Method of torque prediction for automotive air conditioning compressor

Abstract

The torque of a vehicular air-conditioning compressor is predicted by the steps of starting a vehicle air-conditioning system having an engine and engine control module for controlling multiple vehicle functions, calculating a steady state torque value using an rpm value from the engine, calculating an engine RPM transient torque value using the rpm value from the engine, calculating a electronic control valve current transient torque value using a current value applied to an electronic control valve, selecting a final torque value from a group consisting of the steady state torque value, the engine RPM transient torque value, and the electronic control valve transient torque value, and providing a final torque value to the vehicle engine control module to control a predetermined vehicle function.

Claims

1. A method of predicting torque for an automotive air conditioning compressor comprising the steps of: starting an air-conditioning system for a vehicle having an engine and ECM for controlling multiple vehicle functions; calculating a steady state torque value using an RPM value from the engine; calculating an engine RPM transient torque value using the RPM value from the engine; calculating a Iecv transient torque value using a current value applied to an electronic control valve; selecting a final torque value from a group consisting of said steady state torque value, said engine RPM transient torque value, and said Iecv transient torque value; providing said final torque value to the ECM of the vehicle to control a predetermined function of the engine, and controlling a vehicle engine by adjusting an engine performance of the vehicle engine based on the final torque value, wherein said step of calculating a steady state torque value using an RPM value from the engine comprises the steps of: selecting a first set of constants which are employed to calculate a maximum steady state torque value T.sub.max; selecting a second set of constants which are employed to calculate a base steady state torque value T.sub.base; selecting a third set of constants which are employed to calculate a minimum steady state torque value T.sub.min; comparing T.sub.base with T.sub.max and T.sub.base; equating T.sub.max as a steady state torque value T.sub.ss if T.sub.base>T.sub.max; equating T.sub.min as a steady state torque value T.sub.ss if T.sub.baseT.sub.max and T.sub.base<T.sub.min; and equating T.sub.base as a steady state torque value T.sub.ss if T.sub.baseT.sub.max and T.sub.baseT.sub.min.

2. A method of predicting torque for an automotive air conditioning compressor comprising the steps of: starting an air-conditioning system for a vehicle having an engine and ECM for controlling multiple vehicle functions; calculating a steady state torque value using an RPM value from the engine; calculating an engine RPM transient torque value using the RPM value from the engine; calculating a Iecv transient torque value using a current value applied to an electronic control valve; selecting a final torque value from a group consisting of said steady state torque value, said engine RPM transient torque value, and said Iecv transient torque value; providing said final torque value to the ECM of the vehicle to control a predetermined function of the engine, and controlling a vehicle engine by adjusting an engine performance of the vehicle engine based on the final torque value, wherein said step of calculating an ERMP transient torque value using the RPM value from the engine comprises the steps of: calculating a rate of engine RPM change EPRM; calculating a maximum threshold rate of EPRM as EPRM.sub.Lim; equating engine RPM transient torque value T.sub.ERPMT to zero or null if EPRM<EPRM.sub.Lim; equating engine RPM transient torque value T.sub.ERPMT to T.sub.ss(t.sub.i-t.sub.Lim) if EPRMEPRM.sub.Lim and tt.sub.Lim; and equating engine RPM transient torque value T.sub.ERPMT to T.sub.ss(t.sub.0) if EPRMEPRM.sub.Lim and t<t.sub.Lim.

3. A method of predicting torque for an automotive air conditioning compressor comprising the steps of: starting an air-conditioning system for a vehicle having an engine and ECM for controlling multiple vehicle functions; calculating a steady state torque value using an RPM value from the engine; calculating an engine RPM transient torque value using the RPM value from the engine; calculating a Iecv transient torque value using a current value applied to an electronic control valve; selecting a final torque value from a group consisting of said steady state torque value, said engine RPM transient torque value, and said Iecv transient torque value; providing said final torque value to the ECM of the vehicle to control a predetermined function of the engine, and controlling a vehicle engine by adjusting an engine performance of the vehicle engine based on the final torque value, wherein said step of calculating a Iecv transient torque value using a current value applied to an electronic control valve comprises the steps of: calculating a rate of control valve current change I; calculating a maximum threshold rate of change of EVC current I.sub.Lim; equating electronic control valve transient torque value T.sub.IT to zero or null if there is no Iecv transient I<I.sub.Lim; equating electronic control valve transient torque value T.sub.IT to F.sub.1*Pd-F.sub.0 if II.sub.Lim and PdPd.sub.Liml; and equating electronic control valve transient torque value T.sub.IT to zero or null if II.sub.Lim and Pd>Pd.sub.Liml and Pd>Pd.sub.Limh.

