present invention as well as that of the cited document perform the said identical function. However, the Opposition Division had disregarded that the physical quantities to be controlled and the respective circuits are quite different.

It is agreed that the identical function of the different embodiments can be described by the following rule:

w + ws = X.I2*(R2 + R2) = X.I2*.R2 + X.I2*. R2 which means that the slip frequency ws of the induction motor is proportional to the product of the torque component I2* of the motor current and the secondary resistance R2 of the motor, and that a temperature dependent variation R2 of the secondary resistance entails a corresponding variation ws of the slip frequency.

The slip frequency correction value ws (=2 . fs), which is generated as the output of a feedback configuration evaluating the difference between an instruction value and an actual value of the magnetic flux, is introduced in the calculation of the motor control in different ways.

D1, Figure 9 discloses a single control circuit which generates an inclusive output signal

(ws + ws) = X.I2*(R2 + R2) with a determination of the secondary resistance.

In contrast to this, the present invention provides two separate control circuits, the main one of which forms ws = X.I2*.R2 and the auxiliary one separately forms ws = X.I2*. R2.

D1, Figure 9 does not teach or suggest this separate determination or formulation of the slip frequency correction value ws.

The invention leads to a very different physical structure of the circuit: the separate ws-control circuit allows an easy modification of a slip frequency control circuit having no secondary resistance compensation. According to the invention, there is no need to determine the actual amount of the secondary resistance R2 + R2 and, consequently, to include a physical potentiometer in the control circuit as in D1, Figure 9 (potentiometer 40). In view of these physical differences, not suggested by the prior art, the Opposition Division's understanding, based solely on mathematical considerations, is regarded as an undue hindsighted view.

Further embodiments of D1, disclosed in Figures 3 and 4, have the disadvantage that the slip frequency correction value is solely derived from the difference between the instruction value indicative of the desired voltage and a value indicative of the actual voltage of the induction motor, viz there is no multiplication of the value of this difference with the instruction value of the torque component. This means that the slip frequency correction caused by variation of the torque component must be effected via the feedback circuit. Therefore, when the instruction value of the torque component is changed quickly, the slip frequency correction value changes with time lag and the response is not high speed.

X. The Respondent's (Opponent's) submissions can be summarised as follows:

The Appellant admits that the alleged invention aims at achieving the same control characteristics as the prior art according to D1. However, his maintenance that the invention lies in the application of different technical means is objected to. Figure 9 of D1 comprises, like the alleged invention, two control channels: one channel with the elements 18,42 and 26 for the generation of the slip frequency instruction value and another channel, comprising the elements 93,44,34,41,42 and 26 for the generation of the slip frequency correction value.

In the independent claims of the patent in suit, there is no "auxiliary loop" discernible which would not already be present in the cited reference.

Reasons for the Decision

1. The appeal is admissible.

2. Novelty D1 can be regarded as disclosing the closest prior art:

The primary current of an induction motor 15 is controlled via an AC-AC-frequency converter 400 (Figures 3 to 7) or 100 (Figure 9) in response to a primary current instruction I1* which, corresponding to the two notional components into which an induction motor current can be divided, is derived from an instruction value Im* for the excitation component and an instruction value I2* for the torque component. The value of the slip frequency ws (=2 fs) of the induction motor is controlled such that it is equal to the sum (ws* + ws) of a slip frequency instruction value ws*(=2 fs*) and a slip frequency correction value ws(=2 . fs*). From Figures 3 and 4 can be derived that the slip frequency instruction value is evaluated from the torque component I2* via a calculator 25 which represents the relation ws* = K.I2* (page 12, line 32 to page 13, line 15).

Also according to Figures 3 and 4, the slip frequency correction value ws is determined on the basis of a difference between a motor voltage instruction value e1 and an actual motor voltage value e1+ e (page 10, lines 24 to 29). Since the motor voltage instruction value is determined by the product of a magnet flux instruction value * and the motor voltage frequency value (w1+ ws) the said instruction and actual voltage values are indicative of the respective flux values (cf. page 11, lines 14 to 17 and page 16, line 32 to page 17, line 1). A generation of the difference that is directly based on the respective values of the magnetic flux is disclosed in Figure 9 of D1.

According to the embodiments disclosed in D1, Figures 3 and 4, the said difference (=feedback error) determines the slip frequency correction value ( ws = 2 fs*) directly. The method of present Claim 1 differs therefrom by the characterising feature according to which the slip frequency correction value is determined by multiplying the said difference with an instruction value (fs* or I2*) derived from the torque component (I2*).

According to the embodiment of Figure 9 of D1, the said magnetic flux difference ( ) determines a secondary resistance correction value ( R2) instead of a slip frequency correction value. Together with a predetermined and constant secondary resistance value R2 it is fed into a multiplier 42 in which it is multiplied with the torque component instruction value I2*. The result is that it is not the slip frequency correction value ( ws) but the sum (ws*+ ws) of the slip frequency instruction value (ws*) and the slip frequency correction value ( ws) that is determined by multiplying an instruction value (I2*) derived from the torque component and the value of said difference.

Therefore, the method of Claim 1 can be regarded as being novel.

3. Inventive step

3.1. The theory on which the disclosure of the patent in suit is based, is given by a relation that describes the instruction of the slip frequency in a controlling method according to the precharacterising portion of Claim 1:

1 r2 1 fs= ( ) (1) 2 I2+Im Im where r2, I2 and Im denote quantities that are proportional to the secondary resistance, secondary leakage reactance and excitation inductance of the induction motor, Im denotes a quantity proportional to an excitation current component of the primary current, and I 2 denotes a quantity proportional to a torque current component (cf. column 1, lines 32 to 44 of the patent).