4. The method of predicting torque of claim 3, wherein said step of selecting a final torque value T.sub.F from a group consisting of said steady state torque value T.sub.SS, said engine RPM transient torque value T.sub.ERPMT, and said Iecv transient torque value T.sub.IT comprises the steps of: equating T.sub.F to T.sub.SS when T.sub.ERPMT=0 and T.sub.IT=0; equating T.sub.F to T.sub.IT when T.sub.ERPMT=0 and T.sub.IT0; equating T.sub.F to T.sub.ERPMT when T.sub.ERPMT0 and T.sub.IT=0; and equating T.sub.F to T.sub.IT when T.sub.ERPMT0 and T.sub.n0.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

(1) Although the drawings represent a single embodiment of the present invention, the drawings are not necessarily to scale and certain features may be exaggerated in order to better illustrate and explain the present invention. The exemplification set forth herein illustrates an embodiment of the invention, in one form, and such exemplifications are not to be construed as limiting the scope of the invention in any manner.

(2) FIG. 1, is a schematic diagram of an air conditioning system configured for automotive application embodying the present invention;

(3) FIG. 2, is a graphical depiction of empirical bench test data and calculated compressor torque data of the air conditioning system of FIG. 1 under steady-state conditions wherein accuracy is 2 Nm or less;

(4) FIG. 3, is a graphical depiction of empirical bench test data illustrating transient compressor torque when the electronic control valve (ECV) is transitioned from off to on;

(5) FIG. 4, is a graphical depiction of empirical bench test data illustrating transient compressor torque when the electronic control valve (ECV) is transitioned from on to off;

(6) FIG. 5, is a graphical depiction of empirical bench test data illustrating transient compressor torque during acceleration and de-acceleration;

(7) FIG. 6, is a master flow diagram for the compressor torque prediction system embodied in the present invention;

(8) FIG. 7, is a flow diagram for calculating steady state compressor torque in practicing the present invention;

(9) FIG. 8, is a flow diagram for calculating electronic control valve current (Ievc) transient torque in practicing the present invention;

(10) FIG. 9, is a flow diagram for calculating engine revolutions per minute (ERPM) transient torque in practicing the present invention;

(11) FIG. 10, is a flow diagram for selecting a final torque (T.sub.F) value in practicing the present invention;

(12) FIG. 11, is a glossary of definitions and associated instrumentation terminology employed in describing and practicing the present invention;

(13) FIG. 12, is Table 1: Coefficients for Base Equation; and

(14) FIG. 13, is Table 2: Coefficients for Maximum and Minimum Torque Equations.

(15) Although the drawings represent embodiments of the present invention, the drawings are not necessarily to scale and certain features may be exaggerated in order to illustrate and explain the present invention. The exemplification set forth herein illustrates an embodiment of the invention, in one form, and such exemplifications are not to be construed as limiting the scope of the invention in any manner.

DETAILED DESCRIPTION OF INVENTION

(16) Referring to FIG. 1, the present invention is preferably implemented or practiced in controlling an air conditioning system 10 installed within a ground vehicle such as an automobile 12. The prime mover of the air conditioning system 10 is a compressor 14 which is mechanically driven via a belt drive and an electromagnetic clutch (not illustrated) by a motor/engine 16 of the automobile 12. The compressor 14 compresses hot refrigerant vapor and pumps it to a condenser 18 where it is condensed ultimately to a liquid form. The still pressurized liquid refrigerant from condenser 18 is forced through an expansion valve 20, after which, in its expanded, cooled state it enters an evaporator 22 disposed within an airflow passage formed by a ventilation module 24 within an automobile passenger space. The heated refrigerant vapor continues on to re-enter compressor 14 to renew the cycle.

(17) The low-pressure refrigerant flowing through the evaporator 22 exchanges heat with air flowing through ventilation module 24, which is influenced by a blower 26. Blower 26 is disposed in an air duct 28 through which air for air conditioning passes, and air drawn from an air suction port for air conditioning 30 is directed to evaporator 22 by the blower 26. A portion of the air having passed through the evaporator 22 is directed to a heater unit 32 disposed at a downstream position, and the rate between the amount of air to be passed through the heater unit and the amount of air to be bypassed is adjusted by an air mixing damper 34. In this embodiment, an air temperature sensor 36 at the evaporator 22 exit is provided at the exit side of the evaporator 22 for detecting the temperature of the air after passing through the evaporator 22, and the detected signal is inputted into an air conditioning control unit 38 performing the air conditioning. Signals from various other system sensors 40 are also inputted into the air conditioning control unit 38. An engine control unit 42 (ECU) associated with engine 16 is also electrically interfaced with the air conditioning control unit 38. Both the air conditioning control unit 38 and the ECU 42 contain control circuitry, such as a microprocessor, and volatile and/or non-volatile memory operable to practice the torque prediction method described herein below as well as to affect control of the overall air condition system 10, such as by outputting a displacement control signal to the compressor 14.

(18) An embodiment of the invention is an apparatus and methodology used to calculate or predict the torque of an automotive air conditioner in real time for steady state and transient conditions. This methodology was derived to utilize data acquired from sensors readily available on a standard production vehicle. Thus, no additional expense or costs would be incurred to implement this technique. The torque is predicted utilizing the following inputs: Control valve current Outside air temperature Evaporator air outlet temperature Engine speed High side pressure Evaporator blower voltage
Torque, estimated=B1*Vb*Teao*(ToaTeao)*((Pd/Ps)^K11)/CRPM,

(19) where: CRPM=Compressor RPM B1=Correlation Constant Vb=Blower Voltage Teao=Evaporator Air Out Temperature Toa=Outside Air Temperature Pd=High Side Pressure Ps=Saturation Pressure of Refrigerant at Teao (estimated suction pressure) K1=(k1)/k k=specific heat ratio for the superheated refrigerant.

(20) An improved relationship, is: Torque, correlated=2.7*SQRT(Torque, estimated)

(21) This relationship is acceptable to estimate torque at a quasi-steady condition, but may not predict change in torque when, for example, ECV current is changing.

(22) Base Equation

(23) A base equation was derived from the above relationship based on empirical observations and inclusion of an ECV current parameter. Additional equations were derived empirically and are used for unique conditions to improve the accuracy of the calculation for these unique conditions.

(24) The base equation is:

(25) Compressor Torque is calculated as:
Torque=A0+A1*P*(V).sup.A2*T*((P/T).sup.A31)*(I).sup.A4/Rpm

(26) Where: P=High Side Pressure V=Evaporator Blower Voltage T=Evaporator Air Temperature Differential T=Evaporator Air Outlet Temperature I=Control Valve Current Rpm=Engine Speed A0=Torque Offset Constant A1=Torque Scaling Constant A2=Evaporator Scaling Constant A3=Pressure Temperature Scaling Constant A4=Current Scaling Constant

(27) Steady state bench test results for this equation are shown in FIG. 2.

(28) For the steady state conditions tested, the accuracy is 2 Nm or less.

(29) Additional steady state accuracy can be achieved by establishing a set of constants for the based equation for a given rpm range. An example is shown below.

(30) The coefficient values used for a 6CVCc on a T3000 system are shown in Table 1, depicted in FIG. 12.

(31) Advanced Torque Prediction Method

(32) Transient bench test results are shown in FIGS. 2-5. As shown in the transient plots, the predictive accuracy of the base equation, shown as TP, is not equivalent to the steady state results. Additional equations were empirically derived for transient events as well as when Iecv (electronic control valve current) is below a calibratable threshold or the compressor is at full stroke. The results for the additional equations are shown on the transient plots as Adv TP. As shown on the transient plots, the predictive accuracy is greatly improved by utilizing the additional equations.

(33) Transient Conditions

(34) Both compressor rpm and ECV current can change rapidly enough to adversely affect the predicted torque during the transient event. To improve prediction accuracy it is possible to make note of a change in compressor rpm or ECV current and to mathematically compensate for the transient.