With the assumption that the excitation current remains constant (column 1, lines 45 to 47 of the patent in suit) and with the knowledge that it is in the first place the secondary resistance that varies depending upon the operation conditions of the motor (D1, page 6, line 24 to page 7, line 9), the equation (1) can be written:

ws(=2 fs) = K.r2.I2 or with ws and r2 expressed as a sum out of a constant value and a value representing the variation caused by the operating conditions:

(ws+ ws) = K(R2+ R2)I2 (2) This is equation (8) of D1 (page 19), and it is thus clear that this known relation and the embodiments disclosed in D1 and meeting the conditions of this relation must be taken as the starting point for assessing the question of inventive step in the present case. As a matter of course, equation (8) can also be written in the form ws+ ws = K.R2.I2 + K. R2.I2 (3)

3.2. Figure 3 of D1 discloses an embodiment following this equation (3): a first control path (elements 18,25 and 35) forms the slip frequency instruction value ws* = K.R2.I2* and a second path, derived from the feedback arrangement (323,322,93,34) maintaining the magnetic flux constant (viz. also Im constant) and thus generating the necessary slip frequency correction value ws = K. R2.I2*. The slip frequency instruction value ws* and the slip frequency correction value ws are added at the securing point 35 which outputs the desired sum (ws* + ws) of the slip frequency instruction value and the slip frequency correction value.

3.3. Figures 5, 6 and 9 of D1 disclose embodiments which follow equation (2) above: a first control path (element 18) forms the torque component instruction value I2* and a second path, derived from the feedback arrangement (93,34,40,41) forms the constant secondary resistance value plus the secondary resistance correction value (R2 + R2). The torque component instruction value I2* and the corrected secondary resistance value (R2 + R2) are multiplied in the multiplier 42 which outputs the desired sum (ws + ws) of the slip frequency instruction value and the slip frequency correction value (according to equation (2) above).

3.4. The embodiments according to Figures 3 and 4 of D1 have the disadvantage that the flux dependent feedback arrangement generating the slip frequency correction value ws = K.R2.I2* must compensate not only for the slow variations of the secondary resistance R2 but also for the quick variations of the instruction value of the torque component I2*. Thus, in these embodiments, the feedback circuit must quickly react to variations in the primary current. Therefore, there would be an adverse effect on the total control system if the speed of the auxiliary loop were low.

In the embodiment according to Figure 9 of D1 the feedback circuit must, in essence, only respond to the slow variations of the secondary resistance R2, and a low speed of the feedback circuit has no substantial adverse effect on the total control system.

The Board is satisfied that a person skilled in the art of feedback control will readily realise this difference in the dynamic control characteristics of the known controlling methods, despite the fact that it is not expressly discussed in the citation.

It may be regarded as a disadvantage of the embodiment according to Figure 9 of D1 as against the embodiments shown in Figures 3 and 4 that the control path had to be modified by the substitution of a multiplier (42) for the calculator (25) which has only to represent the constant proportionality between I2* and ws*. Then, the problem underlying the claimed method can be understood as the desire to maintain a first control path (as in Figures 3 and 4 of D1) for forming the uncorrected frequency instruction value ws* = K.R2.I2* and to modify the second path in such a way that the feedback arrangement need not respond to the quick variations of the primary current (as according to Figure 9 of D1).

The Board is satisfied that the solution is directly derivable from equation (3) above:

The main loop in the embodiments of D1, Figures 3 and 4 produces at the summing point 35 the signal ws* = K.R2.I2*. the feedback circuit produces the signal ws = K. R2.I2*.

Now, in order to relieve the feedback circuit of the compensation of variations of the torque component I2* it suffices to multiply the output of the amplifier 34 with a signal proportional to I2*, e.g. ws*. This follows readily from a simple evaluation of the known mathematical relations and applies, as it were, the same principle as applied in Figure 9 of D1: the multiplication of the output of the feedback circuit by a signal proportional to I2*.

3.5. The foregoing considerations can be summarised by taking up the relation represented by equations (2) or (3) above and setting the signal provided by the feedback circuit in brackets:

D1, Figures 3 and 4: (ws*) + ( ws) = K(R2.I2*) + [K. R2.I2*](4) D1, Figure 9: (ws + ws) = K.(R2 +[ R2]).I2*(5) Patent in suit (ws*) + ( ws) = K.(R2.I2*) + [K. R2].I2*(6) These equations reveal that the patent in suit makes use of a principle known from D1, Figure 9 (multiplication of the feedback signal by the instruction value derived from the torque component) with the only difference that the first control path generating the unamended slip frequency instruction value ws* = (R2.I2*) remains unaltered as in D1, Figures 3 and 4 making use of the mathematically evident identity (a+b).c = a.c + bc.

3.6. Since the above considerations are starting from the known prior art and make use of only simple mathematical transformations readily envisageable by a skilled person and not introducing any basically new concept, the Appellant's contention that the revocation is based on a hindsight view cannot be upheld.

The method of Claim 1 is therefore only an obvious modification of the prior art, and Claim 1 cannot be maintained because its subject-matter lacks an inventive step.

4. Despite the fact that the appeal has to be dismissed on the ground that Claim 1 cannot be maintained, the Board wishes to indicate that an examination of its own motion of the independent apparatus Claim 2 and the dependent Claims 3 to 6 has not revealed any inventive subject-matter.

ORDER

For these reasons, it is decided that:

The appeal is dismissed.