(35) RPM

(36) In the event of a rapid change in compressor RPM, the following strategy is utilized: Detect an RPM change above a predetermined threshold, called the trigger limit. Calculate torque with the base equation and maintain this value for three seconds. After 3 seconds, restart torque calculation using the Table 1 coefficients from the initial

(37) RPM conditions and the P, T, T, V, Iecv, and RPM values stored in computer memory for 3 seconds (i.e. P, T, T, V, Iecv, and RPM are updated every second, but are 3 seconds old.) Verify that RPM variation has stabilized below the trigger limit. Change to the coefficients for the new RPM and remove the 3 second calculation delay.

(38) The trigger limit for RPM change is calibratable and dependent on compressor and system type.

(39) Iecv

(40) In the event of a rapid change in Iecv, calculate torque based on a linear relationship of Pd with an offset and return to the base equation when Pd stabilizes.

Example

(41) 6CVCc on T3000 system. Iecv starts at 0 A and step changes to 1.0 A.
Torque=0.045Pd36 until Pd2Pd1<5

(42) Predictive accuracy was also improved by adding equations to establish a minimum and maximum compressor torque value for a given set of conditions. The following equations were established to achieve this objective:
Minimum Torque=C0+C1*E rpm
Maximum Torque=D0+D1*Pd

(43) Where:

(44) C0=min torque offset coefficient. Is calibratable based on compressor, system and RPM.

(45) C1=min torque coefficient. Is calibratable based on compressor, system and RPM.

(46) D0=max torque offset coefficient. Is calibratable based on compressor, system and RPM.

(47) D1=max torque coefficient. Is calibratable based on compressor, system and RPM.

(48) The minimum torque equation is used when Iecv is less than min Iecv, a calibratable term.

(49) The maximum torque equation is used when Iecv is greater than max Iecv, a calibratable term, or the predicted torque is greater than the maximum calculated torque.

(50) The coefficient values, min Iecv and max Iecv used for a 6CVCc on a T3000 system are shown in Table 2 depicted in FIG. 13.

(51) This predictive system is intended for real time vehicle use and as such is dependent on the communication speed or frequency of information update between the HVAC module and the ECM.

(52) Note that the calibratable values shown in the preceding are for a specific compressor type on a specific AC system. A different combination of compressor type and AC system may require different calibratable values. It is expected that such values would be unchanged for a given compressor type and AC system combination. That is to say, once the calibratable values are established for a type of compressor on a vehicle AC system, they would be applicable for the entire production run of that compressor/system combination.

(53) Referring to FIG. 6, a master flow diagram 44 for the compressor torque prediction system embodied in the present invention is illustrated wherein a calculation logic block 46 is initialized by an AC start-up step 48, and which, in turn provides a final torque T.sub.f output logic signal to an engine control module (ECM) 58. AC start-up step 48 feeds a logic step 50 for calculating steady state torque (T.sub.SS) which, in turn, feeds a first input of a logic step 52 wherein final torque (T.sub.f) is selected. Logic step 50 also feeds a T.sub.SS signal to a logic step 54 which calculates engine revolutions per minute (ERPM) transient torque, which feeds an engine RPM from a host vehicle engine control module (speed sensor or a calculation based upon vehicle speed) to a second input of logic step 52 as T.sub.erpmt. Finally, logic step 56 calculates IECV transient torque, feeding an electronic control valve transient torque signal (T.sub.it) to a third input of logic step 52.

(54) Referring to FIG. 7, a flow diagram 60 for calculating steady state compressor torque in practicing the present invention expands on a step 62 of calculating steady state torque T corresponding with step 50 of FIG. 6. An engine speed (ERPM) transducer 64 feeds an input to a logic block 66, which operates to select constants based on ERPM, and outputs selected constants D.sub.0 and D.sub.1 to a first input of a following logic block 70. A discharge pressure signal (Pd) from discharge pressure transducer 68 feeds a second input of logic block 70. Logic block 70 operates to calculate maximum steady state torque (T.sub.max). Logic block 70 outputs a T.sub.max signal to a first input of a yes/no logic block 72.

(55) An ERPM transducer 74 feeds an input to a logic block 76, which operates to select constants based on ERPM, and outputs selected constants A.sub.0, A.sub.1, A.sub.2, A.sub.3 and A.sub.4 to a first input of a following logic block 78. A discharge pressure signal (Pd), an evaporator blower fan voltage signal (V), a thermister ambient temperature signal (Ta), an electronic control valve current signal from measurement of current applied to ECV (I) and a ERPM signal, from transducers collectively illustrated as 80, provide a second input to logic block 78, which are also employed in calculating a base steady state torque torque (T.sub.base) output from logic block 78 which is fed to a second input of yes/no logic block 72.

(56) An ERPM transducer signal and a thermistor ambient temperature (Ta) signal from transducers collectively illustrated as 82 feed a logic block 84 which operates to select constants based on ERPM, and output selected constants C.sub.0 and C.sub.1 to a first input of a following logic block 86. A EPRM transducer 88 feeds a signal to a second input of logic block 86, which are employed in calculating a minimum calculated steady state torque (T.sub.min) as an output from logic block 86.

(57) If T.sub.base>T.sub.max, a first output of yes/no logic block 72 feeds a logic block 90 which equates T.sub.SS with T.sub.max, and, in turn, feeds a T.sub.SS output logic block 92. If T.sub.baseT.sub.max, a second output of yes/no logic block 72 feeds a second yes/no logic block 94. The T.sub.min output of logic block 86 feeds a second input of yes/no logic block 94. Yes/no logic blocks 72 and 94 are logically joined by common logic as set forth in phantom logic block 100. If T.sub.base<T.sub.min, a first output of yes/no logic block 94 feeds a logic block 98 which equates T.sub.SS with T.sub.max, and, in turn, feeds T.sub.SS output logic block 92. If T.sub.baseT.sub.min, a second output of yes/no logic block 94 feeds a logic block 96 which equates T.sub.SS with T.sub.base and, in turn, feeds T.sub.SS output block 92.

(58) Referring to FIG. 8, a flow diagram 102 including a calculation logic block 104 for calculating electronic control valve current (Ievc) transient torque in practicing the present invention expands on a step 106 of calculating Ievc corresponding with step 56 of FIG. 6. An electronic control valve current (from measurement of current applied to ECV) transducer (I) 108 feeds an input to a logic block 110 which operates to calculate Iecv rate of change (ROC), and outputs a rate of control valve current change (I) to an input of a yes/no logic block 114. A maximum threshold rate of change of ECV current sensor (I.sub.Lim) 112 feeds a second input to yes/no logic block 114. Logic block further functions, at logic block 120 (in phantom) to check if Iecv is changing. If I<I.sub.Lim, a first output to a logic block 116 equates electronic control valve transient torque (T.sub.IT) to second input of logic block 1300, wherein there is no I.sub.evc transient which, in turn, feeds a T.sub.IT output logic block 118. If II.sub.Lim, a second output of yes/o logic block 114 feeds a first input of a following logic block 122. Logic block operates to calculate Pd rate of change (ROC). A Pd pressure transducer 124 feeds a second input of logic block 122, which outputs a rate of discharge pressure change (Pd) to a first input of a yes/no logic block 126. A minimum threshold rate of discharge pressure (Pd.sub.limi) transducer 128 feeds a second input of yes/no logic block 126. If Pd<Pd.sub.Liml, a first output of yes/no logic block 126 feeds a first input of another yes/no logic block 134. Yes/no logic block 126 and 134 mutually check if Pd is changing charging. If Pd Pd.sub.Liml, a second output of yes/no logic block 126 feeds a first input of a logic block 130. A maximum threshold rate of discharge pressure (Pd.sub.Limh) transducer 136 feeds a second input of logic block 134. If Pd<Pd.sub.Limh, a first output of yes/no logic block 134 feeds a logic block 138 which equates electronic control valve transient torque (T.sub.IT) with zero, i.e. no T.sub.IT transient, which, in turn, provides an output to output logic block 118. If Pd Pd.sub.Limh, a second output of yes/no logic block 134 feeds the input of logic block 130. A transducer 132 feeds calibratable parameters F.sub.0 and F.sub.1 to a second input of logic block 130 which, in turn, feeds the output logic block 118. A common logic block 140 (in phantom) checks yes/no logic blocks 126 and 134 to determine if Pd is changing.

(59) Referring to FIG. 9, a flow diagram 142 for calculating engine revolutions per minute (ERPM) transient torque (T) in practicing the present invention expends on a step 146 of calculating ERPM transient torque corresponding with logic step 54 of FIG. 6. A logic block 150 which operates to calculate ERPM rate of change (ROC), and outputs a rate of engine RPM change (ERPM) to an input of a yes/no logic block 152. A maximum threshold rate of ERPM (ERPM.sub.lim) transducer 156 feeds a second input to yes/no logic block 152. If ERPM<ERPM.sub.Lim, yes/no logic block 152 feeds a logic block 154 which equates engine RPM transient torque (T.sub.ERPMT) to zero which, in turn, outputs a feed to an output logic block 166. A phantom logic block 158 coupled to yes/no logic block 152 checks if ERPM is changing. If ERPMERPM.sub.Lim, yes/no logic block 152 feeds a first input to a present time interval start timer logic block 160, which also receives a T.sub.SS second input. Logic block 160 provides an output to a following logic block 162 to save T.sub.SS at a present time interval (t.sub.i), which, in turn, feeds a first input of a logic block 164 in which T.sub.ERPMT=T.sub.SS when the timer is started and incrementally updated. The output of logic block 164 feeds the output logic block 166. A timer value for a T.sub.ERPMT calculation logic block 168 feeds logic block 164. The output of logic block 160 also feeds a first input of a yes/no logic block 170. A t.sub.Lim transducer 172 feeds a second input of the yes/no logic block 170. If t<t.sub.Lim, yes/no logic block 170 feeds a first input of a no ERPM transient logic block 176 which, in turn, feeds the output logic block 166. A T.sub.SS input feeds a second input of logic block 176. Yes/no logic block 170 is controlled by phantom logic block 174 to determine if the timer has expired.

(60) Referring to FIG. 10, a flow diagram 178 including a calculation block 180 for selecting a final torque (T.sub.F) value in practicing the present invention expands on a step 182 of selecting a final torque (T.sub.F) corresponding with step 52 of FIG. 6. Separate inputs comprising T.sub.SS, T.sub.IT and T.sub.ERPMT are fed into serially coupled yes/no logic blocks 184, 190, 194 and 198. If T.sub.ERPMT=0 and T.sub.IT=0, yes/no logic block 184 feeds a logic block 186 which equates T.sub.F and T.sub.SS which, in turn feeds an output to an output logic block 188. If T.sub.ERPMT0 or T.sub.IT0, a second output of yes/no logic block 184 feeds second yes/no logic block 190. If T.sub.ERPMT=0 and T.sub.IT0, yes/no logic block 190 feeds a logic block 192 which equates T.sub.F and T.sub.IT which, in turn, feeds an output to output logic block 188. If T.sub.ERPMT0 or T.sub.IT=0, a second output of yes/no logic block 190 feeds a third yes/no logic block 194. If T.sub.ERPMT0 and T.sub.IT=0, yes/no logic block 194 feeds a logic block 196 which equates T.sub.F with T.sub.ERPM which, in turn, feeds an output to output logic block 196. If T.sub.ERPMT=0 or T.sub.IT0, a second output of yes/no logic block 194 feeds a fourth yes/no logic block 198. If T.sub.ERPMT0 and I.sub.IT0, yes/no logic block 198 feeds a logic block 200 which equates T.sub.F and T.sub.IT which, in turn, feeds an output to output logic block 188.

(61) Referring to FIG. 11, a schedule of definitions and associated instrumentation, including definitions of terms, calibration parameters and sensor inputs, is provided to assure a clear understanding of the forgoing description of the invention.

(62) While this invention has been described in terms of the preferred embodiments thereof, it is not intended to be so limited, but rather only to the extent set forth in the claims that follow.

(63) It is to be understood that the invention has been described with reference to a specific embodiment with variations to provide the features and advantages previously described and that the embodiment is susceptible of modification as will be apparent to those skilled in the art.

(64) Furthermore, it is contemplated that many alternative, common inexpensive materials can be employed to construct the basic constituent components. Accordingly, the forgoing is not to be construed in a limiting sense.

(65) The invention has been described in an illustrative manner, and it is to be understood that the terminology, which has been used is intended to be in the nature of words of description rather than of limitation.

(66) Obviously, many modifications and variations of the present invention are possible in light of the above teachings. It is, therefore, to be under stood that within the scope of the appended claims, wherein reference numerals are merely for illustrative purposes and convenience and are not in any way limiting, the invention, which is defined by the following claims as interpreted according to the principles of patent law, including the Doctrine of Equivalents, may be practiced otherwise than is specifically described.